Identities in Action: Philip S. Brenner Jan E. Stets Richard T. Serpe Editors

Frontiers in Sociology and Social Research 6
Series Editor: Richard T. Serpe
Philip S. Brenner
Jan E. Stets
Richard T. Serpe Editors
Identities
in Action
Developments in Identity Theory
Frontiers in Sociology and Social Research
Volume 6
Series Editor
Richard T. Serpe, Department of Sociology, Kent University, Kent, OH, USA
Frontiers of Sociology and Social Research is a cutting-edge social science book
series focusing on new directions in sociological and broader social science research.
These new directions could be novel theoretical paradigms, developing topical areas
of research, innovative methodologies, and/or substantive findings that exemplify
and anticipate trends in subfields. The series is predicated on the observation that any
field of knowledge in contemporary times is a dynamic rapidly changing body of
perspectives and understanding that continuously builds upon the foundation of
extant scholarship. The series encourages manuscript submissions from both new
and established scholars of sociology, anthropology, social policy, and other allied
disciplines.
More information about this series at http://www.springer.com/series/8690

Philip S. Brenner • Jan E. Stets • Richard T. Serpe
Editors
Identities in Action
Developments in Identity Theory
Editors
Philip S. Brenner Jan E. Stets
Department of Sociology Department of Sociology
University of Massachusetts Boston University of California
Boston, Massachusetts, USA Riverside, California, USA
Richard T. Serpe
Department of Sociology
Kent State University
Kent, Ohio, USA
ISSN 2523-3424 ISSN 2523-3432 (electronic)

Frontiers in Sociology and Social Research
ISBN 978-3-030-76965-9 ISBN 978-3-030-76966-6 (eBook)
https://doi.org/10.1007/978-3-030-76966-6
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland
AG 2021
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by
similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Contents
1 Introduction: An Overview of Identities in Action . . . . . . . . . . . . . . 1

Philip S. Brenner, Jan E. Stets, and Richard T. Serpe
Part I Ethnicity and Race

2 Ethnic Identity Measurement and Verification . . . . . . . . . . . . . . . . 11
Peter J. Burke and Michael M. Harrod
3 Ethnic Identity Achievement, Identity Verification, and
Self-worth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Matthew Grindal, Melanie Kushida, and Tanya Nieri
4 Racial Identity Among Blacks and Whites in the U.S. . . . . . . . . . . . 61
K. Jill Kiecolt, W. Carson Byrd, Hans Momplaisir,
and Michael Hughes
5 White Racial Identity and Reparations for Slavery . . . . . . . . . . . . . 85
Ashley V. Reichelmann and Matthew O. Hunt
6 Public Confederate Monuments and Racial Identity among
White Americans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Ryan D. Talbert and C. André Christie-Mizell
Part II Family, Religion, and Work

7 The Mother Identity and Self-Esteem . . . . . . . . . . . . . . . . . . . . . . . 133
Jan E. Stets and Juha Lee
8 From Stoke to Swaddle: Surfer-Dads, Voluntary and Obligatory
Identities, and Identity Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Andrew Quinn
v
vi Contents
9 Religious Identity in a Proximate Social Structure: Mothers,

Fathers, and the Religious Socialization of Their Children . . . . . . . 179
Philip S. Brenner
10 Identity, Attribution, and Self-Esteem: When Employment
Status Matters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Michael M. Harrod and Richard T. Serpe
11 General and Specific Sources of Reflected Appraisals . . . . . . . . . . . 223
Mary Gallagher
Part III Networks, Homophily, and the Physical Environment

12 Does the Self Reflect Society? Examining the Relationship
Between Personal Network Assortativity and Role-Based
Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Mark H. Walker
13 The Structure of Friendship: How Network Density Influences
Identity Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Kelly L. Markowski
14 Using Identity Theory to Understand Homophily in Groups . . . . . . 285
Jan E. Stets, John Aldecoa, Quinn Bloom, and Joel Winegar
15 Relational Positioning As Intermediate Social Structure in
Identity Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
Timothy R. Rose and Philip S. Brenner
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
List of Contributors
John Aldecoa Department of Sociology, California State University, Sacramento,

CA, USA
Quinn Bloom Department of Sociology, University of California, Riverside, CA,
USA
Philip S. Brenner University of Massachusetts, Boston, Boston, MA, USA
Peter J. Burke University of California, Riverside, Riverside, CA, USA
W. Carson Byrd University of Michigan, Ann Arbor, MI, USA
C. Andre Christie-Mizell Department of Sociology, Vanderbilt University,
Nashville, TN, USA
Mary Gallagher Kent State University, Stark, Stark, OH, USA
Matthew Grindal University of Idaho, Moscow, ID, USA
Michael M. Harrod Central Washington University, Ellensburg, WA, USA
Michael Hughes Virginia Tech, Blacksburg, VA, USA
Matthew O. Hunt Department of Sociology, Northeastern University, Boston,
MA, USA
K. Jill Kiecolt Virginia Tech, Blacksburg, VA, USA
Melanie Kushida University of California, Riverside, Riverside, CA, USA
Juha Lee University of California, Riverside, Riverside, CA, USA
Kelly L. Markowski University of Nebraska-Lincoln, Lincoln, NE, USA
Hans Momplaisir Trinity Washington University, Washington, DC, USA
Tanya Nieri University of California, Riverside, Riverside, CA, USA
Andrew Quinn University of California, Riverside, Riverside, CA, USA
vii
viii List of Contributors
Ashley V. Reichelmann Virginia Tech, Blacksburg, VA, USA

Timothy R. Rose Kent State University, Kent, OH, USA
Richard T. Serpe Kent State University, Kent, OH, USA
Jan E. Stets University of California, Riverside, Riverside, CA, USA
Ryan D. Talbert Department of Sociology, University of Connecticut, Storrs, CT,
USA
Mark H. Walker Louisiana State University, Baton Rouge, LA, USA
Joel Winegar Department of Sociology, University of California, Riverside, CA,
USA
Chapter 1
Introduction: An Overview of Identities
in Action
Philip S. Brenner, Jan E. Stets, and Richard T. Serpe
“Identity” is an important concept in the social sciences and humanities, widely

employed by sociologists, psychologists, economists, political scientists, philoso-
phers, and other academics, as well as journalists and policymakers. Scholars use the
concept “identity” to connote how we think about ourselves, and how others think
about us. It identifies those with whom we see ourselves as similar as well as those
with whom we see ourselves as different. Often, the conceptualizations of identity
differ because there is not agreement on how to think about identities (Brubaker &
Cooper, 2000). Political scientists see identities in terms of political affiliations and
orientations (e.g., Egan, 2020; Green et al., 2002). Psychologists view identities
primarily in terms of the groups to which individuals belong and those to which they
do not belong (e.g., Hogg, 2014; Hogg et al., 2017). Economists understand
identities as brands that link consumers to their choices in the marketplace (e.g.,
Davis, 2020; Kranton, 2016). Many sociologists use the term identity when studying
race/ethnicity, class, and gender, and their intersections, as classifications, distinc-
tions, or labels (e.g., Dill & Zambrana, 2009; Telles & Paschel, 2014). Although
these different conceptualizations vary in their scope, each lacks a clear, systematic,
and thorough understanding as to what is an identity, how it operates within the self
and in situations, and what implications it has for the self and society over time.
The weakness of these approaches to identity is reflected in the verbs often used
to discuss identities. Identities are described as something one “is,” an integral, hard
P. S. Brenner (*)
Department of Sociology, University of Massachusetts Boston, Boston, MA, USA
e-mail: [email protected]
J. E. Stets
Department of Sociology, University of California, Riverside, CA, USA
R. T. Serpe
Department of Sociology, Kent State University, Kent, OH, USA
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 1

P. S. Brenner et al. (eds.), Identities in Action, Frontiers in Sociology and Social
Research 6, https://doi.org/10.1007/978-3-030-76966-6_1
2 P. S. Brenner et al.
wired, or stable characteristic. Identities also are described as something one “has,” a
possession to be owned, held, and kept. However, individuals are not simply an
identity, nor do they hold a single identity. Rather, people maintain and enact many
identities. Moreover, identities exist in an ecology of identities within the self. These
identities help constitute a stable self, while at the same time are a dynamic feature of
the self as they are enacted, assessed, supported, altered, or abandoned within and
across situations.
Identity theory, the focus of this volume, investigates the dynamic process of
identities in action. How do identities work when they are engaged with the identities
of other individuals in situations? Do they work together, are they independent of
each other, or do they exist in a state of bitter conflict? Further, how are identities
conceptualized within the self, and how does this conceptualization relate others’
conceptualizations of one’s identities in situations? Here, the issue of identity
verification becomes important.
Identity theory provides a clear, thorough and systematic conceptualization as to
what are identities, and how we can understand them in action (Burke & Stets, 2009;
Stryker, [1980] 2002). In this volume, we demonstrate the usefulness of this
theoretical approach. Identity theory seeks to explain how identities work in a variety
of ways and in a variety of settings. Its careful and systematic approach is useful for
researchers inside and outside of sociology whose work focuses on identities either
in a central or peripheral way. Identity theory can be of use to researchers studying
racial and ethnic issues, identity politics and partisanship; identity development
across the life course; brand loyalty and brand identity, and more.
This introductory chapter provides an overview of some of the important findings
that emerge in this volume. Most of the chapters were presented and discussed at the
Third Biennial Conference on Identity Theory and Research held in 2018 at the
University of Massachusetts, Boston. The chapters covers a wide array of issues. The
volume is organized into three parts on the theme of identities in action. They include
ethnicity and race; family, religion, and work; and networks, homophily, and the
physical environment.
The first part of the volume features five chapters that apply and extend identity
theory to ethnicity and race. These chapters stay true to the volume’s focus on action
by examining the crucial outcomes linked to these identities, such as action for or
against racial justice.
In Chap. 2, Burke and Harrod develop and test a measure of ethnic identity.
Employing a mixed-method approach in which 775 ethnically diverse college
students first describe in their own words what ethnicity and race means to them,
they then use respondents’ descriptions to create an ethnic identity scale. This scale
has two dimensions: a heritage dimension that honors the traditions and culture of
the ethnicity, and a personal dimension that involves enacting one’s ethnicity such as
maintaining ties with friends and neighbors who share one’s ethnic identity and
showing an outward appearance of one’s ethnicity such as wearing ethnically
appropriate clothing. Using the ethnic identity scale, they then examine some out-
comes of the ethnic identity such as whether ethnic identity non-verification reduces
self-esteem, and it does, lowering worth, efficacy, and authenticity-based
1 Introduction: An Overview of Identities in Action 3
self-esteem; whether the effects of non-verification occur uniformly for different

ethnic groups, and they do, for Whites, Blacks, Latinos, and Asians; and where there
is a lower sense of ethnic identity across generations, and there is, with a reduced
sense of one’s ethnic identity for third generations respondents compared to first and
second-generation respondents. The important contribution that this chapter makes
to identity theory is a measure of one’s ethnic identity that can be employed in future
research.
In Chap. 3, Grindal, Kushida, and Nieri examine how ethnic identity achieve-
ment, that is, the process where people develop a secure and stable sense of their
ethnic group, increases individuals’ feelings of self-worth. They examine identity
verification as an important theoretical mechanism that helps establish this relation-
ship by analyzing data from 1500 ethnically diverse college students. Turning to
principles in identity theory, they argue that the process of developing a stable ethnic
group involves people entering homogenous proximate social structures in which
they develop ties to others with the same ethnic identity, this produces greater ethnic
identity verification for individuals, an increase in feeling good about their ethnic
group, and an increase in feeling good about themselves, more generally. Their
findings support these theoretical relationships, which hold for men and women, as
well as White, Black, Latino, and Asian respondents. In general, we see how
important the verification process is in fostering positive evaluations of others in
one’s group as well as oneself.
Kiecolt, Byrd, Momplaisir, and Hughes draw upon identity theory, social identity
theory, and life course theory to understand the functioning of the racial identities of
Black and White Americans in Chap. 4. They analyze data from the 2014 Identity
Module of the General Social Survey (GSS) and examine different dimensions of the
racial social identity. These dimensions include the importance, salience, verifica-
tion, pride, and public respect individuals might obtain from their racial identity.
They find that, compared to Whites, the racial identity of Blacks is more important
and salient to them, more likely to be verified, a source of more pride, but also per-
ceived to receive less public respect. Consistent with life course theory, they reveal
that one’s racial identity is stronger among those from the civil rights era compared
to those from the post-civil rights era. Additionally, the racial identity of Blacks
compared to Whites is more important, salient, and a greater source of pride. Overall,
we see that one’s racial social identity appears to be more relevant in the lives of
Blacks compared to Whites.
Reichelmann and Hunt in Chap. 5 also use the 2014 Identity Module of the GSS
to examine the different dimensions of the white racial identity as it relates to
attitudes regarding actions designed to repair the harm caused by slavery and its
subsequent vestige of racial inequality. They examine a wide range of actions that
involve (1) explicit reparations that range from the symbolic (an apology) to material
(financial payments); (2) government intervention that improves racial inequalities
such as university scholarships for black students and preferential hiring and pro-
motion of blacks; and (3) taking personal responsibility for improving race relations.
They find that pride in being white is associated with increased opposition toward
reparation policies. Alternatively, the feeling that whites are respected as a group in
society is associated with decreased opposition or increased support toward repara-
tion such as a willingness to support symbolic and material reparative actions and
take responsibility for improving race relations. These findings suggest that identity
researchers may need to focus more attention on group pride and respect as it appears
to impact how responsive individuals are to improving racial inequality in society.
In Chap. 6, Talbert and Christie-Mizell also examine the white racial identity as it
relates to the history of slavery by focusing on individuals’ responses to the presence
of county and state Confederate monuments. Relying on a nationally representative
sample of almost 1000 respondents, they find that the greater the number of
Confederate monuments at the state level, the more likely it is that individuals will
claim a prominent white racial identity. The researchers argue that Confederate
monuments legitimize the ideology of black inferiority and white supremacy and
foster the idea that slavery was secondary to the Confederates efforts, thereby
strengthening a positive white group identity. Their research reveals how cultural
artifacts can serve as a reminder of boundaries between groups, thereby reinforcing
intergroup conflict and the importance of one’s own racial group identity.
The second part of the volume focuses on identities linked to the roles people
perform in their everyday lives, such as being a parent or a worker. It also investi-
gates how these role identities are related to and influence other identities people
hold and enact.
In Chap. 7, Stets and Lee study how the prominence, salience, and verification of
the mother identity relates to the self-esteem outcomes of feeling competent or
effective (efficacy-based esteem) and feeling valued or accepted in the mother role
(worth-based esteem). Drawing on a sample of 100 low-income, Hispanic mothers,
they find that only the verification process plays a significant role in how mothers
feel about themselves. Those who experience identity verification in the mother role
report an increase in efficacy and worth-based esteem. Stets and Lee discuss how
mothering is not simply about doing or about being competent in the mother role,
which is often emphasized in family research, but it also is about being or feeling
worthy or a sense of belonging to something larger than themselves such as being a
part of a family. Whether verification of the mother identity increases worthiness and
acceptance in other races/ethnicities beyond Hispanics is an empirical question. In
the Hispanic culture, familism is valued including large families, family solidarity,
and close bonds with extended kin. When the mother identity is verified, this may
have a greater positive impact on belongingness for Hispanic women than other
racial/ethnic groups in society.
Quinn focuses on the father identity in Chap. 8. Based on interviews with
15 surfer-dads, he traces how the surfer identity helps shape other identities the
surfers have such as the spouse identity and worker identity. In some instances, men
prioritize the surfer identity relative to their spouse and worker identity. Perhaps
more importantly, Quinn reveals how the surfer identity, a voluntary identity, affects
taking on the father identity, an obligatory identity. For example, while the men see
the surfer identity as carrying the meaning of being “selfish” in which surfing
facilitates meeting their own needs, satisfying these needs for several years in the
surfer identity appears to make the men ready to take on the challenges associated
with the more “selfless” father identity in which the needs of a newborn become
paramount. Over time, the father identity is prioritized over the surfer identity, and
rather than focusing on a loss of their surfer identity, the men enjoy the addition of
the father identity into their lives. In general, Quinn’s work provides a nuanced
understanding of voluntary and obligatory identities, and how identity change is
experienced.
The types of identities individuals assume as they develop, and the meanings of
these identities are often influenced by how mothers and fathers socialize their
children. In Chap. 9, Brenner examines the intergenerational transmission of iden-
tities through an analysis of the religious identity. Relying on a sample of over
500 middle-class American families from eight communities in different regions of
the country, Brenner finds that the primary way children are socialized into their
religious identity is through their mothers’ religious identities. Children have a
prominent and salient religious identity and engage in religious behavior when
their mothers have a prominent and salient religious identity and behave in religious
ways as well. Thus, compared to fathers, mothers appear to play a larger role in the
religious development of their offspring.
In Chap. 10, Harrod and Serpe explore the relationship between the worker
identity, attributions, and self-esteem. Using a sample of 640 employed and
160 unemployed people, they look at how those who hold the normative identity
as employed differ from those who hold the counter-normative identity as unem-
ployed. Harrod and Serpe examine if effects of causal attributions to ability, effort,
task difficulty for those who are employed and unemployed on identity prominence
and self-esteem differ. Using the two-dimensions of the Rosenberg Self-esteem
scale, self-confidence, and self-denigration, they again look to see if those employed
and unemployed differ in the effects of identity prominence on the two self-esteem
dimensions. They find that the effects of attribution on identity are modest. Only
attributions to ability are significant on identity prominence and the effects do not
differ for those employed and unemployed. The effects of identity prominence on
self-esteem demonstrate a significant pattern that differs for the employed and
unemployed. For those who are employed, identity prominence increases self-
confidence but does not show a significant effect on self-denigration. The unem-
ployed pattern shows that identity prominence does not significantly effect self-
confidence but significantly lowers the level of self-esteem on self-denigration.
Overall, findings do not demonstrate a clear relationship between attributions of
identity prominence, however, they suggest that when there is a theoretical reason
for analyzing the two dimensions of Rosenberg’s Self-esteem scale, it may provide a
clearer picture of how people evaluate their self-esteem.
The second section of the volume ends with an examination by Gallagher in
Chap. 11 of the relative strength of specific others (such as close others including
family and friends) and general others (non-specific persons who individuals may
think of) as a source of feedback as to who individuals are across four identities: the
spouse/partner, friend, worker, and religious/spiritual identity. At issue is whether
each source (specific or general others) has an influence on how individuals see
themselves, or whether one source of feedback, such as the feedback from specific
others, has a stronger influence on one’s self-views than feedback from general
others. Examining data on about 500 adults in the U.S., Gallagher finds, in general,
that both sources of feedback positively influence how others evaluate themselves in
each of their identities. For example, the more positively respondents feel that
specific and general others evaluate them in their spouse, worker, and religious
identities, the more positively respondents see themselves in those identities. The
results support the important symbolic interactionist tenet that feedback from others
is an important ingredient in how individuals see themselves. Indeed, at least for
some central identities, the source of the feedback, whether it be from specific or
general others, is relatively less important.
The chapters in the third section of the volume extend and apply identity theory to
the social and physical environments. We see how one’s social and physical
surroundings shape identities and are shaped by identities.
In Chap. 12, Walker focuses on the intersection between identity theory and
social network analysis by investigating individuals’ personal networks within and
between five role-based groups: family members, coworkers, church contacts,
schoolmates, and friends. Relying on a sample of over 400 students at a large
midwestern university, he examines whether individuals are more likely to be tied
to one another within their role-based group than outside of their role-based group.
He finds that this occurs. Work-based ties have the highest level of homophily while
religion-based ties have the lowest level of homophily. However, he also finds that
there is considerable overlap in social relations between role-based groups. For this
reason, intragroup and intergroup relations may both be important in understanding
the salience of identities. If commitment to an identity is a function of who one is
connected to in the social structure, then Walker’s research suggests we consider
within as well as between group ties as structuring people’s identities.
Like Walker, Markowski, in Chap. 12, focuses on how social networks can be
brought into identity theory. In particular, she is interested in the social context in
which identity verification occurs by analyzing the density of one’s social networks,
or the proportion of all of one’s social contacts that are interconnected. Drawing on a
sample of almost 300 respondents, she finds that higher levels of density buffer the
negative emotions that result from identity non-verification. Dense networks often
are portrayed as having strong emotional bonds, thus the social support that they
offer may be one of the defining mechanisms in offsetting the negative feelings of
non-verification. Additionally, dense networks are likely characterized as mutually
verifying, such that non-verifying experiences may be minimized over the greater
preponderance of verifying experiences.
In Chap. 13, Stets, Aldecoa, Bloom, and Winegar theorize about homophily in
groups through the lens of identity theory. They maintain that when individuals
share similar meanings on an identity, homophily emerges, and the more identity
meanings are shared, the stronger should be the homophily. They also argue that
what is important in shared meanings creating homophily is identity verification.
Verification should increase interaction and encourage homophily. Identity
verification also should increase the prominence and salience of identities in

homophilous groups. Finally, they maintain that person identities are more likely
to be an influential aspect of homophilous groups than role or social/group identities
because as master identities, since person identities are instrumental in shaping role
and social/group identities, they also should be instrumental in shaping the devel-
opment of homophilous groups. Stets and her colleagues apply their theoretical ideas
to discuss how the meanings associated with the moral person identity may influence
individuals’ political identity and be an important basis of partisan politics in
the U.S.
The final chapter by Rose and Brenner study the influence of the physical
environment on identities. They investigate how relational positioning, that is, the
distance between locations where identities are enacted, and the time or effort it takes
to traverse a distance facilitates or constrains contact with others who share one’s
identities. They examine data on about 1200 respondents and use the religious
identity as an example to study the effects of the physical environment. They find
that as relational positioning increases, the frequency of reaching the location of
religious identity enactment is reduced, and the composition of one’s daily religious
contacts becomes less homogenous. In turn, less homogenous social contacts reduce
the prominence of one’s identity. Thus, we see the impact that the physical envi-
ronment has on the centrality and enactment of one’s identities.
In general, this volume advances identity research in several important ways. We
get a deeper and richer understanding of racial and ethnic identities. Racial and
ethnic relations are a frequent topic in current national conversations. Given that by
the middle of this century whites will no longer comprise the majority in our society
(United States Census Bureau, 2012), identity researchers need to continue to
advance theoretical and empirical work on the diverse identities that represent our
society. And, as we continue to live our lives balancing work and family identities
and within a complex network of others, we need to more closely study the different
ways that the social and physical world shapes the meanings as to who we are.
References
Brubaker, R., & Cooper, F. (2000). Beyond ‘identity’. Theory and Society, 29, 1–47.
Burke, P. J., & Stets, J. E. (2009). Identity theory. Oxford University Press.
Davis, J. B. (2020). The status of the concept of identity in economics. Forum for Social
Economics, 50, 1–9.
Dill, B. T., & Zambrana, R. E. (2009). Critical thinking about inequality: An emerging lens. In B. T.
Dill & R. E. Zambrana (Eds.), Emerging intersections: Race, class, and gender in theory,
policy, and practice (pp. 1–21). Rutgers.
Egan, P. J. (2020). Identity as dependent variable: How Americans shift their identities to align with
their politics. American Journal of Political Science, 64, 699–716.
Green, D. P., Palmquist, B., & Schickler, E. (2002). Partisan hearts and minds: Political parties
and the social identities of voters. Yale University Press.
Hogg, M. A. (2014). From uncertainty to extremism: Social categorization and identity processes.
Current Directions in Psychological Science, 23, 338–342.
Hogg, M. A., Abrams, D., & Brewer, M. B. (2017). Social identity: The role of self in group
processes and intergroup relations. Group Processes and Intergroup Relations, 20, 570–581.
Kranton, R. E. (2016). Identity economics 2016: Where do social distinctions and norms come
from? American Economic Review, 106, 405–409.
Stryker, S. [1980]. (2002). Symbolic Interactionism: A Social Structural Version. Blackburn Press.
Telles, E., & Paschel, T. (2014). Who is black, white, or mixed race? How skin color, status, and
nation shape racial classification in Latin America. American Journal of Sociology, 120,
864–907.
United States Census Bureau. (2012). U.S. Census Bureau Projections show a slower growing,
older, more diverse nation a half century from now. https://www.census.gov/newsroom/
releases/archives/population/cb12-243.html
Part I
Ethnicity and Race
Chapter 2
Ethnic Identity Measurement
and Verification
Peter J. Burke and Michael M. Harrod
Introduction
The social psychological components of actors’ racial/ethnic1 identity represent a

fertile, but heretofore largely under theorized, component of both the race/ethnicity
and social psychological literatures. Within this paper we move toward strengthen-
ing both traditions by highlighting what each offers the other. To date, a considerable
amount of the race/ethnicity literature has focused on how large-scale factors such as
immigration patterns, ethnic replenishment, discrimination and assimilation affect
actors (Jimenez, 2008; Lee & Bean, 2004; Portes & Rumbaut, 2001; Waters &
Jimenez, 2005). However, this literature has been remiss to fully address the internal
aspects of one’s ethnic identity and provide a coherent theoretical approach across
studies.
The social psychological literature, on the other hand, has faced the opposite
problem; considerable attention has been given to the internalized functioning of
identities, but suffers from a paucity of effort to understand potential racial/ethnic
Prepared for the 2018 Identity Conference, Boston, MA.

1
The terms race, race/ethnicity, and ethnicity will be used interchangeably throughout this paper.
Nevertheless, we recognize that race and ethnicity are sometimes conceptualized as separate and/or
overlapping concepts. We have chosen to follow recent scholarship arguing that distinguishing
between these terms fails to account for respondents’ lived experiences (Brown et al., 2006; Hitlin
et al., 2007; Lee & Bean, 2004; Wimmer, 2008).
P. J. Burke (*)
M. M. Harrod
Department of Sociology, Central Washington University, Ellensburg, WA, USA

12 P. J. Burke and M. M. Harrod
differences. Hunt and colleagues (Hunt et al., 2000) argue that the dearth of social
psychological theory and research investigating possible racial/ethnic differences
implicitly assumes the theories are invariant to racial/ethnic differences. However, as
scholars of race/ethnicity have clearly shown, the foundational place that racial/
ethnic meaning holds within American society makes this an assumption in need of
exploration (Harris & Sim, 2002; Lee & Bean, 2004; Waters, 1990; Waters &
Jimenez, 2005).
This oversight is even more conspicuous when considering the overarching
salience that race/ethnicity plays in society, where race/ethnicity is argued to repre-
sent a “master status” influencing most other identities an individual holds and it
influences how others treat the individual (Lacy, 2007; Stryker, 1980; Waters, 1990).
Moreover, it is argued that a decisive component of American culture includes a
meaning system where race and ethnicity act as central organizing principles within
American society (Omi & Winant, 1994). These differences are associated with the
stratification system, with those of European descent generally being favored. The
effect is that some groups (e.g., Whites) have been afforded greater rights, better
treatment and more privileges than other groups (e.g., African Americans or Latinos)
(Bean & Stevens, 2005; Bonilla-Silva, 1997; Nagel, 1994; Pager & Shepherd, 2008).
The research reported here allows us to begin addressing both limitations.
Viewing persons’ ethnic identity from an identity theory perspective provides a
theoretical explanation of the internal dynamics of how one’s racial/ethnic identity
enactment, maintenance and action unfolds, which informs the racial/ethnic identity
literature. It also allows us to begin exploring how members of different racial/ethnic
groups understand and enact their ethnic identity. As such, it is an opportunity to
evaluate how well identity theory functions across different racial/ethnic groups. In
this way, it allows us to examine the heretofore untested assumption that identity
theory’s theoretical predictions work equally well regardless of one’s racial/ethnic
background. The third goal of this research is to more deeply explore a social or
group-based identity instead of a role or person-based identity, in this case, one’s
racial/ethnic identity. To date, most of the empirical work of identity theory has
focused on the role and person bases of identity. We seek to broaden this scope and
application by empirically evaluating a social or group-based identity (Burke &
Stets, 1999; Stets & Burke, 1994; Stryker & Serpe, 1994). We begin with a brief
review of the theoretical orientation employed.
Theory
Identity Theory
Identity theory traces its roots to the structural symbolic interactionist framework
emphasizing how people acquire and internalize self-meanings, in the form of
identities, from their statuses, social groups and positions within the social structure
(Burke & Stets, 2009; Stryker, 1980, 1981). According to identity theory, persons
2 Ethnic Identity Measurement and Verification 13
develop a sense of self that is intricately tied to society. Holding a role or being in a
social group leads one to internalize the meanings and expectations for oneself that,
in turn, come to inform both the individual and others who the individual
is. Following these roots, identity theory argues that actors derive self-relevant
meanings by taking the self as an object as others would see him/her. Through this
self-reflective process, an individual acquires definitions of the self by looking at
oneself as a social object as others see the self (Cooley, 1964 [1902]; Mead, 1934).
Building on structural symbolic interaction theory, identity theory argues that
people have internalized what it means to be a member of their ethnic/racial group.
These identity meanings then guide the behaviors one undertakes to be consistent
with those self-meanings. Enacting identity behaviors consistent with the actor’s
identity comes to illustrate to self and others that these behaviors reflect what it
means to hold the enacted identity. It is this process that encompasses one of identity
theory’s central tenets: meanings are manifest in behavior and practices (Burke &
Stets, 2009).
Perceptual Control Process
Identity theory provides an in-depth explanation of the multifaceted nature of an

individual’s identities ranging from how identities are organized, regulated, and
maintained to how they change and the subsequent implications for one’s psycho-
logical well-being (Burke & Stets, 2009). An identity is a set of meanings, held in the
identity standard, that define the individual—what it means to be, for example, a
teacher (role identity) or a Latino (group or category identity), or a person with
integrity (person identity) (Osgood et al., 1957). A core aspect of identity theory is
identity verification, the idea that people seek to have others view them as holding
the same meanings as they hold in their identity standards (Burke, 1991; Burke &
Stets, 2009). To achieve identity verification, people act to control their perception of
identity relevant meanings in the situation (often through reflected appraisals) to
match the meanings in their identity standard—a perceptual control system (Burke &
Cast, 1997; Burke & Stets, 2009; Powers, 1973; Stets & Burke, 2014a).
Self-meanings are held in an “identity standard” that serves as the referent for the
individual guiding her/his perceptions. When an identity is activated in a situation,
the identity becomes engaged in a feedback loop comprised of four main parts:
(1) the identity standard (the meanings one holds for an identity); (2) perceptual
input of self-relevant meanings from the situation (i.e., one’s perceptions of the
meaningful feedback that the self obtains from others in a situation); (3) a compar-
ator which compares a person’s perceptions of meanings in a situation with one’s
identity standard; and (4) the output (i.e., one’s behavior) to the environment as a
result of the comparison (Burke, 1991).
Once an identity is activated, people monitor perceived meanings in the situation
(how things are). The comparator receives these perceived meanings and compares
them to the meanings in the identity standard (how things are supposed to be). When
the perceived meanings within in the situation match one’s identity standard, this is
identity verification, which produces positive emotions as well as increased feelings

of self-worth and self-efficacy (Burke & Harrod, 2005; Burke & Stets, 1999; Cast &
Burke, 2002; Stets & Burke, 2014a). But when there is a discrepancy between the
perceived meanings from the situation and the meanings in the identity standard, this
is identity non-verification, which triggers negative emotions, distress and decreased
self-esteem. Consequently, people act to change meanings in the situation to reduce
the discrepancy and make them match their identity standard.
Ethnic Identity
People come to define themselves or derive self-meanings from the positions and
groups they identify with (Deaux & Martin, 2003; Rosenberg, 1979; Stets & Burke,
2000; Stryker, 1980). The culturally shared meanings held within an actor’s identity
standard for a social or group-based identity is expressed by one’s actions and
interpreted by others who react towards the person’s actions (Alba, 1990; Brown
& Lesane-Brown, 2006; Hitlin et al., 2007; Hogg, 2018; Jimenez, 2004, 2008; Lee &
Bean, 2007). The internalized amalgam of a group’s cultural meanings and one’s
own idiosyncratic meanings composes one’s ethnic identity standard. Social cate-
gories define group members’ group-based identity by (1) defining group boundaries
where some are categorized as “in-group” members while others are defined as “out-
group” members and (2) constructing group norms for appropriate expressions of
social group meanings that are internalized as part of an individual’s identity
standard (Brown et al., 2006; Brown & Lesane-Brown, 2006; Hitlin et al., 2007;
Hogg, 2018; Jimenez, 2004; Lee & Bean, 2007).
Seeing oneself and being seen by others as a group member is especially
important for social or group-based identities because the behaviors and rituals
related to membership in an ethnic group distinguish group members from
non-members and “define an ethnic uniqueness worth maintaining” for the individ-
ual (Alba, 1990, p. 84; Jimenez, 2008; Stets & Burke, 2000). Therefore, these
actions, practices, and behaviors are understood from an identity theory perspective
as the observable manifestations of the self-meanings people continually regulate as
part of their racial/ethnic identity. This is where the race/ethnicity and social
psychological literatures intersect. One learns, understands, and hones their ethnic
identity from interacting with other in-group and out-group members and these
meanings become internalized over time reducing uncertainty in the actor’s ethnic
identity.
While one’s group identification is critical, it is not the only way someone is
identified as part of a group. Another important aspect of one’s ethnic identity
development is how someone is viewed and treated by others as belonging or not
belonging to a group (Frank et al., 2010; Jimenez, 2004, 2008; Khanna, 2004, 2010;
Portes & Rumbaut, 2006; Waters, 1999). For example, Waters (1999) discusses how
West Indian immigrants and their children actively resist efforts by others to see and
treat them as African Americans. Waters’ respondents highlighted that being seen
and treated as an African American is tied to negative stereotypes and leads to poor
treatment from others; identification as a West Indian, on the other hand, is linked to
positive stereotypes (e.g., being hard working) and results in better treatment from
others. Membership or identification with a group, however, does not mean that
there is only one ethnic experience. Based on an individual’s own experience and
position within the group, s/he may choose to emphasize certain aspects of the
group’s general meanings and/or other members of the group may impose bound-
aries on the actor shaping the final composition of his/her unique ethnic identity
standard (Deaux & Martin, 2003; Jimenez, 2004, 2008).
An important manner by which people come to develop their racial/ethnic
identity is by determining the symbolic boundaries of how their group is similar
to, and different from, other racial/ethnic groups (Howard, 2000; Jimenez, 2008;
Phinney, 2006; Reid & Hogg, 2005). Group norms define the group’s essential
meanings, which, in turn, inform group members of what characteristics, feelings,
and behaviors are appropriate to exhibit as a member of an ethnic group. Individual
group members then adopt these norms and meanings to fit in as part of their ethnic
identity. Opportunities to enact and perform one’s ethnic identity are important
dimensions of having the identity (Burke & Stets, 2009; Deaux & Martin, 2003).
However, as Jimenez (2008) illustrates, with an influx of first-generation immi-
grants, termed ethnic replenishment, second and later generation persons occasion-
ally struggle to enact their ethnic identity in ways considered authentic by more
recent immigrants. For example, Jimenez (2008) notes for Mexican Americans that
authentic identity performances are characterized by things such as knowledge of
Mexican history and the ability to speak Spanish. Poor performance on these tasks is
taken as “evidence” that someone is not truly Mexican American. Jimenez notes that
some respondents avoid certain locations to prevent having their authenticity as a
Mexican American challenged.
The race/ethnicity literature interestingly shows that when one’s racial/ethnic
membership is challenged, people’s actions and feelings mirror the predictions of
identity theory concerning people’s feelings and actions when they experience
identity non-verification (Burke, 1991; Burke & Cast, 1997; Burke & Harrod,
2005). Several studies within the racial/ethnic literature highlight the negative
feelings and actions people experience when their ethnic identity is not accepted
by others (Jimenez, 2004; Khanna, 2004, 2010; Waters, 1999). Waters (1999)
example of the various activities in which West Indian immigrants engage to not
be identified as African Americans illustrates this point well. These examples are
consistent with identity theory’s prediction that actors want others to see them as
they see themselves (i.e., identity verification). And when perceptions during inter-
action fail to match their self-view, the actor takes actions such as avoiding locations
where his/her identity is questioned. Linkages such as this between the race/ethnicity
and social psychological literatures bolster our attempt to illuminate identity theory’s
applicability for understanding people’s racial/ethnic identity enactment as well as
augmenting identity theory’s broader applicability.
Despite the extensive and insightful research tradition of race/ethnicity, there
have been few efforts to quantitatively measure ethnic identity. The few attempts to
do so come largely from psychology, rather than sociology. Moreover, the disparate
definitions and measures of ethnic identity in the literature have complicated efforts
to draw general conclusions (Phinney, 1990). Ethnic identity has been conceptual-
ized in a myriad of ways such as feelings of belonging and commitment (Singh,
1977; Ting-Toomey, 1981), a sense of shared values and attitudes (White & Burke,
1987), cultural knowledge (Alba, 1990), and language and behaviors (Alba, 1990;
Jimenez, 2008; Khanna, 2004; Rogler et al., 1980; Waters, 1990). To move toward
some level of agreement among researchers studying ethnic identity, Phinney
emphasized that there needs to be more “focus on the common elements that
apply to groups” (Phinney, 1990, p. 499, emphasis added). Phinney developed the
Multigroup Ethnic Identity Measure (MEIM) scale to assess the common dimen-
sions of meaning shared across different groups (Phinney, 1992). However, it is
unclear that there is a theoretical underpinning for the development of the 20 items
comprising the MEIM scale. The MEIM uses a mixture of one’s feelings for one’s
own group and other ethnic groups, the degree of clarity in one’s ethnic background,
and the respondent’s ethnic practices and behaviors.
More recently, the MEIM scale was revised drawing on Erikson’s (1968) work on
identity development and achievement. The revision focuses on the respondents’
sense of ethnic identity exploration (e.g. reading or learning cultural practices) and
commitment—factors believed to be germane to the development of one’s ethnic
identity (Phinney & Ong, 2007). This recent revision emphasizes a perspective in
which ethnic identity begins in childhood and is refined through early adulthood
until becoming stable as people reduce their level of uncertainty about who they are
and what it means. However, the revision has focused its attention on development
in the beginning stages of one’s life-course and has foregone the more holistic
conceptualization of ethnic identity found in the original MEIM scale.
Following the principles of identity theory, we developed measures of respon-
dents’ ethnic identity taking into consideration the recent revisions of the current
MEIM. Our questions focus on examining the meanings, expectations, and behav-
iors one has regarding his/her ethnic identity. These measures also represent a
complement to the more recent MEIM scale in that the resulting scale focuses our
understanding of ethnic identity in those areas that Phinney and colleagues suggest
should be stable.
The Research Questions
We explore two aspects of ethnic identity informed by identity theory. First, we

explore the meanings of one’s ethnic identity as manifest in various practices and
behaviors to construct measures of ethnic identity useful for different race/ethnic
groups across several immigrant generations. We also explore variations in ethnic
identity and levels of uncertainty about their identity across ethnic groups and
generations in the U.S. Additionally, we examine how the identity verification
process operates for a social or group-based identity like one’s ethnic identity rather
than a role or person identity as much prior work as explored; and we examine the
negative impact of nonverification on the three components of self-esteem and

whether that varies across ethnic groups and generation (Stets & Burke, 2014b).
We discuss each question in turn below.
Race/ethnicity, in part, defines one’s position within the social matrices of
American society, where one is afforded differential resources, status and prestige
(Bonilla-Silva, 1997, 2003; Hill Collins, 2000; Krysan & Bader, 2009; Pager &
Shepherd, 2008). This is consistent with prior identity theory research finding one’s
ability to access resources and status affects one’s ability to achieve identity verifi-
cation (Burke, 2008; Cast et al., 1999; Stets & Cast, 2007; Stets & Harrod, 2004).
For example, researchers have begun explicating how the amount of skin color
pigmentation differentially affects how people of color are treated in the U.S. (Bean
& Stevens, 2005; Blake et al., 2017; Frank et al., 2010; Keith et al., 2017). For
example, Bean and Stevens (2005) find that light-skinned immigrants have an easier
time being incorporated into American society than immigrants with darker skin.
Because racial/ethnic minority groups have historically been accorded fewer
resources, and are viewed as having fewer resources, we expect minority members
will experience greater difficulty verifying their various role identities (Burke, 2008).
On the other hand, some research suggests that ethnic minorities, especially blacks,
may have less difficulty than whites in verifying their minority identity rather than
other role identities (Hunt, 2003, 2007; Stets & Fares, 2019).
Since we are examining the verification of various ethnic identities rather than
other role identities, we argue that being a member of a racial/ethnic minority group
generally results in having fewer resources, and because of this, racial/ethnic minor-
ities should experience greater difficulty verifying their ethnic identity. Therefore,
we propose the following hypothesis:
H1: Members of racial/ethnic minority groups will experience more difficulty
verifying their ethnic identity than Whites.
We also expect generational status to affect the ability of individuals to verify
their ethnic identity (Kasinitz et al., 2004; Portes & Rumbaut, 2001, 2006).
Researchers argue that the straight-line assimilation of the pre-1965 immigrants,
who were largely European, does not reflect the experiences of these more recent,
largely non-European immigrants. Instead, as Portes and Rumbaut (2006) point out,
the children of racial/ethnic minority immigrants tend to selectively choose from
mainstream society, making past approaches less applicable to current conditions.
Generational status has been related to one’s immersion in American culture, which
has been linked with one’s self-esteem (Harker, 2001) and educational achievement
(Mouw & Xie, 1999). Second-generation individuals should have greater exposure
to mainstream culture than first-generation individuals do, but less exposure than
third-generation respondents.
Accordingly, we argue that the degree of immersion and/or experience one has
within American society should affect both the nature of one’s ethnic identity and the
difficulty one experiences verifying their ethnic identity. Second-generation respon-
dents can be caught between cultures; they experience life in fundamentally different
ways than their immigrant parents and more recent arrivals as well as
third-generation respondents (Gans, 1992; Jimenez, 2004; Kasinitz et al., 2004;

Waters & Jimenez 2005). Thus, second-generation individuals are caught in a
cultural crossfire trying to satisfy the expectations of both cultures simultaneously
and this can lead to greater difficulties achieving identity verification. Therefore, we
propose the following hypothesis:
H2: Individuals from the second-generation will experience greater difficulty
verifying their ethnic identities as compared to the other generations.
Researchers have defined both the level of knowledge about one’s ethnicity and
the certainty about the meanings of one’s ethnic identity as important components of
an ethnic identity (Phinney, 1992). Knowledge can be thought of as the familiarity
one has with his/her racial/ethnic background, history, and traditions. The greater a
person’s knowledge of their racial/ethnic background, the more coherent will be
their understanding and presentation of their ethnic identity. They should also have a
better sense of the appropriate meanings one is seeking to verify through his/her
actions. Khanna (2004) finds that respondents’ level of ethnically shared cultural
knowledge affects how group members react and treat claims of belonging to a given
group. Those who can demonstrate greater cultural knowledge are treated as belong-
ing to the group, and this positive treatment from other group members, increases the
actor’s sense of legitimacy as a group member. On the other hand, Khanna finds that
when people are unable to demonstrate their cultural knowledge, they are treated as
an inauthentic member of the racial/ethnic group. This leads us to propose the
following hypothesis:
H3: The more knowledge an individual has about his/her ethnicity, the less
difficulty one will experience in verifying one’s ethnic identity.
Identity certainty refers to the degree to which people hold a fixed and known set
of ethnic meanings for themselves. Uncertainty about one’s ethnic identity makes it
more difficult to achieve identity verification and negatively impacts the verification
process because it is difficult for the actor to discern what feedback from others is
relevant for their identity. For example, Backman et al. (1963) asked subjects to
report aspects of themselves that five others would attribute to them as well as
aspects few other people would attribute to them. They hypothesized and found that
people were more likely to change their initial self-assessments after receiving
contradictory information for aspects of the self that were not shared by others
than for aspects of the self that were shared by several others. This is relevant here
because people hone their social or group-based identities by interacting within
social contexts with others that share the same group-based identity. However,
those uncertain of their ethnic identity will have more difficulty locating the appro-
priate contexts and social networks within which to practice and perfect the bound-
aries of their ethnic identity as well as knowing what feedback to attend to. This
leads us to present the following hypothesis:
H4: The more uncertain an individual is of their ethnic identity, the more
difficulty the individual will experience verifying his/her ethnic identity.
Procedures
Sample
The data used for this study were obtained from a survey administered to student
volunteers in several upper and lower division undergraduate sociology courses
offered at a large, ethnically diverse Southwestern university.2 Students were offered
extra course credit in exchange for their participation. We utilized the categories of
the U.S. Census for identifying race/ethnicity. Respondents were asked to classify
themselves in one of the following seven categories: American Indian or Alaskan
Native, Asian, Black or African American, Native Hawaiian or Pacific Islander,
White (non-Hispanic), Multiracial, or Hispanic or Latino. Because there were too
few respondents for meaningful analysis, those who classified themselves as Amer-
ican Indian or Native Hawaiian were omitted from the analysis. This left 775 respon-
dents with 283 Asians (37%), 250 Latinos (32%), 120 Whites (15%), 65 African
Americans (8%), and 57 classifying themselves as Multiracial (7%).
While our sample is a convenience sample, we are not trying to generalize our
findings to some population. Rather we focus on the development of measures of
ethnic identity and the testing of theoretical processes concerning those identities
and their verification. Our goal is to determine how well identity theory and the race/
ethnicity literature can be linked and utilized to strengthen one another.
Measures
Ethnic identity. Items were developed from two sources: an initial and separate
sample of 90 students’ written descriptions of what it meant for them to be part of
their racial/ethnic group, and the published works of Phinney (1992) and Alba
(1990) on ethnic identity. In keeping with the symbolic interactionist underpinnings
of identity theory (Burke & Tully, 1977), we were interested in developing a
grounded approach to measuring the ethnic identity meanings held by our student
sample. Ninety students enrolled in an introductory social psychology course wrote
an essay reflecting what their race/ethnicity meant to them. Students wrote about
how they viewed themselves in terms of their racial/ethnic background; what sorts of
activities they were involved with based on their race/ethnicity; what their race/
ethnicity meant to them; and when they were most aware of their race/ethnicity.
Papers were then analyzed for common themes.
Several recurring themes surfaced in students’ construal of their racial/ethnic
identity. Most of the themes discussed mirrored the extant research. They included
2
The racial/ethnic background of the undergraduate population was approximately: eight percent
Black, 40% Asian/Pacific Islander, and 31% Hispanic and 16% white and about 5% other groups,
including Native American.
eating/preparing ethnic foods; looking like and being correctly identified as a

member of one’s self-identified ethnicity; listening to certain types of music; being
around ethnically similar others; participating in traditions, holidays and festivals;
speaking a relevant language/dialect; wearing certain clothing; and participating in
religious traditions—all of which have been identified elsewhere as ways that people
enact their ethnic identities (Alba, 1990; Gans, 1979; Jimenez, 2004, 2008; Umana-
Taylor et al., 2009; Wimmer, 2008).
Eating and preparing food was one of the most frequent themes in the students’
essays. For example, one indicative quote illustrates the importance of preparing and
eating specific foods for ethnic identity: “Another way that I take pride in identifying
my ethnic identity as Mexican-American is [by] knowing how to prepare signature
meals that are recognized in our culture.” Ethnic traditions were also a frequently
cited aspect of one’s ethnic identity. One student stated, “Personally, I think it is
through the observation of our traditions I can feel that I belong to [my] Asian ethnic
group. . .” and, “There are just so many things to say with each other since we know
and share a lot of the same traditions.” Participating in ethnic traditions (e.g. festivals
and eating certain foods), for these respondents, signified an important sense of
belonging and kindred understanding among members of the same ethnic group.
Finally, looking like and being recognized as a member of one’s racial/ethnic
group was frequently discussed. For example, “My general appearance suggests
German/English Caucasian blood, which I do identify with, and represents the
majority of my ancestry.” Overall, these recurring themes highlight how the actions
and behaviors of the respondents based on their ethnicity help to define one as a
member of an ethnic group, the boundaries of that group, and increase solidarity with
the group. The more frequently encountered descriptions of ethnic practices gleaned
from the students’ essays were utilized in this study’s measurement of ethnic
identity.
In addition to the themes of the essays, we consulted the work on the measure-
ment of ethnic identity conducted by Alba (1990) and Phinney (1992) to construct
the ethnic identity measures. Many of the items in Alba and Phinney’s measures of
ethnic identity overlapped with the themes of the initial students’ essays. Drawing on
these sources, we utilized identity theory to guide the construction of the scale.
Explicitly, we focused on identity theory’s emphasis that meaning is manifest in
behavior. leading us to use those items pertaining to the actions and behavior people
use to exemplify ethnic group membership.
For example, a desire to be around ethnically similar others that frequently
surfaced in our participants’ written descriptions and the item “feeling a special
sense of relationship to someone else because of your ethnic background” in Alba’s
(1990) “ethnic experiences” scale were used to inform the construction of the
following three items: (1) “having friends with the same ethnic background,”
(2) “dating within my ethnic community” and (3) “being in my ethnic community.”
These latter three items were included in this study’s measure. From these sources, a
scale of ethnic/racial identity meanings resulted in a list of 14 items shown in
Table 2.1.
Table 2.1 Factor loadings using principle components of 14 ethnic identity meaning items with
promax rotation (N ¼ 775)
Items Heritage Personal
Eating foods associated with my ethnicity 0.72 0.01
Engaging in ethnic traditions 0.90 0.04
Speaking the language associated with my ethnicity 0.71 0.05
Sharing my ethnic heritage with my family 0.89 0.11
Holding on to my ethnic beliefs and attitudes 0.78 0.08
Observing the religious traditions associated with my ethnicity 0.58 0.24
Participating in ethnic holidays/festivals 0.76 0.09
Having friends with the same ethnic background 0.28 0.60
Looking like my ethnicity 0.04 0.75
Listening to music associated with my ethnicity 0.24 0.62
Being in my ethnic community 0.29 0.62
Dating within my ethnic community 0.09 0.81
Wearing clothing styles associated with my ethnicity 0.13 0.89
Maintaining ethnic gender roles 0.05 0.79
Eigenvalues 6.19 5.83
Omega reliability .93 .92
For each item, respondents were asked the degree of importance they placed on
each feature as an expression of their identity. The response categories included “not
important, a little important, somewhat important, very important, and extremely
important” and were coded 1–5, where a higher score indicates a greater degree of
importance placed on a given ethnic identity meaning.
Ethnic Identity was measured using these 14 items. To determine the dimension-
ality of these items measuring ethnic identity, an exploratory factor analysis was
conducted using a principle component factor solution with promax rotation to
preserve the correlation between the factors. This resulted in two factors, each
with seven items. These are also indicated in Table 2.1. An examination of the
loadings on the two factors suggests that the meanings highlighted in the first factor
center on the group defining traditions, sharing, collective culture, and heritage.
Engaging in these practices reaffirms the collective group and its traditions. The first
item, “eating foods associated with my ethnicity,” which might seem more like the
role items of factor two, is most highly correlated with the second item, “engaging in
ethnic traditions” suggesting that the meaning of the foods item is more aligned with
traditions than just personal behaviors. We label this first factor the heritage dimen-
sion of ethnic identity as it is based on the group defining meanings of the ethnic
category—the being ethnic dimension.
The second factor seems more personal, individual, and oriented toward ethnic
role performance: acting to keep ties through friends, neighbors, dating, and looks.
The item “being in my ethnic community” on the surface might seem more like the
communal items of factor one, but it is most highly correlated with “having friends
with the same ethnic background” suggesting more of an individual acting to keep
ties rather than maintaining the cultural traditions associated with the first factor. The
last item, “maintaining ethnic gender roles”, might be seen as aligning with tradi-
tions, but it is most highly correlated with “wearing clothing styles associated with
my ethnicity”, suggesting it is more of maintaining appearances rather than uphold-
ing traditions. We label this second factor the personal dimension of ethnic identity
based on the individual’s way of playing out the role of ethnic group member—the
doing ethnicity dimension.
As the results show, these two dimensions are not completely independent.
Across this sample of different ethnicities, the correlation between the two dimen-
sions is .65, indicating that they share about 42% of the variance in common—the
being and doing of ethnicity go together. From the results of the factor analysis, we
then created two separate scales by standardizing responses to each item, making
sure they are oriented in the proper direction, and then summing the items for each
factor. The heritage dimension of ethnic identity yielded a scale with an omega
reliability (Heise & Bohrnstedt, 1970) of .92 and the personal dimension of ethnic
identity yielded a scale with an omega reliability of .91.
The measure of difficulty in verifying one’s ethnic identity was also created using
the 14 items indicating ethnic identity meanings/practices. Respondents were asked
to identify the degree of difficulty they had in carrying out the specific practices
indicated in each item. The response categories included “no difficulty at all, not
much difficulty, some difficulty, quite a bit of difficulty, and a great deal of
difficulty” and were coded 1–5 with a higher number indicating greatest difficulty
enacting ethnic/racial meanings—and thus greater identity non-verification. These
items were standardized and summed to form a scale with an omega reliability of
.95.3
The measure of ethnic/racial identity knowledge was comprised of four items in
which respondents were asked to indicate the degree of knowledge they have of their
race/ethnicity. Example items included “I know a great deal about my ethnic
background” and “In general, I have a clear sense of my ethnicity”, and the response
categories were “strongly agree, agree, disagree, and strongly disagree.” These items
were coded 1–4, standardized and summed to form a scale with an omega reliability
of .88.
The measure of ethnic/racial identity uncertainty is comprised of four items
indicating the degree of uncertainty an individual has about her/his race/ethnicity.
Respondents indicated the degree to which they agreed or disagreed with statements
such as “one day I might have one opinion about my ethnicity, and on another, I
might have a different opinion,” “I spend a lot of time wondering about my
ethnicity,” and “my beliefs about my ethnicity seem to change frequently.” These
3
Note that this measure of verification difficulty does not compare a measure of the respondent’s
identity standard with a measure of their reflected appraisals, but, instead, taps directly the
respondent’s feeling of discrepancy—that is, of not being able to have perceptions from the
environment match one’s identity standard.
Table 2.2 Means and stan- Variable Mean Std. dev.

dard deviations for all vari-
Ethnicity
ables (N ¼ 775)
Asian .37 .48
Black .08 .28
Latino .32 .47
Multiracial .07 .25
White .15 .36
Knowledge .00 .83
Uncertainty .00 .78
Generation 1 .18 .38
Ethnic identity
Heritage .00 .79
Personal .00 .78
Verification difficulty .00 .68
Self-worth .00 .77
Self-efficacy .00 .74
Authenticity .00 .72
items were coded 1–4, standardized, and summed to form a scale with an omega
reliability of .83.
The measure of generational status follows the generational classification utilized
by Portes and Rumbaut (2001). We classify a respondent as first-generation if the
respondent and his or her parents were born outside of the U.S.4 Respondents
reporting that they were born in the U.S. but one or both of their parents were
born outside of the U.S. were coded as second-generation. Lastly, we coded those
respondents reporting that they and both parents were born within the U.S. as third-
generation.
Finally, we measured the three components of self-esteem identified by Stets and
Burke (2014b): self-worth, self-efficacy, and authenticity. Self-worth was measured
with 7 items and an omega reliability of .92, the efficacy scale had 7 items and an
omega reliability of .90, and the authenticity scale had 7 items with an omega
reliability of .89. The items and loadings for the items on these scales are presented
in Appendix. High scores on the scales represent high levels of the esteem compo-
nent. Table 2.2 presents the means and standard deviations for all the variables
included in the analyses, while Table 2.3 presents the zero-order correlations among
the variables used in the analyses.
4
We do not have the data to determine when these respondents immigrated to the U.S. Thus, we are
unable to determine if any of these individuals should be labeled as the 1.5 generation.
24
Table 2.3 Zero-order correlations among variables (n ¼ 775)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1 Asian 1.00
2 Black .23* 1.00
3 Latino .52* .21* 1.00
4 Multiracial .21* .09* .19* 1.00
5 White .32* .13* .29* .12* 1.00
6 Knowledge .06 .11* .00 .05 .20* 1.00
7 Uncertainty .22* .02 .11* .00 .17* .05 1.00
8 Generation 1 .29* .11* .10* .04 .14* .10* .05 1.00
9 Generation 2 .10* .18* .28* .00 .35* .11* .03 .53* 1.00
10 Generation 3 .36* .30* .22* .04 .52* .22* .08* .28* .67* 1.00
11 Heritage .07 .00 .23* .05 .36* .45* .02 .16* .18* .35* 1.00
12 Personal .18* .07* .00 .08* .23* .34* .09* .21* .00 .19* .67* 1.00
13 Ver difficulty .13* .01 .10* .13* .14* .07 .21* .03 .08* .12* .04 .05 1.00
14 Self-worth .24* .10* .13* .01 .08* .16* .15* .06 .06 .13* .08* .00 .18* 1.00
15 Efficacy .22* .03 .11* .03 .11* .02 .33* .13* .01 .13* .01 .11* .22* .56* 1.00
16 Authenticity .18* .03 .15* .02 .08* .01 .27* .07* .01 .07* .02 .09* .20* .53* .56*
*p .05
P. J. Burke and M. M. Harrod
Results
Before looking at the results of the measures of ethnic identity, we address the first
question of the validity of these measures. Within identity theory, difficulty verifying
a social or group-based identity is posited to lead to lower feelings of self-esteem
(Burke & Harrod, 2002; Burke & Stets, 2009; Stets & Burke, 2000, 2014b). We thus
want to check that the ethnic identity meanings here are perceptually controlled
meanings as predicted by identity theory (Powers, 1973). To do this, we look at the
consequences for all three components of self-esteem when individuals have diffi-
culty in verifying their ethnic identity. We use structural equation modeling to
estimate the effects of difficulty in verifying ethnic identity on the three components
of self-esteem, allowing the error residuals to correlate. We also included ethnicity
and difficulty by ethnicity interactions and other controls presented in Table 2.4.
There, we see that difficulty in verifying the meanings associated with one’s ethnic
identity does significantly reduce an individual’s reported level of all three compo-
nents of self-esteem. Further, testing for variation of this effect by ethnic group and
generation shows no significant interaction by either ethnic group (χ2(12) ¼ 9.32,
p ¼ .68) or generation (χ2(6) ¼ 8.51, p ¼ .20). These results demonstrate that the
meanings being measured with the ethnic identity scale presented here are valid
indicators of ethnic identity for all ethnic groups and that failure to verify the identity
reduces self-esteem equally for all groups. We also see in these results that ethnic
identity uncertainty and knowledge influence the esteem components. Table 2.4
shows that being more certain about one’s ethnic identity increases all three esteem
components, though lacking ethnic identity knowledge only hurts the self-worth
component of self-esteem having to do with feelings of belonging and acceptance.
Each of the ethnic identity scales (heritage and personal) were examined across
ethnic groups and generations using an analysis of covariance with ethnic group and
generation as factors and knowledge and uncertainty as covariates to gain a better
understanding of how ethnic identity varies across racial/ethnic groups and
Table 2.4 Standardized coefficients for the regression of self-esteem on ethnic identity
nonverification, knowledge, and uncertainty (N ¼ 775)
Self-esteem components
Variables Worth Efficacy Authenticity
Difficulty verifying ethnic identity .12* .15* .13*
Ethnic group (reference group white)
Asian .22* .13* .10*
Black .03 .01 .03
Multiracial .03 .02 .02
Latino .02 .00 .05
Ethnic identity knowledge .16* .01 .02
Ethnic identity uncertainty .09* .26* .22*
R2 .12 .15 .11
*p <¼ .05
Table 2.5 Mean heritage Generation

(group) ethnic identity scores
Ethnic group 1 2 3 Total
by ethnic group and genera-
tion (N ¼ 775) Asian .40 .03 .48 .09
Black .05 .46 .18 .00
Multiracial .03 .08 .43 .18
Latino .32 .40 .02 .34
White .05 .33 1.00 .84
Total .34 .16 .58 .00
Table 2.6 Tests of effects β p

from an analysis of covari-
Knowledge .38 .01
ance of Table 2.5
Uncertainty .06 .03
Ethnic group differences .01
Generation differences .01
Ethnic group Generation .60
Table 2.7 Mean personal Generation

(role) ethnic identity scores by
ethnic group and generation
(N ¼ 775) Asian .69 .03 .17 .24
Black .56 .23 .21 .23
Latino .07 .05 .25 .00
White .23 .44 .62 .55
Total .45 .00 .32 .00
generations. Table 2.5 presents the means of the heritage, group, or being dimension
of respondents’ ethnic identity by ethnicity and generation, and Table 2.7 presents
results for the personal, role, or doing dimension by ethnicity and generation as well.
Looking first at Table 2.5, we see that the degree to which individuals endorse the
importance of the heritage or group dimension of ethnic identity meanings does
indeed vary by ethnicity and generation. Not surprisingly, we see that the lowest
reported heritage scores are for Whites, and the highest scores are reported by
Latinos, with the other groups falling somewhere in-between.
Table 2.6 shows that there are no race/ethnicity-by-generation interaction effects
for the heritage dimension of ethnic identity. Looking at the tables for generational
differences, we see a decline in the importance of the ethnic heritage dimension from
the first to the third-generations, with the largest shift occurring between the second
and third-generation respondents. Finally, we see that the more respondents know
about their ethnicity, the more important is the heritage component of their ethnic
identity. Ethnic identity uncertainty, on the other hand, reduces the importance of the
heritage dimension.
Turning next to the personal dimension or role aspect of ethnic identity, Table 2.7
shows significant variation in the importance of this dimension by ethnicity but not

Knowledge .30 .01
ance of Table 2.7
Uncertainty .04 .29
Table 2.9 Mean ethnic iden- Generation

tity verification difficulty
scores by ethnic group and
generation (N ¼ 775) Asian .13 .15 .77 .18
Black .59 .34 .12 .03
Multiracial 1.17 .55 .01 .46
Latino .39 .14 .03 .15
White .12 .33 .50 .34
Total .07 .07 .20 .00
by generation on the average. There are, however, significant interaction effects of

ethnicity and generation, as seen in Table 2.8. On average, Whites and Multiracial
groups report the lowest importance of personal ethnic meanings, while Asian and
African Americans report the highest, although this varies by generation. Looking
next at generational status, we see declines from the first to the second to the third-
generation for Asians, Blacks, and Whites. There are no declines for Multiracial
individuals by generation, while for Latinos there is a slight increase from generation
1 to 2, and a decline from generation 2 to 3. Also, within Table 2.8, we see the strong
effects that ethnic identity knowledge has on the personal, role, or doing dimension
of ethnic identity measured here. Respondents with the greatest knowledge about
their ethnicity held personal meanings most strongly. Ethnic identity uncertainty, on
the other hand, has no significant effect on the strength to which respondents
endorsed the personal (role) meanings of their ethnic identity.
In sum, the degree to which respondents see either the heritage dimension or the
personal dimension of ethnic identity meanings as important varies by ethnicity and
generation, though not in a simple way.
We test hypotheses concerning difficulty of ethnic identity verification in
Table 2.9. Here we see that Whites have the least difficulty verifying their ethnic
identity while Multiracial individuals have the most difficulty, thus supporting
Hypothesis 1. Recall that Hypothesis 1 argues that members of minority groups
will experience greater difficulty verifying their ethnic identity than Whites, likely
because they have, on the average, fewer resources to manage the verification. We
also see that there is a significant effect of generation on the verification of respon-
dent’s ethnic identity with both first- and second-generation respondents having
more difficulty than third-generation on the average.
However, a closer examination of these results reveals a more complex picture as
there is a significant interaction between race/ethnicity and generation with respect to
the difficulty verifying one’s ethnic identity. We see increasing difficulty for Asians

Knowledge .13 .01
ance of Table 2.9
Uncertainty .18 .01
when moving from the first to the third-generation with the biggest jump in difficulty
occurring between generations two and three. African Americans, on the other hand,
have decreasing difficulty when moving across the generations as do Multiracial
individuals. Latinos, as a group, have fewer ethnic identity verification problems, but
these difficulties increase slightly from generation one to generation three. Multira-
cial individuals have the most difficulty verifying their ethnic identity in the first-
generation. And, while this difficulty drops off over generations, it still averages the
highest of all ethnic groups. Multiracial individuals have few shared traditions that
can be affirmed.
Overall, the results show that Whites and Latinos experience the least difficulty in
verifying their ethnic identities whereas Multiracial and Asian respondents have the
greatest reported difficulty. With respect to hypothesis two, only Whites conform to
the hypothesized pattern of most difficulty verifying ethnic identity in generation
two. Instead, each ethnic group seems to have its own level and pattern of identity
verification difficulty. These results demonstrate the varied nature of ethnic identity
verification when examined by one’s race/ethnicity and generational status.
Finally, we see support for both hypotheses three and four concerning the effect
of knowledge and uncertainty on ethnic identity verification. Increasing knowledge
about one’s race/ethnic identity facilitates the verification of that identity, while
uncertainty about one’s identity makes verification more difficult as hypothesized
(Table 2.10).
We now turn to examine variations in both ethnic identity knowledge and
uncertainty across ethnic groups and generations. Recall that knowledge refers to
an understanding of the history and traditions of the group, while uncertainty refers
to the stability and consistency of the ethnic identity meanings for the individual.
Interestingly, as shown in Table 2.3, there is little relationship between identity
uncertainty and knowledge.
Table 2.11 shows the results for the level of knowledge one has of his/her
ethnicity. Knowledge about one’s ethnic identity does vary by ethnicity, with Whites
displaying the lowest level of knowledge and African Americans displaying the
highest. By generation, the level of knowledge about one’s ethnic identity decreases
when moving from generation one to three. Looking at Table 2.12, we see that there
are no significant interaction effects of ethnic identity knowledge and generation.
Finally, Table 2.13 shows the results for ethnic identity uncertainty. Again, there
are differences by ethnicity with Asian respondents reporting the greatest level of
uncertainty and White and Latino respondents reporting the least. The high level of
uncertainty for Asians may well be because this ethnic category is not endemic to
Table 2.11 Mean ethnic Generation

identity knowledge scores by
(N ¼ 775) Asian .25 .02 .45 .07
Black .26 1.06 .10 .36
Latino .13 .03 .23 .01
White .52 .14 .67 .47
Total .23 .10 .38 .00
Table 2.12 Tests of effects p

ance of Table 2.11
Table 2.13 Mean ethnic Generation

identity uncertainty scores by
Asian .24 .30 .44 .29
Black .19 .09 .11 .06
Latino .26 .18 .15 .16
White .12 .51 .38 .38
Total .11 .02 .13 .00
Table 2.14 Tests of effects p

ance of Table 2.13
their origin—being from many different countries, each with its own character and
traditions. Only in the U.S. are they lumped together ethnically. Looking at
Table 2.14 we can see there are no effects for generation and no significant
interaction effect of ethnicity-by-generation.
Discussion and Conclusion
This study began as an effort to explicate how the race/ethnicity and social psycho-
logical literatures can inform each other. Our goal in this study was three-fold. First,
we wanted to present a coherent social psychological framework (identity theory) to
help inform existing findings within the race/ethnicity literature, while also
extending identity theory’s range through the examination of a social or group
identity, which is perhaps the least explored of the identity bases (social, role, and
person) within sociology (Burke & Stets, 2009). Previous studies in the race/
ethnicity literature have found that people desire to have others confirm their
racial/ethnic identity, and when this does not occur, people report feeling frustrated
and uneasy (Jimenez, 2004; Khanna, 2004, 2010; Waters, 1999). However, these
studies do not discuss why people should experience such negative feelings and
outcomes. Identity theory, on the other hand, has a well-developed theoretical and
empirical program addressing precisely why and how people seek identity verifica-
tion and report unease when that identity verification fails to happen.
Following a mixed-methods approach, we drew on identity theory to create a
measure of racial/ethnic identity, based on ethnic meanings as expressed through
behavior. From both a grounded approach and review of the relevant race/ethnicity
literature (e.g. Alba, 1990; Phinney, 1992) we were able to develop items to measure
what it means to the respondents to hold the ethnic identity they have. In doing this,
we find, for our ethnically diverse sample, that there are two related dimensions to
ethnic identity, one based on common heritage and the other on the individual role
performances that attach one to the ethnic community. These scales build on the
original MEIM scale but extend its scope to understand how ethnic identity is
regulated and maintained after it is formed.
Both the group membership ethnic meanings (heritage) and the role performance
ethnic meanings (personal) were held in varying degrees by respondents. The full
meaning of ethnicity thus appears spread across both the being and doing of
ethnicity. Our use of identity theory as a guiding theoretical framework in this
study led us to emphasize the meanings of one’s ethnic identity, as understood by
the respondents themselves and expressed through behaviors—factors not always
emphasized in the race/ethnicity literature. We suggest that the personal or role
dimension of ethnic identity is comprised of meanings expressed through behaviors
that serve to sustain the self in interaction as an ethnic (doing ethnicity), while the
heritage or group dimension is composed of meanings expressed through behaviors
that serve to sustain and maintain the group or collectivity (being ethnic).
With respect to the personal or role dimension, the actions of presenting oneself
as a member of a specific racial/ethnic group (e.g. wearing ethnically appropriate
clothing styles) and interacting with ethnically similar others (being/dating within
my ethnic community) embody the individual as a distinct member of a group. These
actions enable the individual to be identified with their ethnic group and not another,
and they enable the individual to maintain their group membership through interac-
tion with those who are ethnically similar. Because self-presentation (Goffman,
1959) and interaction (Mead, 1934) are fundamental ways in which individuals
develop and sustain their self-understandings, it is important to recognize that it is
through these actions that an individual meaningfully defines who he or she is as a
member of the group and, thus, sustains the self as a member of one group rather than
another.
The heritage dimension of ethnic identity, on the other hand, relates to the broader
social group and is generally composed of behaviors/meanings that serve to sustain
the group and culture. Engaging in ethnic traditions: participating in ethnic holidays/
festivals and holding on to one’s ethnic beliefs and attitudes are actions that serve to
maintain a group’s culture, norms, and meanings. A social group depends on group
members’ enactment of these activities without which it would become destabilized
and likely dissipate. Conversely, if the group no longer existed, neither would one’s
ethnic identity.
For our third goal, we explored how well the predictions of identity theory hold
for a currently unexamined group-based identity by testing the effects of difficulty in
identity verification on feelings of self-esteem. Using the three components of self-
esteem proposed by Stets and Burke (2014b), we confirmed identity theory’s
expectation that respondents who reported difficulty confirming their ethnic identity
also reported lower levels of self-worth, efficacy, and authenticity even when
controlling for knowledge and uncertainty about their ethnic identity. We also
found that the effect of non-verification was the same for all ethnic groups studied,
including Whites, as well as for all three generations, which results are consistent
with prior research on the effects of identity non-verification, giving us confidence in
our measure of ethnic identity. Additionally, we find that having high levels of
uncertainty about one’s ethnic identity leads to a decrease in all components of self-
esteem. Lack of knowledge about one’s ethnic identity, however, reduces only self-
worth but not efficacy or authenticity.
When we examined variations in the degree to which individuals from different
ethnic groups and generations endorsed the common meanings of ethnic identity
indicating the different views about ethnicity that are held across groups and
generations that were consistent with prior research (Kasinitz et al., 2004; Portes
& Rumbaut, 2001, 2006). For example, results show that third-generation respon-
dents report a significantly lower sense of ethnic identity (both personal and heritage
dimensions) than either first- or second-generation respondents.
The Multiracial group had consistently low personal ethnic identity meanings
across all three generations, and were among the lowest heritage ethnic identity
meanings, likely because there is not a single group around which these meanings
may develop. Whites were similarly low, especially after the first-generation though
the reasons are likely different. As part of the dominant culture in which the blending
of different white ethnicities has occurred, Whites may find less that brings them
together as a group. Interestingly, even among groups that had lower levels of group
and role meanings of ethnic identity, members still sought to control those levels of
meaning and they experienced lower self-esteem when they had difficulty doing so.
With respect to the verification of ethnic identities, the results present mixed
support of hypothesis two that second-generation respondents would experience
greater difficulties. This was only true of Whites. Other ethnic groups had their
own pattern. Latinos and Asians had increasing difficulty over the three generations,
while Multiracial and black respondents found verification easier over the genera-
tions. And, while Whites had fewer verification difficulties than other groups in
accordance with hypothesis one, this was primarily true in the third-generation.
Thus, we see that generational experiences matter, but depend upon the racial/ethnic
group with which one identifies.
Our findings suggest that the level of knowledge of one’s ethnic identity varies
across the various racial/ethnic groups and generational status. Whites, as the
dominant group within U.S. society, know the least about their ethnic background
and African Americans, as the historically most oppressed group, reports the greatest
ethnic identity knowledge. This finding is not too surprising considering that
researchers have found that Whites typically report difficulties identifying unique
aspects of “white culture” because it is taken as normative (Bonilla-Silva, 2003;
Gallagher, 2003; Lopez, 1997; Perry, 2007). Knowledge is also influenced by
generational status, with first-generation respondents reporting the greatest knowl-
edge of their ethnic identity and third-generation respondents reporting the least.
Uncertainty about one’s ethnic identity varies by racial/ethnic group, with Asian
respondents reporting the most uncertainty and Whites and Latinos reporting
the least. The level of uncertainty for Asian respondents may be partly attributed
to the fact that many different Asian cultures and ethnicities are aggregated into the
umbrella term “Asian” in the U.S. The meaning of “Asian” is thus inconsistent and
confusing with no common history or community (Okamoto, 2014).
While these findings come from an exploratory study, they provide some inter-
esting prospects for future research and efforts to better understanding of people’s
actions and reactions to events. We know that the drive for identity verification leads
people to choose social situations where others see and react to them as they see
themselves and avoid other environments that lead to identity non-verification. How
this plays out to create different groupings and patterns of interaction and even
residence would be a topic for future research to explore.
It is the personal or role dimension of ethnic identity whereby persons try to
control meanings related to how they portray themselves and be seen by others. As
suggested by this study, this occurs through trying to engage and maintain social
relationships linked to their ethnicity. Our subjects reported that they choose to
spend time within their ethnic community and to dress in ways that confirms the
individual self-views. Appreciation of the personal dimension of one’s ethnic iden-
tity should be increasingly relevant as immigration patterns and ethnic replenishment
continue to change, leading to the sorts of identity issues highlighted by Jimenez
(2008) or the importance within the race/ethnicity literature about understanding a
Multiracial identity (Brunsma, 2005; Campbell & Troyer, 2007; Khanna, 2004,
2010). A greater appreciation of how individuals construct their ethnic identity
will help researchers understand people’s actions, even when they may not be
recognized, ethnically, by others in the same way.
Understanding the heritage dimension also has implications for the race/ethnicity
literature because it provides a mechanism for someone to belong in their ethnic
community and be part of the community and the traditions that bind them. While it
is important that someone has an internalized sense of belonging to a group, it is
equally important that they find situations and others with whom to enact their
identity (Burke & Stets, 1999; Deaux & Martin, 2003; Stryker, 1980). Enacting
one’s ethnic identity with others strengthens and reinforces existing understandings
of the identity and allows for a sense of commonality.
Ultimately, this exploratory study suggests that identity theory has a great deal to
offer researchers working within the race/ethnicity literature. It also shows that an
appreciation of the race/ethnicity literature has a lot to offer social psychological
research generally, and identity theory researchers, specifically. We can see that
there are important differences in how members from different racial/ethnic groups
conceptualize and express their ethnic identity. But despite these important differ-
ences, we found that our measure is applicable across a wide range of different racial/
ethnic groups. This research helps us understand why people choose to associate
with certain people and certain locations and avoid others depending on whether
others verify their ethnic identity.
Appendix
Factor Analysis of Self-Worth Scale Items
Items Loadings
1. I feel I am a person of worth, at least on an equal basis with others.a .72
2. I feel that I have a number of good qualities.a .79
3. I take a positive attitude toward myself.a .81
4. On the whole, I am satisfied with myself.a .75
5. I usually feel good about myself. .82
6. I feel I have much to offer as a person. .79
7. I have a lot of confidence in the actions I undertake in my life. .71
Omega reliability: .92
a
Item from Rosenberg Self-Esteem Scale (Rosenberg, 1965)
Factor Analysis of Self-Efficacy Scale Items
Items Loadings
1. I feel as if what happens to me is mostly determined by other people. .69
2. I certainly feel helpless at times.a .74
3. There is no way I can solve some of the problems I have.b .76
4. Sometimes I feel that I’m not able to accomplish what I want. .74
5. I have little control over the things that happen to me.b .79
6. I often feel unable to deal with the problems of life.b .78
7. There is little I can do to change many of the important things in my life.b .70
a
Item from Rosenberg Self-Esteem Scale (Rosenberg, 1965)
b
Item from Pearlin Mastery Scale (Pearlin & Schooler, 1978)
Factor Analysis of Authenticity Scale Items
Items Loadings
1. I feel most people don’t know the “real” me. .68
2. I find I can almost always be myself. .76
3. I feel people expect me to be different than I really am. .61
4. I think most people accept who I really am. .72
5. I just wish I were more able to be myself. .78
6. I feel the way in which I generally act reflects the “real” me. .71
7. I often do not feel I am myself. .76
References
Alba, R. D. (1990). Ethnic identity: The transformation of white America. Yale University Press.
Backman, C. W., Secord, P. F., & Peirce, J. R. (1963). Resistance to change in the self-concept as a
function of consensus among significant others. Sociometry, 26, 102–111.
Bean, F. D., & Stevens, G. (2005). America’s newcomers and the dynamics of diversity. Russell
Sage Foundation.
Blake, J. J., Keith, V. M., Luo, W., Le, H., & Salter, P. (2017). The role of colorism in explaining
African American females’ suspension risk. School Psychology Quarterly, 32, 118–130.
Bonilla-Silva, E. (1997). Rethinking racism: Towards a structural interpretation. American Socio-
logical Review, 62, 465–480.
Bonilla-Silva, E. (2003). Racism without racists: Color-blind racsim and the persistence of racial
inequality in the United States. Rowman and Littlefield.
Brown, J. S., Hitlin, S., & Elder, G. H., Jr. (2006). The greater complexity of lived race: An
extension of Harris and Sim. Social Science Quarterly, 87, 411–431.
Brown, T. N., & Lesane-Brown, C. L. (2006). Race socialization messages across historical time.
Social Psychology Quarterly, 69, 201–213.
Brunsma, D. L. (2005). Interracial families and the racial identification of mixed-race children:
Evidence from the early childhood longitudinal study. Social Forces, 84, 1131–1157.
Burke, P. J., & Harrod, M. M. (2002). Too good to be believed? Paper presented at the American
Sociological Association, Chicago, Il.
Burke, P. J. (1991). Identity processes and social stress. American Sociological Review, 56,
836–849.
Burke, P. J. (2008). Identity, social status, and emotion. In D. T. Robinson & J. Clay-Warner (Eds.),
Social structure and emotion (pp. 75–93). Elsevier.
Burke, P. J., & Cast, A. D. (1997). Stability and change in the gender identities of newly married
couples. Social Psychology Quarterly, 60, 277–290.
Burke, P. J., & Harrod, M. M. (2005). Too good to be believed? Social Psychology Quarterly, 68,
359–374.
Burke, P. J., & Stets, J. E. (1999). Trust and commitment through self-verification. Social Psychol-
ogy Quarterly, 62, 347–366.
Burke, P. J., & Tully, J. C. (1977). The measurement of role identity. Social Forces, 55, 881–897.
Campbell, M. E., & Troyer, L. (2007). The implications of racial misclassification by observers.
American Sociological Review, 72, 750–765.
Cast, A. D., & Burke, P. J. (2002). A theory of self-esteem. Social Forces, 80, 1041–1068.
Cast, A. D., Stets, J. E., & Burke, P. J. (1999). Does the self conform to the views of others? Social
Psychology Quarterly, 62, 68–82.
Cooley, C. H. (1964 [1902]). The social self - The meaning of the “I”. In Human nature and the
social order (pp. 168–210). New York: Schocken Books.
Deaux, K., & Martin, D. (2003). Interpersonal networks and social categories: Specifying levels of
context in identity processes. Social Psychology Quarterly, 66, 101–117.
Erikson, E. (1968). Identity: Youth and crisis. Norton.
Frank, R., Redstone, I., & Bo, L. (2010). Latino immigrants and the U.S. racial order: How and
where do they fit in? American Sociological Review, 75, 378–401.
Gallagher, C. A. (2003). Playing the white ethnic card: Using ethnic identity to deny contemporary
racism. In A. W. Doane & E. Bonilla-Silva (Eds.), White out: The continuing significant of
racism (pp. 145–158). Routledge.
Gans, H. J. (1979). Symbolic ethnicity: The future of ethnic groups and cultures in America. Ethnic
and Racial Studies, 2, 1–20.
Gans, H. J. (1992). Second-generation decline: Scenarios for the economic and ethnic futures of the
post-1965 American immigrants. Ethnic and Racial Studies, 15, 173–192.
Goffman, E. (1959). Presentation of self in everyday life. Doubleday.
Harker, K. (2001). Immigrant generation, assimilation, and adolescent psychological well-being.
Social Forces, 79, 969–1004.
Harris, D. R., & Sim, J. J. (2002). Who is multiracial? Assessing the complexity of lived race.
American Sociological Review, 67, 614–627.
Heise, D. R., & Bohrnstedt, G. W. (1970). Validity, invalidity, and reliability. In F. B. Edgar &
G. W. Bohrnstedt (Eds.), Sociological methodology (pp. 104–129). Jossey-Bass.
Hill Collins, P. (2000). Black feminist thought: Knowledge, consciousness, and the politics of
empowerment. Routledge.
Hitlin, S., Scott Brown, J., & Elder, G. H., Jr. (2007). Measuring Latinos: Racial and ethnic
classification and self-understanding. Social Forces, 86, 587–611.
Hogg, M. A. (2018). Social identity theory. In P. J. Burke (Ed.), Contemporary social psychological
theories (pp. 112–138). Stanford University Press.
Howard, J. A. (2000). Social psychology of identities. Annual Review of Sociology, 26, 367–393.
Hunt, M. O. (2003). Identities and inequalities - Exploring links between self and stratification
processes. Advances in Identity Theory and Research, 71–84.
Hunt, M. O. (2007). African American, hispanic, and white beliefs about black/white inequality,
1977-2004. American Sociological Review, 72, 390–415.
Hunt, M. O., Jackson, P. B., Powell, B., & Steelman, L. C. (2000). Color-blind: The treatment of
race and ethnicity in social psychology. Social Psychology Quarterly, 63, 352–364.
Jimenez, T. R. (2004). Negotiating ethnic boundaries: Multiethnic Mexican Americans and ethnic
identity in the United States. Ethnicities, 4, 75–97.
Jimenez, T. R. (2008). Mexican American replenishment and the continuing significance of
ethnicity and race. American Journal of Sociology, 113, 1527–1567.
Kasinitz, P., Mollenkopf, J. H., & Waters, M. C. (2004). Becoming New Yorkers: Ethnographies of
the new second-generation. Russell Sage Foundation.
Keith, V. M., Nguyen, A. W., Taylor, R. J., Chatters, L. M., & Mouzon, D. M. (2017).
Microaggressions, discrimination, and phenotype among African Americans: A latent class
analysis of the impact of skin tone and Bmi. Sociological Inquiry, 87, 233–255.
Khanna, N. (2004). The role of reflected appraisals in racial identity: The case of multiracial Asians.
Khanna, N. (2010). “If You’re Half Black, You’re Just Black”: Reflected appraisals and the
persistence of the one-drop rule. The Sociological Quarterly, 51, 96–121.
Krysan, M., & Bader, M. D. M. (2009). Racial blind spots: Black-white-Latino differences in
community knowledge. Social Problems, 56, 677–701.
Lacy, K. R. (2007). Blue-chip black: Race, class, and status in the United States in the new black
middle class. University of California Press.
Lee, J., & Bean, F. D. (2004). America’s changing color lines: Immigration, race/ethnicity, and
multiracial identification. Annual Review of Sociology, 30, 221–242.
Lee, J., & Bean, F. D. (2007). Reinventing the color line: Immigration and America’s new racial/
ethnic divide. Social Forces, 86, 561–586.
Lopez, I. H. (1997). White by law: The legal construction of race. New York University Press.
Mead, G. H. (1934). Mind, self, and society from the standpoint of a social behaviorist. University
of Chicago Press.
Mouw, T., & Xie, Y. (1999). Bilingualism and the academic achievement of first- and second-
generation Asian Americans: Accommodation with or without assimilation? American Socio-
logical Review, 64, 232–252.
Nagel, J. (1994). Constructing ethnicity: Creating and recreating ethnic identity and culture. Social
Problems, 41, 152–176.
Okamoto, D. G. (2014). Redefining race: Asian American panethnicity and shifting ethnic bound-
aries. Russell Sage Foundation.
Omi, M., & Winant, H. (1994). Racial formation in the United States: From the 1960s to the 1980s.
Routledge and Kegan Paul.
Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement of meaning. University
of Illinois Press.
Pager, D., & Shepherd, H. (2008). The sociology of discrimination: Racial discrimination in
employment, housing, credit, and consumer markets. Annual Review of Sociology, 34, 181–209.
Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social
Behavior, 19, 2–21.
Perry, P. (2007). White universal identity as a “sense of group position”. Symbolic Interactionism,
30, 375–393.
Phinney, J. S. (1990). Ethnic identity in adolescents and adults: Review of research. Psychological
Bulletin, 108, 499–514.
Phinney, J. S. (1992). The multigroup ethnic identity measure: A new scale for use with diverse
groups. Journal of Adolescent Research, 7, 156–176.
Phinney, J. S. (2006). Ethnic identity exploration in emerging adulthood. In J. Arnett & J. L. Tanner
(Eds.), Coming of age in the 21st century: The lives and contexts of emerging adults
(pp. 117–134). American Psychological Association.
Phinney, J. S., & Ong, A. D. (2007). Conceptualization and measurement of ethnic identity: Current
status and future directions. Journal of Counseling Psychology, 54, 271–281.
Portes, A., & Rumbaut, R. G. (Eds.). (2001). Ethnicities: Children of immigrants in America.
University of California Press.
Portes, A., & Rumbaut, R. G. (2006). Immigrant America: A portrait. University of California
Press.
Powers, W. T. (1973). Behavior: The control of perception. Aldine.
Reid, S. A., & Hogg, M. A. (2005). Uncertainty reduction, self-enhancement, and ingroup identi-
fication. Personality and Social Psychology Bulletin, 31, 804–817.
Rogler, L. H., Cooney, R. S., & Ortiz, V. (1980). Intergenerational change in ethnic identity in the
Puerto Rican family. International Migration Review, 14, 193–214.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton University Press.
Rosenberg, M. (1979). Conceiving the self. Basic Books Inc..
Singh, V. P. (1977). Some theoretical and methodological problems in the study of ethnic identity:
A cross-cultural perspective. New York Academy of Sciences: Annals, 285, 32–42.
Stets, J. E., & Burke, P. J. (1994). Inconsistent self-views in the control identity model. Social
Science Research, 23, 236–262.
Stets, J. E., & Burke, P. J. (2000). Identity theory and social identity theory. Social Psychology
Quarterly, 63, 224–237.
Stets, J. E., & Burke, P. J. (2014a). Emotions and identity nonverification. Social Psychology
Quarterly, 77, 387–410.
Stets, J. E., & Burke, P. J. (2014b). Self-esteem and identities. Sociological Perspectives, 57, 1–25.
Stets, J. E., & Cast, A. D. (2007). Resources and identity verification from an identity theory
perspective. Sociological Perspectives, 50, 517–543.
Stets, J. E., & Fares, P. (2019). The effects of race/ethnicity and race/ethnic identification on general
trust. Social Science Research, 1–14.
Stets, J. E., & Harrod, M. M. (2004). Verification across multiple identities: The role of status.
Stryker, S. (1980). Symbolic interactionism: A social structural version. Benjamin Cummings.
Stryker, S. (1981). Symbolic interactionism: Themes and variations. In M. Rosenberg & R. H.
Turner (Eds.), Social psychology: Sociological perspectives (pp. 3–29). Basic Books.
Stryker, S., & Serpe, R. T. (1994). Identity salience and psychological centrality: Equivalent,
overlapping, or complementary concepts. Social Psychology Quarterly, 57, 16–35.
Ting-Toomey, S. (1981). Ethnic identity and close friendship in Chinese-American college stu-
dents. International Journal of Intercultural Relations, 5, 383–406.
Umana-Taylor, A. J., Alfaro, E. C., Bamaca, M. Y., & Guimond, A. B. (2009). The central role of
familial ethnic socialization in Latino adolescents’ cultural orientation. Journal of Marriage and
Family, 71, 46–60.
Waters, M. C. (1990). Ethnic options: Choosing identities in America. University of California
Press.
Waters, M. C. (1999). Black identities: West Indian immigrant dreams and American realities.
Harvard University Press.
Waters, M. C., & Jimenez, T. R. (2005). Assessing immigrant assimilation: New empirical and
theoretical challenges. Annual Review of Sociology, 31, 105.25.
White, C. L., & Burke, P. J. (1987). Ethnic role identity among black and white college students: An
interactionist approach. Sociological Perspectives, 30, 310–331.
Wimmer, A. (2008). The making and unmaking of ethnic boundaries: A multilevel process theory.
American Journal of Sociology, 113, 970–1022.
Chapter 3
Ethnic Identity Achievement, Identity
Verification, and Self-worth
Matthew Grindal, Melanie Kushida, and Tanya Nieri
People of color have historically experienced structural impediments as well as

discrimination and prejudice tied to their ethnic group membership (Kessler et al.,
1999; Pager & Shepherd, 2008). Prior research has found that this mistreatment is
associated with lower self-esteem (Zeiders et al., 2013), suggesting that discrimina-
tion and prejudice pose a threat to the self. Scholars have studied how a strong ethnic
identity can neutralize the harmful impact of discrimination and prejudice by
increasing the understanding and awareness of the positive elements of one’s ethnic
group, thus fostering a positive self-concept and good psychological health (see
Phinney, 1990; Phinney et al., 1997; Quintana, 2007; Smith & Silva, 2011).
Phinney’s (1989, 1990, 1993) and Phinney and Ong (2007) work on ethnic
identity achievement has been central to understanding the underlying process of
ethnic identity formation, the transformation of how people view their ethnic
identity, and thus the resilience offered by a strong ethnic identity. In her model of
ethnic identity formation, she theorizes that when people first experience an identity
crisis tied to their ethnicity, which may be triggered by a negative evaluation of their
ethnic group, they are motivated to explore the underlying social meanings of their
ethnic identity. Through this exploration process, people develop a more sophisti-
cated understanding of their ethnic group membership (commitment), including the
positive features of their ethnic heritage, which leads to a stable and secure under-
standing of their ethnic group (Phinney, 1989). Ethnic identity achievement is the
ultimate outcome of this process and reflects an ethnic identity that is marked by high
levels of commitment and exploration (Phinney, 1992).
M. Grindal (*)
University of Idaho, Moscow, ID, USA
M. Kushida · T. Nieri
University of California at Riverside, Riverside, CA, USA
e-mail: [email protected]; [email protected]

40 M. Grindal et al.
Research examining the impact of ethnic identity achievement confirms that it

contributes to positive psychological health including enhanced self-esteem
(Costigan et al., 2010; Ghavami et al., 2011; Phinney & Alipuria, 1990; Phinney
& Chavira, 1992; Yip & Fuligni, 2002). Prior work articulating the theoretical
mechanism underlying this relationship suggests that the mature and developed
sense of one’s ethnic identity marked by an achieved ethnic identity enables people
to enhance the positive feelings they have for their ethnic identity and thus enhance
one’s self-esteem (Ghavami et al., 2011). However, this explanation does not factor
in the interactional dynamics, which may account for why an achieved ethnic
identity enhances the positive feelings for one’s ethnic identity. In this study, we
address this limitation by testing a theoretical model that integrates insights from the
ethnic identity literature with identity theory (Burke & Stets, 2009).
We conceptualize ethnic identity achievement as the outcome of a developmental
process, which creates a personal network of others who share the same ethnic
identity and ethnic identity meanings. As a result, people are better able to verify
their ethnic identity and develop positive feelings for their ethnic identity (group-
specific self-worth), which, in turn, enhances their global self-worth. Drawing on a
multiethnic sample of college students, we test the relationship between ethnic
identity achievement and self-esteem, and the extent to which this relationship is
accounted for by identity verification and group-specific self-worth.
By integrating the concept of ethnic identity achievement with identity theory,
this paper helps theoretically clarify the importance of holding a strong ethnic
identity for positive self-worth and extends identity theory by clearly articulating
the function of group-specific self-worth in the verification of social identities. This
research also has applied implications. With a greater understanding of the theoret-
ical processes linking ethnic identity achievement to self-worth, administrators of
leadership and cultural awareness programs, designed to foster cultural and ethnic
exploration, can better structure and assess their programs to guarantee their success.
We address these themes in the discussion.
Theory
Ethnic Identity Achievement
Ethnic identity has been generally understood to be more than a categorical label,
instead signifying the strength of identification with one’s ethnic group (Phinney,
1990). Within this broad definition, there have been many specific conceptualiza-
tions (see Ashmore et al., 2004; Phinney & Ong, 2007). These include the affective
and cognitive attachment to one’s ethnic group, behavioral involvement in the
practices of one’s ethnic group, positive feelings for one’s ethnic group, the psy-
chological centrality and salience of one’s ethnic group membership, and belief in
the cultural values of one’s ethnic group (Knight et al., 2010).
3 Ethnic Identity Achievement, Identity Verification, and Self-worth 41
Underlying these various conceptualizations is the idea that ethnic identity is not a
static phenomenon, but rather is a developmental process of identity formation in
which people over time gain a greater understanding of what it means to be a
member of their ethnic group (Cross, 1991; Phinney, 1989; Quintana, 2007; Smith
& Silva, 2011). As a result of this developmental process, people may emerge with a
strong ethnic identity, acquire a more stable and secure understanding of their ethnic
identity, and develop positive feelings for their ethnic group (Phinney, 1989).
Based on past work in identity development (Erikson, 1968; Marcia, 1966, 1980),
this identity formation process has been studied in relation to the concept of ethnic
identity achievement (Phinney, 1989, 1993). Ethnic identity achievement refers to
the outcome of this developmental process where people develop a secure and stable
sense of their ethnic group membership informed by an exploration of its underlying
social meanings (Phinney, 1989, 1993; Phinney & Ong, 2007). At the beginning of
this identity formation process, people have an unexamined ethnic identity, which
reflects little concern for their ethnic identity or a willingness to define it in
accordance with the views of others (Phinney, 1990). At some point, a person may
develop an awareness of the political and social significance of their ethnic identity,
possibly by experiencing an act of prejudice and discrimination, which compels
them to start exploring the meanings of their ethnic identity (Phinney, 1989).
The process of exploration may include attending cultural events or speaking with
friends about their shared ethnic heritage and values. As people explore the mean-
ings of their ethnic identity and embed themselves in the cultural practices of their
ethnic group, they become more committed to their ethnic identity, which is defined
as a clear understanding of one’s ethnic group membership (Phinney, 1990). The
identity formation optimally concludes with an achieved ethnic identity marked by
high levels of commitment and exploration (Phinney, 1989). Those with an achieved
ethnic identity have a clear, stable, and secure sense of their ethnic identity, informed
and buttressed by the knowledge derived from the exploration of its social
meanings (Phinney, 1989).
An achieved ethnic identity is regarded as key to helping people of color navigate
the social obstacles they experience, which are often linked to negative stereotypes
of their ethnic identity (Quintana, 2007). The achievement process allows them to
overcome these obstacles by immersing themselves in the positive elements of their
ethnic culture and heritage, and thus feel good about themselves as a person.
Research supporting this hypothesis has found ethnic identity to be positively
associated with psychological well-being (Branscombe et al., 1999; Kiang et al.,
2006) and self-esteem (Martinez & Dukes, 1997; Phinney et al., 1997). Other work
examining ethnic identity achievement, specifically, also has found that it is associ-
ated with positive psychological health outcomes including self-esteem (Costigan
et al., 2010; Ghavami et al., 2011; Phinney & Alipuria, 1990; Phinney & Chavira,
1992; Yip & Fuligni, 2002).
Despite this prior research, work examining the theoretical mechanisms underly-
ing the relationship between ethnic identity achievement and self-esteem has been
limited. As noted above, an important focus for many of these developmental
scholars has been on how an achieved ethnic identity gives people the knowledge
and understanding of their ethnic heritage and culture to develop positive feelings for
their ethnic group (Phinney, 1989, 1991). Examining a sample of college students,
Ghavami et al. (2011) found that those with higher levels of ethnic identity achieve-
ment reported more positive feelings for their ethnic group (affirmation) and thus
experience enhanced self-esteem. This explanation, however, does not factor in the
interactional dynamic, fostered by the achievement process, which could account for
the positive feelings people develop for their ethnic group.
Research and theory within the ethnic identity achievement model offer a starting
point for understanding how this interactional dynamic may manifest. Prior work has
found that those with an achieved ethnic identity feel stronger ingroup ties to
members of their ethnic group (Ghavami et al., 2011). In a qualitative study, Case
and Hernandez (2013) examined the experiences of Latino college students in a four-
year leadership program designed to facilitate exploration of the students’ Latino
heritage. These authors found that the bonds developed with other students in the
first year of the program were beneficial and a “necessary building block” for the
students’ continued success (Case & Hernandez, 2013, p. 88). The students eventu-
ally developed “a level of comfort for their Latino/a peer group” (Case & Hernandez,
2013, p. 88), suggesting strong ties to members of their ethnic group.
Connecting this prior work on ingroup ties to principles from identity theory, we
offer a further elaboration of this interactional dynamic. We argue that the explora-
tion process undergirding an achieved ethnic identity fosters the development of a
personal network of others with whom people develop ingroup ties based on their
shared ethnic identity meanings. These shared identity meanings help people secure
verification of their ethnic identity, which enhances the positive feelings they hold
for their ethnic group, and thus their self-esteem. Below, we cover this theoretical
process as outlined by identity theory. We then discuss our framing of self-esteem as
two distinct dimensions: group-specific self-worth that is tied to one’s ethnic identity
and global self-worth that is tied to one’s overall self-concept.
Identity Theory, Verification, and Self-esteem
Rooted in structural symbolic interactionism (Stryker, 1980), identity theory has

examined the central role of identity verification as a motivation for self-relevant
behavior, and its subsequent impact on psychological health outcomes such as
emotions and self-esteem (Burke & Stets, 2009). Identity theory posits that for a
given identity, people hold a set of meanings that define themselves, otherwise
known as their identity standard. People monitor their social interactions, perceiving
self-relevant meanings in the situation, and comparing those self-relevant meanings
to their identity standard meanings. Part of the self-relevant meanings are comprised
of individuals’ views of how they think others see them in the situation, known as
their reflected appraisals.
When the reflected appraisals of one’s identity in the situation do not match the
meanings held in the identity standard, people experience negative emotions. This
influences changes in behaviors and cognitions so that reflected appraisal meanings
better match identity standard meanings, thus achieving identity verification (Burke
& Stets, 2009; Stets et al., 2020). When people verify their identities, they experi-
ence positive emotions and enhanced self-esteem. These theoretical propositions
have been supported empirically, with identity verification linked to lower levels of
negative emotions and depression (Burke & Harrod, 2005) and higher self-esteem
(Burke & Harrod, 2005; Burke & Stets, 1999; Cast & Burke, 2002; Stets & Burke,
2014).
Although research in identity theory has not studied the relationship between
ethnic identity achievement and the development of ingroup ties, a growing body of
work has examined how other identity dynamics are impacted by one’s immediate
social networks, such as one’s friends, peers, and family (i.e., proximate social
structures) (Merolla et al., 2012; Serpe et al., 2019; Stryker et al., 2005; Yarrison,
2016). Examining college students, Merolla et al. (2012) found that continued
participation in a science training program, a proximate social structure that exposed
respondents to other science students, enhanced the number of ties and the affective
strength of ties to other science students.
Implied in this past work is the potential benefit of not just proximate social
structures, but proximate social structures in which people share the same identities
and identity meanings. Prior research suggests that when people are nested within
these homogenous proximate social structures (Yarrison, 2016), they are more likely
to receive support for their identity and consequentially report their identity as
prominent or important, and thus experience greater self-esteem. These findings
have been supported empirically (Serpe et al., 2019). This same research has
suggested that homogenous proximate social structures also might contribute to
enhanced identity verification (Serpe et al., 2019).
Exploring this question in greater detail, research has examined how identity
verification is influenced by personal networks of others within proximate social
structures with whom people share the same identities and identity meanings
(Grindal & Trettevik, 2019; Stets et al., 2021). Examining college students, Grindal
and Trettevik (2019) found that when people perceived similar student identity
meanings with their close friends, and were thus were nested within a homogenous
proximate social structure, they experienced greater identity verification of their
student identity.
Stets et al. (2021) offer a theoretical elaboration of this process by drawing on the
concept of homophily or the tendency of people to form ties with similar others.
While acknowledging prior literature that broader structural factors influence the
placement of similar others within intermediate and proximate social structures, they
argue that agentic identity-based processes strengthen these homophilous ties. Spe-
cifically, they suggest a reciprocal relationship between identity verification and
group-based ties, which reinforce homophily within the group. Not only does being
around others with shared identity meanings help people secure identity verification,
but the affective responses to identity verification further reinforce and strengthen
these homophilous ties and the ability to secure future identity verification.
We suggest that the process of attaining an achieved ethnic identity plays out
through these identity processes. When people are faced with negative evaluations of
their ethnic group, part of the exploration process underlying the attainment of an
achieved ethnic identity involves people developing ties to others within their
proximate social structure with whom they share ethnic identity meanings, effec-
tively making the proximate social structure more homogenous and helping them
secure identity verification. Thus, the exploration process not only provides people
the knowledge of the positive elements of their ethnic heritage and culture to feel
more positively about their ethnic identity, it also helps them secure consistent
verification of their ethnic identity. This verification further enhances the positive
feelings they hold for their ethnic group, thereby enhancing their self-esteem.
Dimensions of Self-esteem
Self-esteem has traditionally been defined and studied as a unidimensional construct

involving the positive or negative evaluation of the global self-concept (Huang &
Dong, 2012; Rosenberg, 1979). Research in identity theory has expanded on the
concept of self-esteem in two important ways. First, it has distinguished between
three dimensions of self-esteem: self-worth, self-efficacy, and self-authenticity (Stets
& Burke, 2014). Second, it has distinguished between self-esteem tied to a particular
identity (specific self-esteem) and self-esteem tied to the global self-concept (global
self-esteem) (Ervin & Stryker, 2001; Rosenberg et al., 1995; Serpe et al., 2019). We
consider each of these in turn, and then their dual involvement in the identity
verification process.
Stets and Burke (2014) found empirical support that global self-esteem is com-
posed of three dimensions, each of which can be generally tied to one of the three
bases of identity. Authenticity-based esteem refers to feelings that reflect one’s true
or authentic self. This basis of self-esteem is tied to person identities, which are those
sets of self-meanings that define oneself as unique from others, such as being more
(or less) dominant or moral compared to others. When the dominant identity or
moral identity is verified, individuals should feel authentic. Efficacy-based esteem is
the feeling that individuals have control over their environment. This dimension of
self-esteem is tied to role identities, which are those sets of self-meanings that people
have for the roles they occupy in society such as being a parent, student, or
profession. When the parent, student, or professor identity is verified, individuals
should feel effective over their world.
Finally, worth-based esteem is how worthwhile and valuable people feel them-
selves to be. This dimension is most consistent with Rosenberg (1979) original
conceptualization of self-esteem (Markowski & Serpe, 2018). Worth-based esteem
is tied to social identities, which are the sets of meanings people have for their
membership in a category such as membership in a racial or ethnic category or a
religious category, or membership in a group such as membership in a union or local

softball team. When a categorical or group identity is verified, individuals should
experience a sense of belongingness and inclusion. Since our focus is on the ethnic
identity, we address the dimension of worth-based esteem in more detail.
Membership in social identities reflects the human motivation to be liked and find
acceptance and belonging with others (Stets & Burke, 2014). People seek member-
ship in and develop meanings for social identities that help ensure they receive
acceptance from others and are well-integrated into the social group. Identity
verification operates as the “glue” that reaffirms these feelings. When people receive
verification of their social identity, it confirms that they are worthy and liked (Stets &
Burke, 2014). Ethnic identities operate in this manner. The process of ethnic identity
achievement moves people into personal networks composed of others who share
ethnic identity meanings. These others provide verification of one’s ethnic identity,
which reaffirms that one is liked, accepted by their ethnic group, and have worth.
Thus, we expect identity verification of one’s ethnic identity should enhance self-
worth.
Research in identity theory also has drawn a distinction between specific self-
esteem and global self-esteem (Ervin & Stryker, 2001; Rosenberg et al., 1995; Serpe
et al., 2019). A key tenet of identity theory is that the self is multifaceted, made up of
many identities (Burke & Stets, 2009). Thus, evaluations of one’s self can be specific
to a given identity, which is known as specific self-esteem, or made generally of the
whole self, which is known as global self-esteem. As the self is multifaceted and
composed of many identities, scholars view the evaluation of the global self (global
self-esteem) as being largely sustained by the evaluations of the individual identities
within the self (Brenner et al., 2018; Hoelter, 1986; Rosenberg et al., 1995). For
instance, Hoelter (1986) found that evaluations of 12 specific facets of the self
collectively explained about 60% of the variance in global self-esteem. Thus, insofar
as global self-esteem to some extent reflects the specific evaluations of all the
identities in the self, a change in the specific self-esteem of one identity should
have an incremental impact on one’s global self-esteem (Rosenberg et al., 1995).
Prior research has found support for this proposition with specific self-esteem
positively influencing global self-esteem (Brenner et al., 2018; Hoelter 1986;
Rosenberg et al., 1995; Serpe et al., 2019), especially when the identity in question
is prominent or important (Rosenberg et al., 1995).
In addition, prior work suggests that identity processes, including identity veri-
fication, may influence global self-esteem indirectly by enhancing the specific self-
esteem tied to a given identity (Serpe et al., 2019). When an identity is verified,
people receive confirmation of who they are within that identity. Thus, the positive
sense of self that emerges with identity verification should principally be linked to
the verified identity. As verification increases the specific self-esteem tied to the
verified identity, we should then expect the latter to incrementally influence one’s
global self-esteem.
Given these theoretical extensions of the self-esteem concept, we examine how
the verification of one’s ethnic identity impacts self-worth, the dimension of self-
esteem tied to the verification of social identities. Further, we distinguish between
group-specific self-worth tied to the ethnic identity and global self-worth tied to the
global self-concept. We expect that the verification of one’s ethnic identity should
enhance global self-worth by increasing group-specific self-worth, or the feeling that
one is liked and accepted by members of their ethnic group. Prior work has examined
the relationship between the verification of social identities and global self-worth
(Burke & Harrod, 2021; Stets & Burke, 2014). Also, research in the ethnic identity
literature has examined how concepts analogous to group-specific self-esteem posi-
tively impact one’s global self-esteem (Costigan et al., 2010; Ghavami et al., 2011).
However, no known work has examined how the verification of social identities
impact global self-worth by enhancing group-specific self-worth. We turn next to a
summary of the theoretical model and its specified hypotheses.
Summary and Hypotheses
We posit that the relationship between ethnic identity achievement and enhanced
global self-worth can be, in part, explained through identity verification. The process
through which people develop an achieved ethnic identity places them within a
personal network of others with whom they share ethnic identity meanings. This
sharing of meanings fosters verification of one’s ethnic identity, which increases
their sense of belonging and worth as a member of their ethnic group (group-specific
self-worth). Since global self-worth is sustained through the evaluations of specific
social identities, we further expect specific self-esteem tied to one’s ethnic identity to
be positively associated with global self-esteem. Therefore, we hypothesize:
H1: Ethnic identity achievement is positively associated with ethnic identity
verification.
H2: Ethnic identity verification is positively associated with group-specific self-
worth.
H3: Group-specific self-worth is positively associated with global self-worth.
Consistent with the original formulation of the ethnic identity achievement
model, we expect that the exploration process underlying the attainment of an
achieved ethnic identity also exposes people to the positive elements of their ethnic
heritage and culture. This aids in counteracting the negative stereotypes and obsta-
cles many ethnic group members face in their everyday lives, thus enhancing the
positive feelings people hold about their ethnic group. Therefore, we hypothesize:
H4: Ethnic identity achievement is positively associated with group-specific self-
worth.
The full theoretical model is displayed in Fig. 3.1. In addition to the four
hypothesized pathways, we also test the non-hypothesized pathways (marked by
dashed lines) to assess mediation as well as the effects of four demographic
control variables: ethnicity, gender, age, and socioeconomic status.
Group Specific
Self-Worth
Ethnic Identy
Achievement
Global
Self-Worth
Ethnic Identy
Verificaon
Demographics:
Gender
SES
Ethnicity
Age
Fig. 3.1 Theoretical Model
Method
Data
The cross-sectional survey data for this study are drawn from undergraduate students
at a large southwestern university in 2016–2017. The university has an ethnically
diverse undergraduate student body with an approximate ethnic-racial distribution of
41% Latino, 34% Asian, 12% White, 4% African American, and 6% belonging to
another ethnic-racial group or multiple ethnic-racial groups. A total of 1686 respon-
dents completed the survey.
Sample
We included respondents in our analysis if they identified as white, Asian, African

American, or Latino and provided complete data. The subsamples sizes for the other
ethnic groups (e.g., Native American, Middle Eastern descent) were not large
enough to permit meaningful by-group comparisons. The final sample consisted of
1504 respondents with an ethnic-racial breakdown of 53% Latino, 30% Asian, 11%
white, and 6% African American. The gender breakdown of the sample was 74%
female and 26% male. This figure varies from the larger school population because
our sample was drawn primarily from sociology classes, which are populated more
by women than men. The median parental income of the respondents was $37,500,
and the median level of highest parental education completed was some high school.
Measures
Global Self-worth: Global self-worth was measured with a seven-item scale drawn
from recent identity theory research (Stets & Burke, 2014). Stets and Burke (2014)
developed and validated a three-factor measure of self-esteem. Each factor corre-
sponds to one of the three bases of identity. We focus on the self-worth items that
correspond to the ethnic social identity. Each item measured how positively the
respondent felt about themselves (e.g., “I feel I am a person of worth, at least on an
equal plane with others.” “I feel that I have a number of good qualities.”). An
exploratory factor analysis indicated a one-factor solution (see Table 3.1). The
items were averaged to create a scale ranging from 0–4 with a higher score reflecting
greater global self-worth (α ¼ 0.94).
Group-Specific Self-worth: We measured group-specific self-worth with four items
drawn from a previously validated scale from the social identity theory literature
(Cameron, 2004). Cameron’s scale measures three dimensions of social identity, one
of which is affect or the extent to which people hold a positive or negative emotional
evaluation of themselves as a member of their social group (e.g., “In general, I am
glad to be a member of my ethnic group.” “Generally, I feel good when I think about
myself as a member of my ethnic group.”). We used these four items as our measure
of group-specific self-worth. Two negatively worded items were reverse coded. Each
of the four items had six response options ranging from “strongly disagree” (0) to
“strongly agree” (5). An exploratory factor analysis indicated a one-factor solution
(see Table 3.2). The items were averaged to create a scale ranging from 0–5 with a
higher score reflecting greater group-specific self-worth (α ¼ 0.80).
Ethnic Identity Verification We measured ethnic identity verification with a
validated 14-item scale drawn from prior identity theory research (Burke
Table 3.1 Factor Loadings for Global Self-Worth Items

Items Loadings
I feel I am a person of worth, at least on an equal plane with others 0.74
I feel that I have a number of good qualities 0.84
I take a positive attitude toward myself 0.86
On the whole, I am satisfied with myself 0.84
I usually feel good about myself 0.87
I feel I have much to offer as a person 0.80
I have a lot of confidence in the actions I undertake in my life 0.82
Eigenvalue 4.77
Cronbach’s alpha 0.94
Table 3.2 Factor loadings for group-specific self-worth items

Items Loadings
In general, I am glad to be a member of my ethnic group 0.81
Generally, I feel good when I think about myself as a member of my ethnic group 0.77
I often regret that I am member of my ethnic groupa 0.69
I don’t feel good about being a member of my ethnic groupa 0.73
Eigenvalue 2.25
a
Reverse-coded
& Harrod, 2021)1 The items measure ethnic identity verification by assessing the
degree of difficulty people have engaging in seven ethnic practices tied to the
personal dimension of their ethnic identity (e.g., “Listening to music associated
with my ethnicity.”), and seven ethnic practices tied to the heritage dimension of
their ethnic identity (e.g., “Engaging in ethnic traditions.”). Traditionally, identity
theorists have measured identity verification by asking respondents to self-report
identity meanings and the meanings of the reflected appraisals they receive for
that identity, from which measures of identity discrepancy and verification are
calculated. This measure deviates from this traditional approach by reporting the
respondents’ direct subjective assessment of the identity discrepancy (Burke
& Harrod, 2021). Each of the fourteen items had five response options ranging
from “no difficulty at all” (0) to a “great deal of difficulty” (4).
An exploratory factor analysis indicated the presence of a two-factor solution
consisting of the heritage and personal dimensions (see Table 3.3). There was a clear
factor structure with each item loading highly on its specified factor (λ > 0.40). Since
there is no theoretical reason to believe the dimensions would operate differently in
our model, we averaged the 14 items into one measure of identity verification. The
scale ranged from 0 to 4, with a higher score indicating greater difficulty verifying
one’s identity (or greater identity non-verification). We then reverse coded the scale
to reflect less difficulty verifying one’s identity (or greater identity verification). This
overall scale demonstrated strong internal reliability (α ¼ 0.90).
Ethnic Identity Achievement We measured ethnic identity achievement with a
seven-item scale. We adapted our measure from the achievement subscale of the
Multigroup Ethnic Identity Measure (MEIM; Phinney, 1992). The MEIM measures
three dimensions of ethnic identity: achievement, affirmation and belonging, and
behaviors. The achievement subscale contains three questions that measure the
commitment to one’s ethnic identity, or a clear understanding of one’s ethnic
group membership (e.g., “I have a clear sense of my ethnic background and what
it means to me.”). The achievement subscale also contains four questions that
measure the exploration of one’s ethnic identity, or the extent to which one has
1
One item is different. The 14th item in the original scale is “maintaining ethnic gender roles.” Our
corresponding item is “maintaining an ethnic home.”
Table 3.3 Factor loadings for identity verification items

Factor loadings
Items Heritage Personal
Eating foods associated with my ethnicity 0.52 0.03
Engaging in ethnic traditions 0.74 0.08
Speaking the language associated with my ethnicity 0.61 0.04
Sharing my ethnic heritage with my family 0.81 0.05
Holding on to my ethnic beliefs and attitudes 0.73 0.10
Observing the religious traditions associated with my ethnicity 0.63 0.05
Participating in ethnic holidays/festivals 0.79 0.03
Having friends with the same ethnic background 0.02 0.67
Looking like my ethnicity 0.04 0.60
Listening to music associated with my ethnicity 0.13 0.56
Being in my ethnic community 0.10 0.72
Dating within my ethnic community 0.07 0.77
Wearing clothes associated with my ethnicity 0.08 0.65
Maintaining an ethnic home 0.21 0.61
Eigenvalue 5.09 4.78
Cronbach’s alpha 0.88 0.86
Bolded values indicate factor loadings greater than 0.40 and statistical significance of p < 0.05
explored the social meanings of their ethnic group membership (e.g., “I have spent
time trying to find out more about my own ethnic group, such as its history,
traditions, and customs.”). Prior psychometric research of the MEIM found that
two of the items in the achievement subscale performed poorly in their analyses,
which they suggested could be due to the items being negatively worded (Roberts
et al., 1999). To address this limitation, these two items were positively worded for
this study. Each item had five response options ranging from strongly disagree (0) to
strongly agree (4).
An exploratory factor analysis of these seven items indicated the presence of a
one-factor solution with the commitment and exploration items loading highly on
this one factor (see Table 3.4). This provides evidence that exploration and commit-
ment are two interactive processes that lead to the emergence of an achieved ethnic
identity, thus justifying the measurement of ethnic identity achievement as a unidi-
mensional construct (Phinney, 1992). We averaged the seven items in the scale, with
values ranging from 0 to 4 with a higher score indicating greater ethnic identity
achievement (α ¼ 0.87).
Demographics We measured four background factors: ethnicity, gender, age, and
socioeconomic status. Each of these demographic variables has been linked to global
self-worth and/or the ethnic identity constructs in this study. For instance, prior
research has found men and people with higher socioeconomic status report greater
levels of global self-worth (McMullin & Cairney, 2004; Twenge & Keith Campbell,
2002). Relatedly, prior research has found that age is positively associated with
global self-worth during early- and mid-adulthood (Robins et al., 2002). Also,
Table 3.4 Factor loadings for ethnic identity achievement items

Items Loadings
I have a clear sense of my ethnic background and what it means to me 0.71
I think a lot about how my life will be affected by my ethnic group membership 0.62
I am very clear about the role of my ethnicity in my life 0.73
I understand pretty well what my ethnic group membership means to me, in terms 0.70
how to relate to my own group and other groups
I have spent time trying to find out more about my own ethnic group, such as its 0.78
history, traditions, and customs
In order to learn about my ethnic background, I have often talked to other people 0.71
about my ethnic group
I have spent a lot of time trying to learn more about the culture and history of my 0.76
ethnic group
Eigenvalue 3.61
whites and those with other high-status markers (e.g., men) experience greater
identity verification (e.g., Burke & Harrod, 2021; Burke & Stets, 2009), while
African Americans and Latinos report greater levels of group-specific self-worth
(private regard) than whites and Asians (Hughes et al., 2017) and people of color
report higher levels of ethnic identity achievement compared to whites (Phinney &
Alipuria, 1990).
We measured ethnicity with four categories: African American, Asian, Latino,
and white (white is the reference group). We measured gender with two categories:
female and male (male is the reference group). Age was measured in years. Socio-
economic status was composed of two ordinal measures: mother’s education and
parental income. Since these two items were measured on different metrics, they
were each standardized and then averaged together to create a scale of socioeco-
nomic status with a higher score reflecting higher socioeconomic status.
Analytical Strategy
To test the theoretical model, we used structural equation modeling with maximum
likelihood estimation to test the parameter estimates of the structural model.2 We
tested the hypothesized effects of ethnic identity achievement on group-specific self-
worth and ethnic identity verification (net of controls), the direct effect of ethnic
identity verification on group-specific self-worth (net of controls and ethnic identity
achievement), and the direct effect of group-specific self-worth on global self-worth
(net of ethnic identity achievement, and ethnic identity verification). To assess
mediation and provide a more stringent test of the hypotheses, we also tested the
2
Since the model is just-identified, there were no goodness of fit statistics to estimate.
effects of the nonhypothesized pathways and the effects of the demographic control
variables on identity verification, group-specific self-worth, and global self-
worth. Lastly, we ran three groups models to test for ethnic and gender differences
in the effects of the theoretical model.
Findings
Table 3.5 shows the descriptive statistics and correlations of the observed variables.
The respondents reported moderate-to-high levels of self-esteem (M ¼ 2.09,
SD ¼ 0.62). For the ethnic identity variables, the respondents reported high levels
of group-specific self-worth (M ¼ 4.02, SD ¼ 1.05), high levels of ethnic identity
verification (M ¼ 3.13, SD ¼ 0.71), and moderate levels of ethnic identity achieve-
ment (M ¼ 2.61, SD ¼ 0.81). An examination of the correlations show preliminary
support for the hypotheses specified in the theoretical model. Ethnic identity
achievement was positively associated with group-specific self-worth (r ¼ .35,
p < 0.001) and identity verification (r ¼ 0.32, p < 0.001). Identity verification
was positively associated with group-specific self-worth (r ¼ .27, p < 0.001), which
was associated with global self-worth (r ¼ .27, p < 0.001).
The standardized path coefficients of the structural model are shown in Table 3.6.
In support of our first hypothesis, we found that ethnic identity achievement was
positively associated with identity verification (β ¼ 0.33, p < 0.001). In terms of the
background factors, we found that SES was negatively associated with identity
verification (β ¼ .06, p < 0.05). Further, compared to whites, all three other ethnic
groups reported lower identity verification: Asians (β ¼ .24, p < 0.001), African
Americans (β ¼ 0.18, p < 0.001), and Latinos (β ¼ 0.15, p < 0.001).
In support of our second hypothesis, we found that identity verification was
positively associated with group-specific self-worth (β ¼ 0.17, p < 0.001). We
also found that females reported greater group-specific self-worth (β ¼ 0.06,
p < 0.01), and age was inversely associated with group-specific self-worth
(β ¼ 0.05, p < 0.05). Further, compared to whites, all three other ethnic groups
reported greater group-specific self-worth: Asians (β ¼ 0.09, p < 0.05), African
Americans (β ¼ 0.09, p < 0.01), and Latinos (β ¼ 0.27, p < 0.001).
In support of our third hypothesis, we found that group-specific self-worth was
positively associated with global self-worth (β ¼ 0.21, p < 0.001). We additionally
found that age was positively associated with global self-worth (β ¼ 0.11,
p < 0.001) as was SES (β ¼ 0.08, p < 0.01). Further Asians reported lower global
self-worth than whites (β ¼ 0.15, p < 0.001). In support of our fourth hypothesis,
we found that identity achievement was positively associated with group-specific
self-worth (β ¼ 0.26, p < 0.001).
Collectively, these findings provide support for the theoretical model. Consistent
with the identity theory explanation, those with greater levels of ethnic identity
achievement experienced greater identity verification, which enhanced their group-
specific self-worth and, in turn, their global self-worth. Also, consistent with the
Table 3.5 Descriptive statistics and correlations of study variables
Variable 1 2 3 4 Mean SD Min Max
Global self-worth (1) 1.00 2.09 0.64 0 3
Group specific self-worth (2) 0.27*** 1.00 4.02 1.05 0 5
Identity Verification (3) 0.14*** 0.27*** 1.00 3.13 0.71 0 4
Identity Achievement (4) 0.23*** 0.35*** 0.32*** 1.00 2.61 0.81 0 4
Gender 0.00 0.13*** 0.09*** 0.09*** 0.74 0.44 0 1
Age 0.11*** 0.05* 0.01 0.00 20.04 2.32 18 40
SES 0.02 0.13*** 0.12*** 0.10*** 0.00 0.86 1.65 2.05
Asian 0.15*** 0.16*** 0.15*** 0.06* 0.30 0.46 0 1
Black/African American 0.09*** 0.00 0.08** 0.09*** 0.06 0.24 0 1
Latino 0.09*** 0.24*** 0.14*** 0.12*** 0.53 0.50 0 1
White 0.01 0.16*** 0.05* 0.17*** 0.11 0.31 0 1
3 Ethnic Identity Achievement, Identity Verification, and Self-worth
* p < 0.05, ** p < 0.01, *** p < 0.001

53
Table 3.6 Standardized coefficients of structural model

Group-specific self-
Independent variables Identity verification worth Global self-worth
Identity achievement 0.33*** 0.26*** 0.15***
Identity verification – 0.17*** 0.03
Group specific self-Worth – – 0.21***
Female 0.04 0.06** 0.04
Age 0.01 0.05* 0.11***
SES 0.06* 0.02 0.08**
Ethnicity+
Asian 0.24*** 0.09* 0.15***
Black/African American 0.18*** 0.09* 0.03
Latino 0.15*** 0.27*** 0.04
* p < 0.05, ** p < 0.01, *** p < 0.001 tests are two-tailed
traditional theoretical explanation posited by the ethnic identity development

scholars, ethnic identity achievement directly enhanced group-specific self-worth
and, in turn, global self-worth. We also found a direct effect between ethnic identity
achievement and global self-worth (β ¼ 0.15, p < 0.001), but we found no direct
effect between identity verification and global self-worth. These latter two results
suggest that the effect between identity verification and global self-worth was
completely mediated by group-specific self-worth, while the effect of ethnic identity
achievement on global self-worth was only partially mediated by the theoretical
processes specified in the model.
Finally, we ran three sets of groups models to examine the gender and ethnic
variation in the theoretical model. For each of these two grouping variables (gender
and ethnicity), we ran two models: one where the structural coefficients were
constrained to be equal by the groups, and one where the structural coefficients
were unconstrained and allowed to vary by the groups. In the test of the gender
variation in the theoretical model, we found no differences between the constrained
and unconstrained models, indicating no overall gender differences (χ2 ¼ 28.58,
df ¼ 21, p ¼ 0.12).
We ran two sets of groups models to test the ethnic variation in the theoretical
model. In the first analysis we tested for differences among the four ethnic groups
and found no overall differences between the constrained and unconstrained models,
indicating no ethnic differences (χ2 ¼ 53.60, df ¼ 45, p ¼ 0.18). In the second
analysis, we tested for differences between the white and non-white ethnic groups.
Again, we found no overall differences between the constrained and unconstrained
model, indicating no differences between whites and non-whites (χ2 ¼ 18.69,
df ¼ 15, p ¼ 0.23).
We followed each of these tests of the whole theoretical model with more
stringent tests of the variation in each hypothesized structural pathway. These tests
also showed no significant ethnic or gender variation.
Discussion
Ethnic identity achievement has been viewed as essential for helping people over-
come the negative stereotypes and social obstacles tied to their ethnic group and
develop positive self-worth (Phinney, 1989). An achieved ethnic identity exposes
people to the positive elements of their ethnic heritage and culture, thus allowing
them to develop a secure and well-developed understanding of their ethnic group
membership. This enables them to develop positive feelings for their ethnic group
and thus enhance their global self-worth (Ghavami et al., 2011). However, this
explanation is incomplete since it does not explicitly factor in the interactive
dynamics involved in the development of an achieved ethnic identity.
To address this limitation, we integrated the concept of ethnic identity achieve-
ment with identity theory principles. We conceptualized the process of ethnic
identity achievement as one which moves people into homogenous proximate social
structures in which they develop ties to others with the same ethnic identity and
shared ethnic identity meanings. We hypothesized that this should result in greater
ethnic identity verification, thus enhancing group-specific self-worth. Extrapolating
from prior research (Hoelter, 1986), we suggested that global self-worth is to a large
extent sustained by the self-worth tied to social identities, and thus hypothesized that
group-specific self-worth would be positively associated with global self-worth.
Lastly, we hypothesized that the knowledge derived from the process of attaining
an achieved ethnic identity should expose people to the positive elements of their
ethnic heritage and culture, thus directly enhancing their group-specific self-worth
and ultimately their global self-worth. The tests of our theoretical model supported
these hypotheses.
Our findings contribute to the literature on ethnic identity achievement by
identifying identity verification as an additional source of group-specific self-worth
and global self-worth generated by ethnic identity achievement. Interestingly, we
still found a direct effect between ethnic identity achievement and global self-worth
even after accounting for the theoretical processes specified in our model. One
possible explanation is that the reflected appraisals that verify one’s ethnic identity
might also be verifying other higher-order person identities that share similar identity
meanings.
In the hierarchical control model outlined by Burke and Stets (2009), the reflected
appraisals people receive from lower-order social and role identities are also
received by higher-order person identities when those meanings are shared with a
person identity. Although Stets and Burke (2014) argue that verification of social
identities should primarily enhance global self-worth, they note that if the meanings
between a person identity and group identity are shared, we might expect verification
of the person identity to also influence global self-worth. This is consistent with their
findings in which verification of the moral identity not only enhanced self-authen-
ticity but also self-worth. For our study, the verifying reflected appraisals of one’s
ethnic identity might convey that one is dedicated and loyal to their ethnic group.
Ingroup loyalty has been documented as a central dimension of the moral identity
(Graham et al., 2013), thus these reflected appraisals might enhance self-worth by
verifying both the ethnic and moral identities.
We did not find any gender or ethnic differences in the effect sizes of the
theoretical model, including when we tested for differences between whites and
non-whites. This is somewhat surprising given prior research indicating that ethnic
identity is less prominent for whites than for people of color (Burke & Harrod, 2021;
Phinney & Alipuria, 1990). However, a closer look at our findings provides an
alternative explanation for how ethnic differences might play out in our model. We
ran exploratory analyses and found that levels of ethnic identity achievement were
higher for people of color than for whites. Given that people of color experience
greater discrimination and negative treatment due to their ethnicity (Kessler et al.,
1999; Pager & Shepherd, 2008), they may feel more compelled to explore the
meanings of their ethnic identity and develop an achieved ethnic identity. Thus,
while the theoretical processes in our model operate similarly across ethnic groups,
our findings suggest that the process of ethnic identity achievement is of greater
necessity and pursued more often by people of color.
Relatedly, we also found ethnic differences in the levels of identity verification.
However, unlike ethnic identity achievement, whites experienced greater ease in
being able to verify their ethnic identity. This is consistent with prior research
generally (Burke & Stets, 2009) and ethnic identities, in particular (Burke
& Harrod, 2021), where groups with higher status are better able to secure identity
verification. However, we did not find a similar effect for other status-based vari-
ables like gender and socioeconomic status. It is possible that since we looked at
ethnic identities, ethnicity may have been the most influential status marker that
people could use to secure identity verification.
Our findings also help extend identity theory’s understanding of self-esteem.
Prior work in identity theory has noted the importance of viewing self-esteem as a
multidimensional construct (Stets & Burke, 2014) as well as studying self-esteem on
both specific and global levels (Ervin & Stryker, 2001). While prior work has
examined how verification impacts global self-worth (Stets & Burke, 2014; Burke
& Harrod, 2021) and specific self-worth (Markowski & Serpe, 2018), ours is the first
known study to examine how verification impacts both constructs in relation to a
social identity. Our findings implicate the importance of group-specific self-worth in
the verification process. We found that the effect of verification on global self-worth
was fully accounted for by group-specific self-worth. That is, verification enhanced
the positive feelings people had for themselves globally by promoting more positive
feelings for their ethnic group specifically.
Our findings also have applied implications. Schools and colleges have adopted
awareness programs to help people from ethnic groups traditionally targeted by
discrimination and prejudice gain a better understanding of their ethnic group and
help promote positive behavioral and psychological health outcomes (e.g., Case &
Hernandez, 2013). The findings of our study can inform how these programs are
structured to be more effective. For instance, placing program participants into
cohorts can help ensure that they develop shared identity meanings with other
members of the program who are going through a similar exploration of their ethnic
group’s cultural heritage. This can help ensure the emergence of a homogenous
proximate social structure from which participants can secure identity verification.
Relatedly, our findings can inform how these programs are assessed. The success
of these programs can be monitored periodically by assessing how well participants
can secure identity verification and enhance their group-specific self-worth. Since
our findings indicate that the success of these programs is directly contingent on their
ability to facilitate these identity outcomes, these periodic assessments would help
program administrators make any necessary changes to make their programs more
effective.
This study used a cross-sectional design to test our theoretical model. While there
are good theoretical reasons to suppose the time-order of the pathways posited in this
model, it is plausible that some of these relationships exhibit feedback effects. For
instance, the positive repercussions of identity verification including enhanced
group-specific self-worth might drive people to develop stronger ties with members
of their ethnic group within their proximate social structure (Stets et al., 2021). This
might spur further exploration of one’s ethnic identity, thus increasing levels of
ethnic identity achievement. Additionally, prior research indicates that specific self-
esteem not only influences global self-esteem, but also that global self-esteem might
bias one’s evaluation of their specific identities—suggesting a feedback loop
between these self-esteem constructs (Rosenberg et al., 1995). Studies with longi-
tudinal data can more accurately model these processes to identify these feedback
effects. This research would extend and provide greater confidence in our findings.
Our study also sampled students at an ethnically diverse college where students of
color constitute the majority of the undergraduate student body. People in interme-
diate social structures often exhibit a homophily bias where they can come together
based on shared identities and characteristics (Stryker et al., 2005). For people of
color, this can help bring others into their proximate social structures who share the
same ethnic identity and similar ethnic identity meanings, thus spurring the process
of ethnic identity formation identified in this model. However, the same cannot be
said for people of color on college campuses generally, an intermediate social
structure where people of color are often a small minority of the student body – a
pattern that is especially pronounced at selective colleges (Monarrez & Washington,
2020). Thus, our findings may not be representative of students of color who attend
colleges with predominantly white student bodies. Future research that can test this
model among these students can provide a greater understanding of how structural
impediments can impact the tendency towards ethnic homophily and ethnic identity
development.
Despite these limitations, this study has helped theoretically elaborate on the
psychological resilience afforded by the ethnic identity achievement process by
specifying how it helps people secure verification of their ethnic identity, enhance
group-specific self-worth, and global self-worth. These findings also reinforce the
importance of ethnic identity achievement in the lives of people who are subject to
ethnic discrimination and prejudice. Lastly, our results can help guide the imple-
mentation of ethnic awareness programs for youth and young adults by assessing the
extent to which they promote group-specific self-worth and provide a support

network where participants can receive verification of their ethnic identity.
References
Ashmore, R. D., Deaux, K., & McLaughlin-Volpe, T. (2004). An organizing framework for
collective identity: Articulation and significance of multidimensionality. Psychological Bulletin,
130(1), 80–114.
Branscombe, N. R., Schmitt, M. T., & Harvey, R. D. (1999). Perceiving pervasive discrimination
among African Americans: Implications for group identification and Well-being. Journal of
Personality and Social Psychology, 77(1), 135–149.
Brenner, P. S., Serpe, R. T., & Stryker, S. (2018). Role-specific self-efficacy as precedent and
product of the identity model. Sociological Perspectives, 61(1), 57–80.
Burke, P. J., & Harrod, M. M. (2005). Too much of a good thing? Social Psychology Quarterly, 68
(4), 359–374.
ogy Quarterly, 62(4), 347–366.
Burke, P. J., & Harrod, M. M. (2021). Ethnic identity measurement and verification. In
P. S. Brenner, J. E. Stets, & R. T. Serpe (Eds.), Identities in action: New developments in
identity theory (pp. 11–37). Springer.
Cameron, J. E. (2004). A three-factor model of social identity. Self and Identity, 3(3), 239–262.
Case, K. F., & Hernandez, R. (2013). “But still, I’m Latino and I’m proud”: Ethnic identity
exploration in the context of a collegiate cohort program. Christian Higher Education, 12
(1–2), 74–92.
Cast, A. D., & Burke, P. J. (2002). A theory of self-esteem. Social Forces, 80(3), 1041–1068.
Costigan, C. L., Koryzma, C. M., Hua, J. M., & Chance, L. J. (2010). Ethnic identity, achievement,
and psychological adjustment: Examining risk and resilience among youth from immigrant
Chinese families in Canada. Cultural Diversity and Ethnic Minority Psychology, 16(2),
264–273.
Cross, W. E. (1991). Shades of black: Diversity in African American identity. Temple University
Press.
Erikson, E. H. (1968). Identity: Youth and crisis. Norton.
Ervin, L. H., & Stryker, S. (2001). Theorizing the relationship between self-esteem and identity. In
T. J. Owens & S. Stryker (Eds.), Extending self-esteem theory and research: Sociological and
psychological currents (pp. 29–55). Cambridge University Press.
Ghavami, N., Fingerhut, A., Peplau, L. A., Grant, S. K., & Wittig, M. A. (2011). Testing a model of
minority identity achievement, identity affirmation, and psychological Well-being among ethnic
minority and sexual minority individuals. Cultural Diversity and Ethnic Minority Psychology,
17(1), 79–88.
Graham, J., Haidt, J., Koleva, S., Motyl, M., Iyer, R., Wojcik, S. P., & Ditto, P. H. (2013). Moral
foundations theory: The pragmatic validity of moral pluralism. In P. Devine & A. Plant (Eds.),
Advances in experimental social psychology (Vol. 47, pp. 55–130). Academic.
Grindal, M., & Trettevik, R. (2019). Perceived similarity in identity meanings, identity verification,
and positive emotions. In J. E. Stets & R. T. Serpe (Eds.), Identities in everyday life (pp. 35–52).
Oxford University Press.
Hoelter, J. W. (1986). The relationship between specific and global evaluations of self: A compar-
ison of several models. Social Psychology Quarterly, 49(2), 129–141.
Huang, C., & Dong, N. (2012). Factor structures of the Rosenberg self-esteem scale: A meta-
analysis of pattern matrices. European Journal of Psychological Assessment, 28(2), 132–138.
Hughes, D. L., Del Toro, J., & Way, N. (2017). Interrelations among dimensions of ethnic-racial
identity during adolescence. Developmental Psychology, 53(11), 2139–2153.
Kessler, R. C., Mickelson, K. D., & Williams, D. R. (1999). The prevalence, distribution, and
mental health correlates of perceived discrimination in the United States. Journal of Health and
Social Behavior, 40(3), 208–230.
Kiang, L., Yip, T., Gonzales-Backen, M., Witkow, M., & Fuligni, A. J. (2006). Ethnic identity and
the daily psychological Well-being of adolescents from Mexican and Chinese backgrounds.
Child Development, 77(5), 1338–1350.
Knight, G. P., Gonzales, N. A., Saenz, D. S., Bonds, D. D., German, M., Deardorff, J., Roosav,
M. W., & Updegraff, K. A. (2010). The Mexican American cultural values scale for adolescents
and adults. The Journal of Early Adolescence, 30(3), 444–481.
Marcia, J. E. (1966). Development and validation of ego-identity status. Journal of Personality and
Social Psychology, 3(5), 551–558.
Marcia, J. E. (1980). Identity in adolescence. In J. Adelson (Ed.), Handbook of adolescent
psychology (pp. 159–187). Wiley.
Martinez, R. O., & Dukes, R. L. (1997). The effects of ethnic identity, ethnicity, and gender on
adolescent well-being. Journal of Youth and Adolescence, 26(5), 503–516.
Markowski, K. L., & Serpe, R. T. (2018). Identity theory paradigm integration: Assessing the role
of prominence and salience in the verification and self-esteem relationship. Advances in Group
Processes, 35, 75–102.
McMullin, J. A., & Cairney, J. (2004). Self-esteem and the intersection of age, class, and gender.
Journal of Aging Studies, 18(1), 75–90.
Merolla, D. M., Serpe, R. T., Stryker, S., & Wesley Schultz, P. (2012). Structural precursors to
identity processes: The role of proximate social structures. Social Psychology Quarterly, 75(2),
149–172.
Monarrez, T., & Washington, K. (2020). Racial and ethnic representation in postsecondary
education. Urban Institute. https://www.urban.org/research/publication/racial-and-ethnic-repre
sentation-postsecondary-education.
Pager, D., & Shepherd, H. (2008). The sociology of discrimination: Racial discrimination in
employment, housing, credit, and consumer markets. Annual Review of Sociology, 34, 181–209.
Phinney, J. (1993). A three-stage model of ethnic identity development. In M. Bernal & G. Knight
(Eds.), Ethnic identity: Formation and transmission among Hispanics and other minorities
(pp. 61–79). State University of New York Press.
Phinney, J. S. (1989). Stages of ethnic identity development in minority group adolescents. The
Journal of Early Adolescence, 9(1–2), 34–49.
Phinney, J. S. (1990). Ethnic identity in adolescents and adults: Review of research. Psychological
Bulletin, 108(3), 499–514.
Phinney, J. S. (1991). Ethnic identity and self-esteem: A review and integration. Hispanic Journal
of Behavioral Sciences, 13(2), 193–208.
Phinney, J. S. (1992). The multigroup ethnic identity measure: A new scale for use with diverse
groups. Journal of Adolescent Research, 7(2), 156–176.
Phinney, J. S., & Alipuria, L. L. (1990). Ethnic identity in college students from four ethnic groups.
Journal of Adolescence, 13(2), 171–183.
Phinney, J. S., Cantu, C. L., & Kurtz, D. A. (1997). Ethnic and American identity as predictors of
self-esteem among African American, Latino, and white adolescents. Journal of Youth and
Adolescence, 26(2), 165–185.
Phinney, J. S., & Chavira, V. (1992). Ethnic identity and self-esteem: An exploratory longitudinal
study. Journal of Adolescence, 15(3), 271–281.
Phinney, J. S., & Ong, A. D. (2007). Conceptualization and measurement of ethnic identity: Current
status and future directions. Journal of Counseling Psychology, 54(3), 271–281.
Quintana, S. M. (2007). Racial and ethnic identity: Developmental perspectives and research.
Journal of Counseling Psychology, 54(3), 259–270.
Roberts, R. E., Phinney, J. S., Masse, L. C., Richard Chen, Y., Roberts, C. R., & Romero,
A. (1999). The structure of ethnic identity of young adolescents from diverse Ethnocultural
groups. The Journal of Early Adolescence, 19(3), 301–322.
Robins, R. W., Trzesniewski, K. H., Tracy, J. L., Gosling, S. D., & Potter, J. (2002). Global self-
esteem across the life span. Psychology and Aging, 17(3), 423–434.
Rosenberg, M. (1979). Conceiving the self. Basic Books.
Rosenberg, M., Schooler, C., Schoenbach, C., & Rosenberg, F. (1995). Global self-esteem and
specific self-esteem: Different concepts, different outcomes. American Sociological Review, 60
(1), 141–156.
Serpe, R. T., Long-Yarrison, F., Stets, J. E., & Stryker, S. (2019). Multiple identities and self-
esteem. In J. E. Stets & R. T. Serpe (Eds.), Identities in everyday life (pp. 53–71). Oxford
University Press.
Smith, T. B., & Silva, L. (2011). Ethnic identity and personal Well-being of people of color: A
meta-analysis. Journal of Counseling Psychology, 58(1), 42–60.
Stets, J. E., & Burke, P. J. (2014). Self-esteem and identities. Sociological Perspectives, 57(4),
409–433.
Stets, J. E., Savage, S. V., Burke, P. J., & Fares, P. (2020). Cognitive and Behavioral responses to
the identity verification process. In R. Stryker, R. T. Serpe, & B. Powell (Eds.), Identity and
symbolic interaction: Deepening foundation and building bridges (pp. 65–88). Springer.
Stets, J. E., Aldecoa, J., Bloom, Q., & Winegar, J. (2021). Using identity theory to understand
homophily in groups. In P. S. Brenner, J. E. Stets, & R. T. Serpe (Eds.), Identities in action: New
developments in identity theory (pp. 285–302). Springer.
Stryker, S., Serpe, R., & Hunt, M. (2005). Making good on a promise: The impact of larger social
structures on commitment. Advances in Group Processes, 22, 93–123.
Twenge, J. M., & Keith Campbell, W. (2002). Self-esteem and socioeconomic status: A meta-
analytic review. Personality and Social Psychology Review, 6(1), 59–71.
Yarrison, F. (2016). Contextualizing proximate social structure in identity theory. In J. E. Stets &
R. T. Serpe (Eds.), New directions in identity theory and research (pp. 343–365). Oxford
University Press.
Yip, T., & Fuligni, A. J. (2002). Daily variation in ethnic identity, ethnic Behaviors, and psycho-
logical well–being among American adolescents of Chinese descent. Child Development, 73(5),
1557–1572.
Zeiders, K. H., Umana-Taylor, A. J., & Derlan, C. L. (2013). Trajectories of depressive symptoms
and self-esteem in Latino youths: Examining the role of gender and perceived discrimination.
Developmental Psychology, 49(5), 951–963.
Chapter 4
Racial Identity Among Blacks and Whites
in the U.S.
K. Jill Kiecolt, W. Carson Byrd, Hans Momplaisir, and Michael Hughes
Racial Identity Among Blacks and Whites in the U.S.
Racial identity, the self-categorization as a member of a racial group, and the

personal significance of being a group member, informs people’s worldviews and
can motivate political and social action (Stryker et al., 2000). Yet despite great
interest in racial identity, gaps in the literature remain. Studies have discovered that
African Americans develop positive racial identities, and that racial identity is
multidimensional (Cross, 1991; Phinney, 1991; Sellers et al., 1998; Vandiver
et al., 2002). Other studies have documented robust benefits of a strong and positive
racial identity for Blacks’ self-esteem and mental health (Branscombe et al., 1999;
Brown et al., 2002; Hughes et al., 2015; Ida & Christie-Mizell, 2012).
More recently, scholars have advanced an understanding of Whites’ racial iden-
tities (Jardina, 2019; Lewis, 2004; McDermott & Samson, 2005). Whiteness studies
have explored Whites’ racial identity and investigated how Whites’ racial identity in
tandem with their status as a dominant group helps legitimate and perpetuate racial
inequality (Croll, 2007; Hartmann et al., 2009; Hughey, 2012; Hunt & Reichelmann,
2019; Jardina, 2019; McDermott & Samson, 2005).
Despite these advances, more comparative studies are needed. Studies of racial
identity based on representative national samples often deal with one group, such as
K. J. Kiecolt (*) · M. Hughes

Virginia Tech, Blacksburg, VA, USA
W. C. Byrd
University of Michigan, Ann Arbor, MI, USA
H. Momplaisir
Trinity Washington University, Washington, DC, USA

62 K. J. Kiecolt et al.
Blacks or Whites (Croll, 2007; Hunt & Reichelmann, 2019; Jardina, 2019; Thornton
et al., 2013). With some exceptions (Kiecolt & Hughes, 2017; Stets & Fares, 2019),
comparative studies typically have adolescent or college student samples (Johnson
et al., 2005; Syed & Juang, 2014). While these are useful, they leave open the
question of whether racial identity differs across cohorts.
Our study draws on social identity theory, identity theory, and life course theory
to investigate the nature and determinants of different racial identity dimensions
among Blacks and Whites in the U.S. We use data from the 2014 General Social
Survey (Smith et al., 2015) to examine five aspects of racial identity: its importance
to one’s self-concept; its salience, how often it is invoked in interactions; pride in
being a group member; how often others verify it; and perceived public respect for
one’s racial group. By examining multiple dimensions of racial identity, we can gain
a more nuanced understanding of how Blacks and Whites view their racial identities
and experience them in everyday life.
Theory
Race and Racial Identity
Social identities entail awareness of belonging to a social category or group, along

with the value and emotional significance of belonging (Ashmore et al., 2004;
Ellemers & Haslam, 2011; Tajfel & Turner, 1986). Racial identities begin with
identifying as a member of a racial group. Among identifiers, the meaning and
significance of racial identification varies along several dimensions, e.g., how
closely people’s self-views are tied to their racial group (Sellers et al., 1998; Stets
& Fares, 2019). To explain variability in racial identities among Blacks and Whites,
we draw on social identity theory and identity theory, which elucidate how macro-
structural factors such as race influence racial identity through proximal factors and
psychological processes.
Race is a major dimension of social inequality in U.S. society. Racism, an
organized system based on categorizing Whites as superior and other racial groups
such as Blacks as inferior, pervades social institutions (Williams & Mohammed,
2013). As a result, Whites have more access to opportunities and resources than do
Blacks (Bonilla-Silva, 2017; Omi & Winant, 2014). The advantages of whiteness for
navigating society have evolved and are sustained through policy and social inter-
action. As racial identities reflect the system of racial inequality in which they are
anchored (Harris, 1993), race is likely to be related to dimensions of racial identity.
Some identities are more important than other identities to people’s sense of who
they are (Stets & Serpe, 2013). Identity importance refers to how prominent an
identity is to one’s self-concept (Brenner et al., 2014; McCall & Simmons, 1978;
Rosenberg, 1979; Stryker & Serpe, 1994). For example, one’s racial identity may be
more (or less) prominent in one’s self-concept.
4 Racial Identity Among Blacks and Whites in the U.S. 63
We expect racial identity to be more important to one’s self-concept (that is, more
prominent) among Blacks than Whites. In sociology, theory on whiteness links
White privilege to weak racial identities among Whites (Croll, 2007).1 Whites’
long-held privileged status on the racial hierarchy enables them to downplay their
status as racial actors (Lewis, 2004).
Among U.S. adults, 38.2% of Blacks, but only 10.3% of Whites selected their
racial identity as one of their three most important identities (Smith, 2007). Racial
identity was more important for ethnic minority youth as well (Johnson et al., 2005;
Roberts et al., 1999). Racial minority group members also felt closer and more
attached to their group than Whites did (Syed & Juang, 2014; Thornton et al., 2012;
Wong & Cho, 2005). Similarly, African Americans had a stronger sense of common
fate with their racial group than Whites did. About 40% of African Americans but
only 20% of Whites had a strong sense of common fate (Bobo & Johnson, 2000). In
summary, based on theory and previous research, we predict that:
Hypothesis 1a: Racial identity will be more important to Blacks than Whites.
Identity salience refers to the likelihood that an identity will be invoked in
situations. More salient identities are more likely to be enacted and in ways that
correspond with the meaning of the identity (Brenner et al., 2014; Burke & Stets,
2009; Stryker & Serpe, 1994). Greater racial identity salience implies that one’s
racial identity more often guides one’s behavior. Identities that are more important
also tend to be more salient, that is, more often invoked across situations (Brenner
et al., 2014; Burke & Stets, 2009; Stryker & Serpe, 1994).2
Racial identity is likely to be more salient to Blacks than to Whites. In interactions
with Whites, Blacks’ racial identity may be invoked regardless of their wishes
(Brenner et al., 2014). In public places, Blacks’ racial identity is invoked when
they anticipate dealing with incivility from Whites, based on past experiences of it
(Feagin, 1991). In organizations, Blacks’ racial identity becomes salient when they
are required to play down their race in order to make Whites feel comfortable
(Thornhill, 2015).
In contrast, Whites do not have to think about race as often as members of
minority groups do, and many Whites claim not to (Lewis, 2004). Whites often do
1
Optimal distinctiveness theory in psychology provides a psychological rationale for expecting
greater racial identity importance among Blacks than Whites. The theory proposes that people most
identify with groups that satisfy their needs for inclusion and a sense of group distinctiveness
(Leonardelli et al., 2010). For Blacks, identifying with a numerical racial minority group helps
satisfy both needs. For Whites, identifying with their racial group helps satisfy a need for inclusion,
but not for distinctiveness, because Whites are the majority group. Whites may satisfy a need for
group distinctiveness by identifying with an ethnic group rather than their racial group. For Whites,
however, this is a personal choice (Waters, 1990).
2
The Multidimensional Model of Racial Identity (MMRI) defines salience more situationally as
“the extent to which a person’s race is a relevant part of his or her self-concept at a particular
moment in time” (Rowley et al., 1998, p. 717). It depends on what is happening in a situation and
how central a person’s racial identity is. Measures of racial identity based on the MMRI have not
included salience (Sellers et al., 1998).
not think about their race in everyday interactions. Some choose to live in predom-
inantly White environments that support a color-blind ideology (Hagerman, 2014).
They do not necessarily view themselves or other Whites in terms of their race, so
they see interactions with other Whites as race neutral (Lewis, 2004). We predict
that:
Hypothesis 1b: Racial identity will be more salient for Blacks than Whites.
Pride in a group or category refers to how positively (or negatively) people feel
toward their group and their membership in it (Luhtanen & Crocker, 1992; Rosen-
berg, 1979). Pride in one’s racial group is also termed private regard (Sellers et al.,
1998). Pride comes from associating positive meanings with one’s group. For
example, Black Americans have long seen themselves as strong and resilient, having
overcome many injustices (Campbell, 2017).
According to social identity theory, people who identify with a social group or
category want to favorably evaluate their group and to feel proud of belonging to
it. Members of high-status, dominant groups can do so easily because social
consensus ascribes positive characteristics to their group. In a recent study, 54% of
Whites said that Whites had “a great deal” or “a lot” to be proud of, whereas only
14% said “a little” or “none at all” (Jardina, 2019). Members of devalued racial
minority groups are assumed to work harder to feel good about their group (Stets &
Burke, 2000; Tajfel & Turner, 1986) and to counter negative cultural stereotypes
(Williams & Mohammed, 2013). The theory assumes that most group members
succeed in doing so.
Consistent with this assumption, African American and White college students
have high private regard (Sellers et al., 1998; Vandiver et al., 2002; Johnson et al.,
2005). Black adults, too, evaluate their group very positively (Hughes & Demo,
1989; Hughes et al., 2015; Ida & Christie-Mizell, 2012) and reject negative stereo-
types of it (Vandiver et al., 2002). Because minority group members presumably
make more conscious efforts to feel good about their group, we predict that:
Hypothesis 1c: Pride in one’s group will be higher among Blacks than Whites.
Verification refers to the perceived congruence between one’s own and others’
view of oneself (Burke & Stets, 2009). It refers to concordance between the
meanings people assign to their racial identity and their reflected appraisals, that
is, how much people think others see them in the same way they see themselves
(Burke & Stets, 2009).
Racial inequality also may influence identity verification. Identity theory pro-
poses that when people’s identities are verified, they feel good. When identities are
not verified people feel bad (Burke & Stets, 2009). Higher-status people are thought
to have their identities verified more often, because they have more resources for
maintaining their identities. In addition, they are apt to get more identity support
from lower-status others, because of their greater power and status (McLeod, 2015;
Stets & Harrod, 2004).
With respect to race, Whites may want others to verify their claims of being
“moral, logical, rational, objective” as Whites (Hughey & Byrd, 2013, p. 978).
Blacks also want others to verify the positive meanings that most group members
associate with their racial identity (Hughes et al., 2015). However, on average
Blacks have fewer resources than Whites for verifying their identity, e.g., education,
wealth, status, power (Stets & Harrod, 2004). In addition, their positive identity
meanings are more often challenged by experiences of racial discrimination, that is,
unfair treatment based on negative racial stereotypes (Branscombe et al., 1999;
Crocker & Major, 1989; Kiecolt et al., 2016). Blacks may try to preempt encounters
that might fail to verify their positive racial identities (Oyserman & Swim, 2001).
We might expect identity verification to be lower among Blacks than Whites
because, on average, they have fewer resources for identity maintenance, and they
experience more discrimination (Byrd, 2012). But verification should be at least as
high in many contexts, such as interactions with relatives and friends. Moreover,
Whites may get less racial identity verification because they seek it less often, owing
to its presumably lower salience. Given these possibilities, we cannot predict the
direction of any racial differences in identity verification.
Finally, public regard refers to how positively or negatively people believe the
broader society evaluates their social category, in this case, their racial group
(Ashmore et al., 2004; Sellers et al., 1998). It is often measured as how much people
think others respect their racial group. Those who identify with their racial group
recognize its status in society (Monk, 2020). Cultural stereotypes of White superi-
ority and Black inferiority persist, and they help maintain the racial hierarchy
(Williams & Mohammed, 2013). Blacks are aware that their group has a devalued
status in U.S. society (Sellers et al., 1998). Whites recognize their status as the
dominant group (Jardina, 2019). Whites’ sense of their dominant group position
(Blumer, 1958) leads them to expect others to positively value their group.
In studies of public regard among college students (Johnson et al., 2005; Rowley
et al., 1998) and adults (Stets & Fares, 2019), African Americans had lower scores
on public regard than Whites. Theory and research lead us to expect that:
Hypothesis 1d: Public regard will be lower among Blacks than Whites.
Social Relationships and Racial Identity
Identity theory explains how role identities come to be important and brought into
interactions (Stets & Serpe, 2013). It recognizes that larger social structures, such as
gender, age, and race/ethnicity, affect the likelihood of having particular social
relationships, that is, “ties to others in social networks,” as well as the strength of
those ties (Stryker et al., 2005, p. 94). For example, being older is related to a lower
likelihood of being employed and to interacting less often with coworkers. In turn,
the theory would predict, their identities as workers would be less important and
salient.
As scholars have sought to integrate identity theory with social identity theory,
they have recognized that social identities, too, develop and are enacted through
interacting with others in the context of one’s social relationships (Deaux & Martin,
2003; Stets & Burke, 2000). We posit two mechanisms by which social relationships
may strengthen social identities, based on previous theory and research. First, social
relationships may provide a context that supports social identities. For example,
Hispanic university students whose friendships and participation in Hispanic student
organizations supported their ethnic identity assigned more importance to that
identity (Deaux & Martin, 2003). Second, meeting the expectations of one’s role
identities may indirectly strengthen social identities. For example, parents socialize
their children to function in a gendered world. Perhaps as a result, parents rate their
gender identity as more important and salient than nonparents do, and they are
prouder of it (Kiecolt et al., 2019).
Here we investigate whether particular social relationships (marriage or cohabi-
tation, parenthood, or religious attendance) are related to higher scores on racial
identity dimensions that capture people’s views of their own identity. Racial group
membership shapes social networks, which typically are racially homogeneous
(Deaux & Martin, 2003). Relationships with family and friends reinforce Blacks’
racial identities (Deaux & Martin, 2003; Demo & Hughes, 1990; Hughes & Demo,
1989), and they should reinforce Whites’ as well. However, social connections are
unlikely to influence public regard in any systematic way, because it entails
appraisals of the public’s view of one’s group.
Being married or cohabiting may boost some dimensions of racial identity, by
supporting the identity while enacting one’s spousal or partner role. Although being
married or cohabiting is not related to ingroup closeness or positive ingroup evalu-
ations for Blacks or Whites (Hughes et al., 2015), it may heighten racial identity
salience and verification. In their everyday lives, partners have “debriefing” conver-
sations with each other (Vangelisti & Banski, 1993), which may touch on racial
issues. African American couples, for example, provide racism-specific support in
response to their spouse’s accounts of discrimination (Smith et al., 2020). Although
for economic and historical reasons, African Americans are less likely than Whites to
be married (Sassler & Lichter, 2020), they may discuss racial issues more frequently.
On balance, partners’ discussions of racial issues suggest that racial identities are
more often invoked and verified for married and cohabiting people. We tentatively
predict that:
Hypothesis 2a: Married and cohabiting people will score higher on racial identity
salience.
Hypothesis 2b: Married and cohabiting people will score higher on racial identity
verification.
We also explore whether being married or cohabiting is related to other aspects of
racial identity, but we have no evidence on which to base predictions.
Parents also may have stronger racial identities than nonparents, because the role
of parent entails racial socialization. Black parents often seek to socialize children to
be proud of their racial group (Demo & Hughes, 1990; Neblett et al., 2009). Doing
so may heighten their own sense of pride, as well as the importance and salience of
their racial identity. White parents, too, may consciously impart messages about
racial identity to their children. Research on deliberate racial socialization among
White families is sparse (Umaña-Taylor & Hill, 2020). One study suggests that it is
limited to White parents who recognize continuing racial inequality (Hagerman,
2014). Nevertheless, racial socialization may still occur inadvertently. We predict
that parents will have stronger racial identities than nonparents. The effects may be
stronger among Black parents, however.
Hypothesis 3a: Parents will score higher than nonparents on racial identity
importance.
Hypothesis 3b: Parents will score higher than nonparents on racial identity
salience.
Hypothesis 3c: Parents will score higher than nonparents on pride in their racial
identity.
Hypothesis 3d: Parents will score higher than nonparents on racial identity
verification.
Religious attendance also may be related to a stronger racial identity, by provid-
ing a context that supports one’s racial identity. In places of worship, Blacks are
exposed to political messages that foster a sense of community with other Blacks
(Calhoun-Brown, 1996). However, previous studies of Blacks have found inconsis-
tent effects of religious participation on racial identity (Demo & Hughes, 1990;
Hughes & Demo, 1989; Maiya et al., 2020). Religious attendance also may foster
stronger racial identities among Whites. Recent work implicates American Chris-
tianity as a cornerstone of White supremacy (Jones, 2020). As most Whites are
Christians, this implies that religious attendance may reinforce Whites’ racial
identity.
Blacks attend religious services more often than Whites do (Kiecolt & Hughes,
2017). Consequently, on average, a positive effect of religious attendance would
heighten Blacks’ identity scores more. We tentatively predict that:
Hypothesis 4a: Religious attendance will be positively related to racial identity
importance.
Hypothesis 4b: Religious attendance will be positively related to racial identity
salience.
Hypothesis 4c: Religious attendance will be positively related to pride in one’s
racial identity.
Hypothesis 4d: Religious attendance will be positively related to racial identity
verification.
We have described how social relationships may reinforce racial identity.
Although these relationships are not directly connected to racial identity, they may
provide contexts or set role expectations which may reinforce racial identity. The
presence or extensiveness of such social ties, however, does not capture the affective
dimension of social relationships. Identity theory proposes that affective ties to
others strengthen identity centrality and salience.
We propose that in general, feeling lonely and disconnected from others may be
related to a weaker racial identity, as it signals that people lack a sense of belonging
(Bruce et al., 2019; Stets & Burke, 2000). Blacks and other racial minorities scored
slightly higher on loneliness than Whites, but the association disappeared in a
multivariate model (Bruce et al., 2019). Hence, we predict that:
Hypothesis 5a: Feeling lonelier will dampen racial identity importance.
Hypothesis 5b: Feeling lonelier will dampen racial identity salience.
Hypothesis 5c: Feeling lonelier will dampen pride in one’s racial identity.
Hypothesis 5d: Feeling lonelier will dampen racial identity verification.
The Life Course: Cohort Membership and Racial Identity
A life course perspective suggests that the historical context also may influence
identities. Some historical contexts are marked by identity-relevant events in society,
often related to social movements. Such events may lead social identities related to
those movements to take on added meaning and emotional significance (Stryker
et al., 2000). Mannheim’s theory of generational identity proposes that during late
adolescence and early adulthood, people are especially impressionable (Mannheim,
1952). The historical events that occur during these “formative years” have height-
ened significance to them because “they are associated with crystallization of both
personal identities and knowledge of events outside the self” (Griffin, 2004, p. 545).
Such events shape the identities and worldviews that people carry with them
throughout life.
The life course perspective suggests that some birth cohorts will have stronger
racial identities than others. The Civil Rights movement, Barack Obama’s presi-
dency, and the Black Lives Matter movement, which involved frequent and intense
public discussions of race in the U.S., may have produced cohort differences in racial
identity. For example, White Southerners who were young adults during the civil
rights era retained especially vivid collective memories of the Civil Rights move-
ment (Griffin, 2004). But given widespread awareness and media coverage of the
Civil Rights movement, we expect pervasive civil rights era cohort effects on racial
identity dimensions.
The millennial cohort, those born since the early 1980s, also came of age in a
period of heightened discussions of racial inequality. Members were in their
formative years when Barack Obama was elected and served as president, and they
were twice as likely to vote for him as older cohorts (Milkman, 2014). White
millennials were prominent in the Occupy Wall Street movement, which supported
racial equality (Milkman, 2014) and likely heightened awareness of racial identity.
The Black Lives Matter movement, which began in 2013 in response to killings of
unarmed black people and the acquittal of Trayvon Martin’s killer (Leazenby &
Polk, 2020; Ugorji, 2017), may especially have heightened racial identity for Black
millennials. In summary, we predict that:
Hypothesis 6a: The civil rights era cohort and the millennial cohort will score
higher than the middle, post-civil rights cohort on racial identity
importance.
Hypothesis 6b: The civil rights era cohort and the millennial cohort will score
salience.
Hypothesis 6c: The civil rights era cohort and the millennial cohort will score
higher than the middle, post-civil rights cohort on pride in their
racial identity.
Hypothesis 6d: The civil rights era cohort and the millennial cohort will score
verification.
Control Variables
In the analyses below, the effects of gender and education are examined. Among
Whites, women rate their racial identity as more important than men (Jardina, 2019).
However, gender is unrelated to closeness to one’s racial group for Blacks or Whites
(Croll, 2007; Hughes et al., 2015; Wong & Cho, 2005), so we cannot predict whether
it will be associated with other racial identity dimensions. More education has been
linked to feeling either less close or closer to other Blacks (Broman et al., 1988;
Wong & Cho, 2005). For Whites, more education is related both to feeling closer to
other Whites and to rating one’s racial identity as less important (Croll, 2007;
Jardina, 2019; Wong & Cho, 2005). Hence, education may be related to lower racial
identity importance, and to lower salience as well. More educated Blacks and Whites
tend to evaluate their ingroup more positively (Hughes & Demo, 1989; Wodtke,
2012).3 Similarly, education may be related to greater pride in one’s racial identity.
In summary, racial identity is multidimensional and variable (Hunt &
Reichelmann, 2019). It is broadly structured by racial inequality, so we expect to
see differences across dimensions by race. In addition, indicators of social relation-
ships in various contexts, as well as cohort membership, should contribute to its
variability across racial groups. We examine these influences in the analyses below.
Methods
Data
Data for this study came from the 2014 General Social Survey (Smith et al., 2015).
The survey was a national multi-stage probability sample of noninstitutionalized
English-speaking adults aged 18 and older in the United States. Data were collected
using face-to-face computer-assisted interviews. The response rate was 69%. The
analytic sample consisted of the 193 non-Hispanic Blacks and 946 non-Hispanic
Whites who were asked about their racial identities. We excluded Hispanics and
non-Hispanic respondents of other races, due to small sample sizes.
Measures
Racial identity. Dependent variables were five single-items indicators of racial

identity. For each one, the [RACE/ETHNIC TERM] used in the prompt was
respondent’s race for Blacks and Whites. Identity importance was assessed by
asking, “How much is being [RACE/ETHNIC TERM] an important part of how
you see yourself?” Identity salience was measured as, “In general, how much do you
find that being [RACE/ETHNIC TERM] influences or guides how you behave?”
Pride in one’s racial group was measured as, “How proud are you to be [RACE/
ETHNIC TERM]? Identity verification was an item that asked, “Think about how
you see yourself as [RACE/ETHNIC TERM]. How much do you think your friends
see you this way?” Public regard was measured as, “How much do you think people
in the U.S. respect [RACE/ETHNIC TERM]?” Responses to all five questions
ranged from 0—“not at all,” to 10—“completely,” with 5 representing “in the
middle.”
3
Initial models also included region of residence, residence in a rural area, employment status,
income, and occupational prestige as background variables. Later models excluded them because
they were unrelated to any racial identity dimension.
Predictor variables. Respondents were asked what race they considered them-
selves and whether they were Spanish, Hispanic, or Latino/Latina. The regression
analyses compared non-Hispanic Blacks to non-Hispanic Whites.
We measured four aspects of social relationships. Being in a marriage or
cohabiting relationship was measured as a dummy variable (1 ¼ married or
cohabiting, 0 ¼ unmarried, noncohabiting). Parents were compared to nonparents
(1 ¼ parent, 0 ¼ nonparent). Frequency of church attendance ranged from 0 ¼ never
to 8 ¼ every day.
Loneliness, feeling disconnected from others, was a three-item scale (alpha ¼ .78)
that averaged scores on three items: “How often do you feel”: that you lack
companionship? Left out? Isolated from others? Responses were 3 ¼ often,
2 ¼ sometimes, 1 ¼ rarely, or 0 ¼ never. Lacking companionship was correlated
r ¼ .53 with feeling left out and r ¼ .50 with feeling isolated from others. Feeling left
out was correlated r ¼ .63 with feeling isolated. All correlations were highly
significant ( p < .001).
Two sets of birth cohorts, a civil rights and older cohort and a millennial cohort,
were compared to a post-civil rights cohort. Following Griffin (2004), the civil rights
cohort consisted of respondents aged 63–85 years, who were aged 17–25 in the civil
rights era of 1954–1970. We combined this cohort with the pre-civil rights cohort of
26 respondents who were aged 86–89 and older in 2014, who also had lived through
the civil rights movement, and who had similar scores on the racial identity vari-
ables. The millennial cohort consisted of respondents aged 18–31 in 2014, who were
born in 1983 or later. These respondents were aged 17–25 in 2008 when President
Obama was elected, as well as younger respondents who came of age during his
presidency. The post-civil rights cohort consisted of respondents who were aged
32–62 in 2014.
Control variables were gender and educational attainment. Gender was a dummy
variable (0 ¼ male, 1 ¼ female). Educational attainment was measured in years,
from 0 ¼ no formal education to 20 ¼ graduate or professional degree.
Data Analysis
Descriptive statistics were calculated for all study variables for Blacks and Whites,
and the differences between the groups were tested for significance. We computed
correlations among all the study variables. We then computed correlations among
the identity dimensions separately by race. Next, we regressed the five indicators of
racial identity on race and the other predictor variables. To do so, we used structural
equation modeling (SEM) with the maximum likelihood missing value option,
which uses all available information. As a preliminary step, we performed a
multigroup SEM, which tested for differences in effects between Blacks and Whites.
However, as the multigroup analysis did not allow for weighting the data, we
modeled racial differences in effects as interaction terms in the SEM we report.
The data were weighted to adjust for differential probability of inclusion and
nonresponse (Smith et al., 2015).
Results
Table 4.1 shows the descriptive statistics for the variables in the study. Racial
identity was more important to Blacks than Whites (Blacks X ¼ 7.99 vs. Whites
X ¼ 5.22, t ¼ 10.08, p < .001). Similarly, racial identity was more salient to Blacks
than to Whites (Blacks X ¼ 5.96 vs. Whites X ¼ 4.99, t ¼ 2.98, p < .01). Blacks
expressed more pride in their group than Whites did (Blacks X ¼ 9.43 vs. Whites
Table 4.1 Descriptive statistics and unweighted Ns for study variables, by race
Blacks Whites
M (S.E.) or M (S.E.) or
Range proportion N proportion N
Racial identity
Identity importance 0–10 7.99*** 193 5.22*** 932
(0.24) (0.13)
Identity salience 0–10 5.96** 193 4.99** 922
(0.30) (0.12)
Private regard 0–10 9.43*** 192 6.27*** 930
(0.17) (0.13)
Identity verification 0–10 7.62*** 193 6.31*** 900
(0.29) (0.13)
Public regard 0–10 5.00*** 193 6.45*** 905
(0.19) (0.07)
Predictor variables
Married/cohabiting 0, 1 0.45*** 193 0.69*** 946
Parent 0, 1 0.82 193 0.76 946
Religious attendance 0–8 4.34*** 192 3.15*** 945
(0.25) (0.10)
Loneliness 0–3 0.92 187 0.91 942
(0.07) (0.03)
Civil rights & older 0, 1 0.16*** 191 0.29*** 938
cohorts
Post-civil rights cohort 0.69 0.59
Millennial cohort 0.15 0.13
Woman 0, 1 0.63* 193 0.53* 946
Education 0–20 13.36*** 193 14.24*** 945
(0.19) (0.10)
*p < .05
**p < .01
***p < .001
X ¼ 6.27, t ¼ 15.13, p < .001). On racial identity verification, Blacks also stood
out from Whites (Blacks X ¼ 7.62 vs. Whites X ¼ 6.31, t ¼ 4.14, p < .001).
Blacks also perceived less respect for their racial group than Whites did (Blacks
X ¼ 5.00 vs. Whites X ¼ 6.45, t ¼ –7.15, p < .001).
The two groups did not differ on loneliness (Blacks X ¼ 0.92 vs. Whites X ¼ 0.91,
t ¼ 0.06, NS); both scored low, on average. Blacks were less likely than Whites to be
married (Blacks X ¼ 0.45 vs. Whites X ¼ 0.69, t ¼ 6.32, p < .001). Similar
proportions of the two groups were parents (Blacks X ¼ 0.76 vs. Whites X ¼ 0.82,
t ¼ 1.24, NS). Blacks attended religious services more frequently than Whites (Blacks
X ¼ 4.34 vs. Whites X ¼ 3.15, t ¼ 4.41, p < .001).
Fewer Blacks than Whites were in the civil rights and older cohorts, 16% as
compared with 29%. However, roughly equal percentages of Blacks and Whites
were millennials, 15% and 13%, respectively. On average, Blacks had less education
than Whites (Blacks X ¼ 14.24 vs. Whites X ¼ 13.36, t ¼ –4.09, p < .001).
Table 4.2 shows the correlations among the study variables for the total sample.
All the racial identity dimensions except public regard had moderately strong
correlations with each other. Correlations among all the racial identity dimensions
except public regard were moderately strong, from .50 to .68. Correlations for public
regard, in contrast, were in the .2 range. The correlations of Black with the study
variables echo the racial differences shown in Table 4.1. Loneliness was negatively
related to marriage/cohabitation, parenthood, and religious attendance.
Table 4.3 shows the correlations among the racial identity dimensions for Blacks
and Whites. Correlations among the racial identity dimensions were moderate to
high for Whites and more modest for Blacks. For both groups, identity importance,
identity salience, pride, and identity verification were all significantly correlated with
each other. Correlations for Whites were in the .5 to .6 range; those for Blacks ranged
from .19 to .49. Public regard, in contrast, was conceptually distinct from the other
racial identity dimensions. Correlations between it and the other racial identity
dimensions were far smaller. For Blacks, public regard was modestly related to
racial identity importance, but not to any other identity dimensions. By contrast,
public regard was significantly related to all four of the other racial identity dimen-
sions for Whites.
Table 4.4 shows the structural equation models of the five racial identity dimen-
sions. The differences on racial identity between Blacks and Whites remained when
other predictors are added. That is, consistent with Hypotheses 1a and 1b, respec-
tively, racial identity was more important (β ¼ 0.37, p < .001) and salient (β ¼ 0.20,
p < .01) to Blacks than Whites. As Hypothesis 1c predicted, Blacks felt prouder of
their racial identity than Whites did (β ¼ 0.38, p < .001). We made no prediction as
to racial identity verification but found that Blacks’ racial identity was verified more
often than that of Whites (β ¼ 0.16, p < .001). In contrast, as Hypothesis 1d
predicted, Blacks scored lower than Whites on public regard, that is, perceived
respect for their racial ingroup (β ¼ –0.27, p < .001).
74
Table 4.2 Correlations among variables (Ns range from 1069 to 1139)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
(1) Importance 1.00
(2) Salience .63** 1.00
(3) Pride .68** .50** 1.00
(4) Verification .56** .60** .50** 1.00
(5) Public regard .21** .25** .21** .27** 1.00
(6) Black .30** .11** .35** .15** –.27** 1.00
(7) Loneliness .07* .06 .07* .09** .00 .00 1.00
(8) Married/cohabiting –.03 .04 –.04 –.01 .07* –.19** –.24** 1.00
(9) Parent .12** .09** .19** .09** .03 .05 –.08** .19** 1.00
(10) Religious attendance .09** .03 .15** .03 .02 .16** –.08** .01 .16** 1.00
(11) Pre- or civil rights cohort .13** .14 .06 .12** .06 –.10** .02 –.05 .14** .09** 1.00
(12) Post-civil rights cohort –.08** –.09** –.01 –.08* –.05 .08** –.01 .18** .13** .01 –.70** 1.00
(13) Millennial cohort –.08** –.07* –.08** –.06 –.03 .03 .01 –.16** –.39** –.13** –.22** –.47** 1.00
(14) Woman .07* .04 .08** .11** .07* .08* .06 –.09** .12** .14** .01 –.02 .02 1.00
(15) Education –.13** –.10** –.28** –.11** .05 –.12** –.07* .12** –.15** .04 –.00 .01 .01 .01 1.00
*p < .05
**p < .01
K. J. Kiecolt et al.
Table 4.3 Correlations among racial identity variables, by race (Ns range from 192 to 193 for
Blacks and 900 to 923 for Whites)
Identity Identity Pride in Identity
importance salience identity verification
Blacks
Identity salience 0.42***
Pride in identity 0.49*** 0.19**
Identity 0.29*** 0.35*** 0.31***
verification
Public regard 0.15* 0.10 0.14 0.13
Whites
Identity salience 0.67***
Pride in identity 0.66*** 0.54***
Identity 0.59*** 0.65*** 0.51***
verification
Public regard 0.36*** 0.34*** 0.37*** 0.36***
Correlations in bold differ significantly by race at *p < .05
*p < .05
**p < .01
***p < .001
Marital/cohabiting status and parenthood had only scattered effects on racial

identity. As Hypothesis 2a predicted, racial identity was more salient among married
or cohabiting respondents than others. However, contrary to Hypothesis 2b, their
racial identities were not verified more often. Married and cohabiting respondents
also reported greater racial identity importance, but they did not differ from others on
pride in their racial group or perceived respect for their group.
We expected higher levels of racial identity among parents, and perhaps espe-
cially Black parents, who often consciously socialize their children about racial
matters. The only effect that emerged was that parents were prouder of their racial
identity than nonparents, as Hypothesis 3c predicted. Parents did not differ from
nonparents on identity importance (Hypothesis 3a), identity salience (Hypothesis
3b), or identity verification (Hypothesis 3d). None of the effects differed by race.
Religious attendance, too, was only related to being prouder of one’s racial
identity, as Hypothesis 4c predicted. The effect did not differ by race. Contrary to
expectations, religious attendance was unrelated to identity importance (Hypothesis
4a), identity salience (Hypothesis 4b), or identity verification (Hypothesis 4d) for
either Blacks or Whites.
Contrary to Hypothesis 5a, feeling lonelier was related to greater racial identity
importance (β ¼ 0.09, p < .01). Loneliness interacted with race to influence identity
salience. For Blacks, as Hypothesis 5b predicted, loneliness was related to lower
identity salience. For Whites, in contrast, loneliness heightened racial identity
salience. Contrary to Hypothesis 5c, respondents who were lonelier were prouder
of their racial group as well (β ¼ 0.08, p < .05). Surprisingly, they also reported that
Table 4.4 Standardized coefficients for racial identity

Identity Identity Pride in one’s Identity Public
importance salience group verification regard
Blacka 0.37*** 0.20** 0.38*** 0.16*** –
0.27***
Married/cohabitinga 0.07* 0.13*** 0.06 0.06 0.03
Parenta 0.04 0.02 0.10** 0.04 0.02
Religious attendance 0.03 0.00 0.09** –0.02 0.04
Loneliness 0.09** 0.13*** 0.08* 0.09** 0.01
Black Lonelinesss –0.16**
Civil rights/older 0.18*** 0.17*** 0.09* 0.14*** 0.04
cohortsa
Millennial cohorta –0.05 –0.04 –0.04 –0.01 0.01
Black Civil rights/ –0.03 0.03 –0.04
older cohorts
Black Millennial 0.08* 0.10* 0.07*
cohort
Womana 0.06 0.03 0.05 0.10** 0.09*
Black Woman –0.09* –0.09*
Education –0.08** –0.08* –0.23*** –0.09** 0.01
Constant 1.46*** 1.41*** 2.49*** 1.84*** 2.80***
N 1114 1103 1110 1081 1086
a
Reference categories are Whites, unmarried/noncohabiting respondents, nonparents, the post-civil
rights cohort, and men
*p < .05
**p < .01
***p < .001 (two-tailed tests)
their racial identity was more often verified (β ¼ 0.09, p < .01), as against
Hypothesis 5d. Loneliness was unrelated to public regard for one’s racial group.
We predicted that racial identity would be more pronounced among millennials
and the civil rights and older cohort than the post-civil rights cohort. We found some
of the predicted effects of cohort membership on racial identity. For the civil rights
and older cohort, racial identity was more important (Hypothesis 6a) and salient
(Hypothesis 6b) than for the post-civil rights cohort, and their identity was more
often verified (Hypothesis 6d).4 These effects did not differ by race.
Some aspects of racial identity also were heightened among Black millennials.
Black millennials reported higher racial identity importance and salience than the
post-civil rights cohort (but not the civil rights and older cohorts). They also felt
greater pride in their group than the post-civil rights cohort. In contrast, Whites in the
millennial cohort did not differ from Whites in the post-civil rights cohort on any
dimension of racial identity.
4
We also performed supplementary analyses to test whether southern residence at age 16 interacted
with cohort or race to influence racial identity. None of the interactions with cohort or race were
significant. Its only main effect was to strengthen racial identity verification.
On the background variables, gender interacted with race on identity importance

and pride. Black women assigned less importance to their racial identity than Black
men did, and they were somewhat less proud of it than Black men were. In addition,
as a group, women had their racial identity verified more than men did, and they
perceived greater public regard for their racial group. More educated respondents
scored lower on racial identity importance, salience, pride in their group, and
verification. In contrast, education was unrelated to perceived respect for one’s
group.
Discussion
In this study, we drew on social identity theory, identity theory, and life course
theory to investigate several potential influences on racial identity among a national
sample of Blacks and Whites. We examined five dimensions of racial identity:
importance, salience, pride in one’s racial group, verification, and perceived public
regard for one’s racial group. We predicted several racial differences based on social
identity theory, which the findings supported. Racial identity was more important
and salient to Blacks than Whites.
As predicted, Blacks also had higher levels of pride in their group than Whites.
Nevertheless, Whites’ mean score on pride was still above the scale midpoint. As
social identity theory implies, Blacks and other racial minority groups may be
especially motivated to find grounds for feeling proud of their group because they
recognize its devalued status. Their goal is to raise the status of their group (Ellemers
& Haslam, 2011). Dominant group members feel proud of their group as well, but
they do not have to search for positive meanings. In addition, some Whites’ positive
evaluation of their racial group may be tempered by guilt (Jardina, 2019).
Blacks’ racial identity also was verified by their friends more than was Whites’.
Blacks’ higher mean on identity verification may occur because racial matters are
more a feature of everyday interactions for Blacks than they are for Whites. Whites
most often interact with other Whites, and they regard such interactions as race
neutral (Lewis, 2004).
Blacks assessed public regard, that is, respect for their group, as lower than
Whites. Yet even among Blacks, mean level of public regard reached the scale
midpoint. In previous studies, public regard among Blacks was somewhat lower
(Johnson et al., 2005; Sellers et al., 1998), probably owing to differences in the
measures and samples. That Blacks had relatively low public regard but very high
pride in their group suggests that an as yet unknown cultural mechanism may
undermine any negative effects of public regard on positive evaluations of their
racial group.
Consistent with this interpretation, public regard was largely uncorrelated with
other identity dimensions for Blacks (see also Sellers et al., 1998), indicating that
Blacks’ racial identities are largely independent of perceived public respect for their
group. Whatever mechanism is at work may help explain the racial paradox in
self-esteem and mental health—Blacks’ high self-esteem and good mental health,
despite greater exposure to stressors (Williams et al., 2010).
Correlations between public regard and other identity dimensions were moder-
ately high for Whites. That is, Whites’ own views of their racial group seemed most
susceptible to reflected appraisals from the broader society. Alternatively, Whites
with stronger and more positive racial identities may simply assume that the public
shares their beliefs.
Identity theory assumes that people enact social identities in concert with their
social roles, in networks of social relationships. We proposed that marriage or
cohabitation, parenthood, and religious attendance would enhance racial identity.
Parents, for example, may engage in racial socialization (Hagerman, 2014; Maiya
et al., 2020). Religious attendance may affirm one’s sense of belonging to a racial
group, and with it one’s racial identity. However, the social relationships had few
effects on the racial identity dimensions. We believe that the theory is sound, but it
needs to be elaborated. Effects of social relationships on racial identity are likely to
be conditional. For example, parenthood may amplify racial identity for White
parents who teach their children to recognize racial inequality, but not for White
parents who espouse a colorblind racial ideology (Hagerman, 2014). The quality of
social relationships also may matter, as our findings on loneliness indicate.
Our most surprising finding was that loneliness, feeling unconnected to people,
bred a stronger racial identity. Among both Blacks and Whites, loneliness was
related to greater racial identity importance, pride in their group, and verification.
More social ties usually buttress social identities (Deaux & Martin, 2003; Smith,
2007; Stryker et al., 2005), though most evidence on racial identity comes from
studies of racial minority groups (Deaux & Martin, 2003; Demo & Hughes, 1990).
Perhaps lonelier people grow more attached to their racial group to compensate for a
lack of satisfying social relationships. In one study, for example, a stronger racial
identity among Whites was related to lower general trust (Stets and Fares Stets &
Fares, 2019). That lonelier people reported greater verification for their racial
identity seemed especially counterintuitive. Perhaps it was because of selective
perception or a social desirability bias.
Drawing on life course theory, we also examined whether the historical context
influenced racial identity. Living through the Civil Rights movement as an adult had
long-term effects on racial identity. Racial identity was especially important for older
Blacks and Whites, who came of age during an era marked by racial tensions and
sharply demarcated racial boundaries. For both Blacks and Whites in the civil rights
and older cohorts, racial identity had greater salience and was more consonant with
how they thought others viewed them. Griffin (2004), too, found that the events of
the Civil Rights movement shaped Whites’ orientations. Unlike Griffin, though, we
did not find that racial identity in this cohort was more distinctive among native
Southerners than non-Southerners (analyses not shown).
Recent events also magnified the racial identities of Black millennials. Compared
to those in the post-civil rights cohort, Black millennials had higher identity impor-
tance and salience, and they were prouder of their racial group. White millennials’
racial identity did not differ from that of the older post-civil rights cohort. However,
although White millennials do not emphasize their racial identity more than their
older counterparts, neither do they deemphasize their racial identity, as studies of
colorblindness suggest (Bonilla-Silva, 2017; Hartmann et al., 2009). Future research
needs to continue to monitor cohort differences in racial identity and views of racial
inequality in society, especially given continuing struggles for racial justice and
White backlash (Anderson, 2014; Croll, 2007; DiTomaso, 2013; Hartmann et al.,
2009).
This study has several limitations. First, the data were cross-sectional, so we
could not determine the direction of causality. For example, based on identity theory
we proposed that perceived loneliness influences racial identity. As we could not
model reciprocal relationships among the variables, we could not rule out the
possibility that racial identity affects loneliness. It seems unlikely, though, as
identities usually foster ties with communities of others, either real or symbolic
(McLeod, 2015). Second, the data had only single-item measures of racial identity
dimensions. Multi-item scales would be more reliable measures. Third, the data
lacked measures of affective and extensive commitment to role identities that would
have enabled us to better test predictions from identity theory. Fourth, we could not
disentangle cohort differences in racial identity from age differences. We interpreted
differences in racial identity by age as cohort differences, but in addition racial
identity may be more significant at some points in the life course than others. Finally,
having larger samples of Blacks would have enabled us to better detect significant
effects and differences between the groups.
Our study has documented the nature and some of the influences on racial identity
among Blacks and Whites. The findings extend arguments that racial identity differs
in meaning and significance across racial groups. Some of the determinants of racial
identity also differ, owing to both historical and contemporary factors. Scholars need
to continue to investigate the consequences of racial identity for attitudes and
behavior (e.g., Hunt & Reichelmann, 2019; Jardina, 2019). Racial differences in
racial and political attitudes, for example, may be especially sharp among group
members with more distinctive racial identities.
Finally, we need to know much more about the characteristics of contexts and
social relationships that influence the racial identity dimensions. Identity theory has
much to contribute here (Deaux & Martin, 2003; Stryker et al., 2005). How the
characteristics of contexts influence the nature of social relationships within those
contexts, how social relationships influence racial identity, how identity dimensions
influence well-being, and how these dynamics may differ by race are important
questions for future research. For example, given persisting racial inequality (Wil-
liams & Mohammed, 2013), contexts and social relationships that make racial
identity less salient for racial minorities may enhance well-being.
References
Anderson, C. (2014). Ferguson isn’t about black rage against cops. It’s white rage against progress.
Washington Post, August 29. https://www.washingtonpost.com/opinions/ferguson-wasnt-
black-rage-against-copsit-was-white-rage-against-progress/2014/08/29/3055e3f4-2d75-11e4-
bb9b-997ae96fad33_story.html
Ashmore, R. D., Deaux, K., & McLaughlin-Volpe, T. (2004). An organizing framework for
collective identity: articulation and significance of multidimensionality. Psychological Bulletin,
130(1), 80–114.
Blumer, H. (1958). Race prejudice as a sense of group position. Pacific Sociological Review, 1(1),
3–7.
Bobo, L. D., & Johnson, D. (2000). Racial attitudes in a prismatic metropolis: mapping identity,
stereotypes, competition, and views on affirmative action. In L. D. Bobo, M. L. Oliver, J. M.
Johnson Jr., & A. Valenzuela Jr. (Eds.), Prismatic metropolis: inequality in Los Angeles
(pp. 81–161). Russell Sage.
Bonilla-Silva, E. (2017). Racism without racists: color-blind racism and the persistence of racial
inequality in America (5th ed.). Rowman and Littlefield.
Branscombe, N. R., Schmitt, M. T., & Harvey, R. D. (1999). Perceiving pervasive discrimination
among African Americans: Implications for group identification and well-being. Journal of
Brenner, P. S., Serpe, R. T., & Stryker, S. (2014). The causal ordering of prominence and salience in
identity theory: An empirical examination. Social Psychology Quarterly, 77(3), 231–252.
Broman, C. L., Neighbors, H. W., & Jackson, J. S. (1988). Racial group identification among Black
adults. Social Forces, 67(1), 146–158.
Brown, T. N., Sellers, S. L., & Gomez, J. P. (2002). The relationship between internalization and
self-esteem among Black adults. Sociological Focus, 35(1), 55–71.
Bruce, L. D. H., Wu, J. S., Lustig, S. L., Russell, D. W., & Nemecek, D. A. (2019). Loneliness in the
United States: A 2018 national panel survey of demographic, structural, cognitive, and behav-
ioral characteristics. American Journal of Health Promotion, 33(8), 1123–1133.
Byrd, D. R. (2012). Race/ethnicity and self-reported levels of discrimination and psychological
distress, California, 2005. Preventing Chronic Disease, 9, 120042. Retrieved October 2020,
from https://doi.org/10.5888/pcd9.120042
Calhoun-Brown, A. (1996). African American churches and political mobilization: the psycholog-
ical impact of organizational resources. Journal of Politics, 58(4), 935–953.
Campbell, R. D. (2017). “We Pride Ourselves on Being Strong...And Able to Bear a Lot”:
Examining the socio-cultural and historical context of Black Americans’ experiences with
depression and help-seeking. Advances in Social Work, 18(2), 663–681.
Crocker, J., & Major, B. (1989). Social stigma and self-esteem: the self-protective properties of
stigma. Psychological Review, 96(4), 608–630.
Croll, P. R. (2007). Modeling determinants of white racial identity: results from a new national
survey. Social Forces, 86(2), 613–642.
Cross, W. E., Jr. (1991). Shades of black: diversity in African-American identity. Temple University
Press.
Deaux, K., & Martin, D. (2003). Interpersonal networks and social categories: specifying levels of
context in identity processes. Social Psychology Quarterly, 66(2), 101–117.
Demo, D. H., & Hughes, M. (1990). Socialization and racial identity among Black Americans.
Social Psychology Quarterly, 53(4), 364–374.
DiTomaso, N. (2013). The American non-dilemma: racial inequality without racism. Russell Sage
Foundation.
Ellemers, N., & Haslam, S. A. (2011). Social identity theory. In P. A. M. Van Lange, A. W.
Kruglanski, & E. T. Higgins (Eds.), Handbook of theories of social psychology (pp. 379–398).
Sage.
Feagin, J. R. (1991). The continuing significance of race: antiblack discrimination in public places.
American Sociological Review, 56(1), 101–116.
Griffin, L. J. (2004). “Generations and collective memory” revisited: race, region, and memory of
civil rights. American Sociological Review, 69(4), 544–557.
Hagerman, M. A. (2014). White families and race: colour-blind and colour-conscious approaches to
white racial socialization. Ethnic and Racial Studies, 37(14), 2598–2614.
Harris, C. I. (1993). Whiteness as property. Harvard Law Review, 106(8), 1707–1791.
Hartmann, D., Gerteis, J., & Croll, P. R. (2009). An empirical assessment of whiteness theory:
Hidden from how many? Social Problems, 56(3), 403–424.
Hughes, M., & Demo, D. H. (1989). Self-perceptions of Black Americans: self-esteem and personal
efficacy. American Journal of Sociology, 95(1), 132–159.
Hughes, M., Kiecolt, K. J., Keith, V. M., & Demo, D. H. (2015). Racial identity and well-being
among African Americans. Social Psychology Quarterly, 78(1), 25–48.
Hughey, M. W. (2012). Color capital, white debt, and the paradox of strong white racial identities.
Du Bois Review: Social Science Research on Race, 9(1), 169–200.
Hughey, M. W., & Byrd, W. C. (2013). The souls of white folk beyond formation and structure:
bound to identity. Ethnic and Racial Studies, 36(6), 974–981.
Hunt, M. O., & Reichelmann, A. V. (2019). Racial identity and racial attitudes among White
Americans. In J. E. Stets & R. T. Serpe (Eds.), Identities in everyday life (pp. 217–237). Oxford
University Press.
Ida, A. K., & Christie-Mizell, C. A. (2012). Racial group identity, psychosocial resources, and
depressive symptoms: exploring ethnic heterogeneity among Black Americans. Sociological
Focus, 45(1), 1–23.
Jardina, A. (2019). White identity politics. Cambridge University Press.
Johnson, T. M., Robinson Kurpius, S. E., Rayle, A. D., Arredondo, P., & Tovar-Gamero, Z. G.
(2005). The Multidimensional Inventory of Black Identity: its use with Euro-American, Latino,
and Native American undergraduates. Measurement and Evaluation in Counseling and Devel-
opment, 38(2), 92–103.
Jones, R. P. (2020). White too long: the legacy of white supremacy in American christianity. Simon
& Schuster.
Kiecolt, K. J., Momplaisir, H., & Hughes, M. (2016). Racial identity, racial discrimination, and
depressive symptoms among African Americans and Afro-Caribbeans. In J. E. Stets & R. T.
Serpe (Eds.), New directions in identity theory and research (pp. 367–393). Oxford University
Press.
Kiecolt, K. J., & Hughes, M. (2017). Racial identity and the quality of life among blacks and whites
in the U.S. Social Science Research, 67, 59–71.
Kiecolt, K. J., Hughes, M., & Momplaisir, H. (2019). Gender identity among U.S. adults. In J. E.
Stets & R. T. Serpe (Eds.), Identities in everyday life (pp. 195–215). Oxford University Press.
Leazenby, L., & Polk, M. (2020). What you need to know about black lives matter in 10 questions.
Chicago Tribune, September 3. https://www.chicagotribune.com/lifestyles/ct-life-cb-black-
lives-matter-chicago-20200903-xh75kbw5nfdk5joudlsgb2viwq-story.html
Leonardelli, G. J., Pickett, C. L., & Brewer, M. B. (2010). Optimal distinctiveness theory: a
framework for social identity, social cognition, and intergroup relations. Advances in Experi-
mental Social Psychology, 43, 63–113.
Lewis, A. E. (2004). ‘What group?’ Studying whites and whiteness in the era of ‘color-blindness’.
Sociological Theory, 22(4), 623–646.
Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: self-evaluation of one’s social
identity. Personality and Social Psychology Bulletin, 18(3), 302–318.
Maiya, S., Carlo, G., Landor, A. M., & Memmott-Elison, M. K. (2020). Ethnic-racial and religious
identity as mediators of relations between ethnic-racial socialization and prosocial behaviors
among Black young adults. Journal of Black Psychology. https://doi.org/10.1177/
0095798420971388
Mannheim, K. (1952). The problem of generations. In K. Mannheim (Ed.), Essays in the sociology
of knowledge (pp. 276–322). Routledge and Kegan Paul.
McCall, G. J., & Simmons, J. L. (1978). Identities and interactions: an examination of human
associations in everyday life. Free Press.
McDermott, M., & Samson, F. L. (2005). White racial and ethnic identity in the United States.
Annual Review of Sociology, 31, 245–261.
McLeod, J. D. (2015). Why and how inequality matters. Journal of Health and Social Behavior, 56
(2), 149–165.
Milkman, R. (2014). Millennial movements: Occupy Wall Street and the DREAMers. Dissent, 61
(3), 55–59.
Monk, E. P., Jr. (2020). Linked fate and mental health among African Americans. Social Science &
Medicine, 266, 113340. Retrieved September 3, 2020, from https://doi.org/10.1016/j.
socscimed.2020.113340
Neblett, E. W., Smalls, C. P., Ford, K. R., Nguyen, H. X., & Sellers, R. M. (2009). Racial
socialization and racial identity: African American parents’ messages about race as precursors
to identity. Journal of Youth and Adolescence, 38(2), 189–203.
Omi, M., & Winant, H. (2014). Racial formation in the United States (3rd ed.). Routledge.
Oyserman, D., & Swim, J. K. (2001). Stigma: an insider’s view. Journal of Social Issues, 57(1),
1–14.
Phinney, J. S. (1991). Ethnic identity and self-esteem: a review and integration. Hispanic Journal of
Behavioral Sciences, 13(2), 193–208.
Roberts, R. E., Phinney, J. S., Masse, L. C., Chen, Y. R., Roberts, C. R., & Romero, A. (1999). The
structure of ethnic identity of young adolescents from diverse ethnocultural groups. Journal of
Early Adolescence, 19(3), 301–322.
Rosenberg, M. (1979). Conceiving the self. Krieger.
Rowley, S. J., Sellers, R. M., Chavous, T. M., & Smith, M. A. (1998). The relationship between
racial identity and self-esteem in African American college and high school students. Journal of
Sassler, S., & Lichter, D. T. (2020). Cohabitation and marriage: complexity and diversity in union-
formation patterns. Journal of Marriage and Family, 82(1), 35–61.
Sellers, R. M., Smith, M. A., Shelton, J. N., Rowley, S. A. J., & Chavous, T. M. (1998).
Multidimensional model of racial identity: a reconceptualization of African American racial
identity. Personality and Social Psychology Review, 2, 18–39.
Smith, S. M., Williamson, L. D., Branch, H., & Fincham, F. D. (2020). Racial discrimination,
racism-specific support, and self-reported health among African American couples. Journal of
Social and Personal Relationships, 37(3), 779–799.
Smith, T. W. (2007). Social identity and socio-demographic structure. International Journal of
Public Opinion Research, 19(3), 380–390.
Smith, T. W., Marsden, P. V., & Hout, M. (2015). General Social Surveys 1972-2014. National
Opinion Research Center at the University of Chicago.
Quarterly, 63(3), 224–237.
Stets, J. E., & Fares, P. (2019). The effects of race/ethnicity and racial/ethnic identification on
general trust. Social Science Research, 80, 1–14.
Stets, J. E., & Harrod, M. M. (2004). Verification across multiple identities: the role of status. Social
Psychology Quarterly, 67(2), 155–171.
Stets, J. E., & Serpe, R. T. (2013). Identity theory. In J. DeLamater & A. Ward (Eds.), Handbook of
social psychology (pp. 31–60). Springer.
Stryker, S., Owens, T. J., & White, R. W. (2000). Self, identity, and social movements. University of
Minnesota Press.
Stryker, S., & Serpe, R. T. (1994). Identity salience and psychological centrality: equivalent,
overlapping, or complementary concepts? Social Psychology Quarterly, 57(1), 16–35.
Stryker, S., Serpe, R. T., & Hunt, M. O. (2005). Making good on a promise: the impact of larger
social structures on commitments. Advances in Group Processes, 22, 93–123.
Syed, M., & Juang, L. P. (2014). Ethnic identity, identity coherence, and psychological functioning:
testing basic assumptions of the developmental model. Cultural Diversity & Ethnic Minority
Psychology, 20(2), 176–190.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S. Worchel
& L. W. Austin (Eds.), Psychology of intergroup relations (pp. 2–24). Nelson-Hall.
Thornhill, T. (2015). Racial salience and the consequences of making white people uncomfortable:
intra-racial discrimination, racial screening, and the maintenance of white supremacy. Sociology
Compass, 9(8), 694–703.
Thornton, M. C., Taylor, R. J., & Chatters, L. M. (2012). African American, Black Caribbean, and
non-Hispanic White feelings of closeness toward other racial and ethnic groups. Journal of
Black Studies, 43(7), 749–772.
Thornton, M. C., Taylor, R. J., & Chatters, L. M. (2013). African American and Black Caribbean
mutual feelings of closeness: findings from a national probability survey. Journal of Black
Studies, 44(8), 798–828.
Ugorji, B. (2017). Black lives matter: decrypting encrypted racism. Ethnic Studies Review, 37–38,
27–43.
Umaña-Taylor, A. J., & Hill, N. E. (2020). Ethnic-racial socialization in the family: a decade’s
advance on precursors and outcomes. Journal of Marriage and Family, 82(1), 244–271.
Vandiver, B. J., Cross, W. E., Worrell, F. C., & Fhagen-Smith, P. E. (2002). Validating the Cross
Racial Identity Scale. Journal of Counseling Psychology, 49(1), 71–85.
Vangelisti, A. L., & Banski, M. A. (1993). Couples’ debriefing conversations: the impact of gender,
occupation, and demographic characteristics. Family Relations, 42(2), 149–157.
Waters, M. C. (1990). Ethnic options: choosing identities in America. University of California
Press.
Williams, D. R., Costa, M., & Leavell, J. P. (2010). Race and mental health: patterns and
challenges. In T. L. Scheid & T. N. Brown (Eds.), A handbook for the study of mental health:
social contexts, theories, and systems (pp. 268–290). Cambridge University Press.
Williams, D. R., & Mohammed, S. A. (2013). Racism and health I: pathways and scientific
evidence. American Behavioral Scientist, 57(8), 1152–1173.
Wodtke, G. T. (2012). The impact of education on intergroup attitudes: A multiracial analysis.
Wong, C., & Cho, G. E. (2005). Two-headed coins or Kandinskys: white racial identification.
Political Psychology, 26(5), 699–720.
Chapter 5
White Racial Identity and Reparations
for Slavery
Ashley V. Reichelmann and Matthew O. Hunt
My claim, in simplest terms, is that dominant group identities can become salient and
politically meaningful when the conditions that facilitate their invisibility are disturbed;
that is, when the group believes that its status is sincerely challenged. In short, the salience
and assertion of a dominant group identity is reactive; it is an effort to defend the group’s
position within a stratified system. (Jardina, 2019, p. 36)
Introduction
The African American Museum of History and Culture (AAMHC) opened in

Washington DC in 2016. Less than two years later, the Memorial for Peace and
Justice (MPJ) opened in Montgomery, Alabama—one of the geographic centers of
the Civil Rights movement. These two tourist sites are unique because they are each
the first of their kind: the AAMHC is the first museum dedicated to African
Americans on the National Mall, while the MPJ is billed as the first national
memorial to victims of lynching in the United States. Both are substantial in size
and content, epitomizing a sustained movement towards reparations that would
acknowledge and more completely represent some of America’s darkest moments:
the historical violence against, and oppression of, Africans and African Americans.
At the same time, however, the populist wave that Donald Trump rode to power,
followed by high-profile (and deadly) White nationalist demonstrations such as in
Charlottesville, Virginia during the summer of 2017, invite the question: how do
A. V. Reichelmann (*)
Department of Sociology, Virginia Tech, Blacksburg, VA, USA
M. O. Hunt
Department of Sociology, Northeastern University, Boston, MA, USA

86 A. V. Reichelmann and M. O. Hunt
white Americans feel about America’s legacy of slavery and its contemporary
impact on the lives of black Americans? What are their views on possible reparative
measures, and how are such views influenced by their understandings of themselves
as white Americans? Theoretically, such knowledge helps advance contemporary
understandings of the heterogeneous manifestations of White racial identities (WRI),
previously thought to be largely invisible and homogenous. Empirically, they allow
exploration of the relationship between white Americans’ views of their own
identities (including possible recognition of the lasting effects of a history that
privileged their racial group) and their support for modern redistributive social
policies. This particular relationship is of interest given recent work demonstrating
the relevance of racial identity for understanding whites’ political orientations (see
Jardina, 2019; Major et al., 2018; Mutz, 2018 for examples). Systematic examination
of this relationship also contributes to the larger project of understanding WRI as
part of a complex interplay between self-views, perceptions of outgroups, and
historical interactions between groups (Reichelmann, 2020).
Using data from an original survey fielded in 2016, we explore how white
Americans’ racial identities shape their views of reparations to African Americans
for the enslavement of Africans on U.S. soil. Building off of previous work analyz-
ing connections between WRI dimensions and a range of racial attitudes (Hunt &
Reichelmann, 2019), we use Identity Theory (IT) to examine whether and how such
identity elements relate to support for/opposition to an array of reparative measures
including an apology for slavery, a memorial to victims of slavery, and financial
payments to black Americans, as well as government intervention on behalf of black
Americans in the form of special scholarships and policies involving preferential
hiring, workplace fairness, and the improvement of blacks’ social and economic
standing. Such analysis furthers IT’s goal of demonstrating how identity matters in
the social world—in this case on the controversial topic of reparations to African
American descendants of enslaved persons.
Background
Identity involves the ways individuals conceive of themselves, in interaction with

others, using categories provided by their social worlds (Burke & Stets, 2009;
Stryker, [1980]2002). Identity is often discussed along three bases: social, role and
person identities (Burke & Stets, 2009; Stets & Serpe, 2013). Social identities are
rooted in available socio-cultural categories (e.g., race, gender, region). Role iden-
tities correspond to statuses comprising key social institutions (e.g., parent, student,
worker), and person identities refer to how individuals see themselves as distinct
from other individuals. Racial identity is the way people define themselves vis-à-vis
available socially-constructed racial categories in a given community or society. In
the United States, to be “white” has meant, historically, identifying with the polit-
ically and economically dominant racial group borne of European colonialism and
settlement in the Americas; as such, it has also meant not being a member of any of
5 White Racial Identity and Reparations for Slavery 87
the identifiable racialized minority groups in the U.S. (e.g., African Americans,
Asian-Americans, Native Americans, etc.).
All three identity bases shape how people interpret, interact with, and are
responded to by others in their social milieu. Symbolic interactionism provides a
framework for understanding how symbolic meanings—internalized and activated
by the context of an interaction—situationally locate our identities (Howard, 2000).
As such, identities represent a key part of the self and the self-concept where “self”
refers to the internalized conversation regarding “who am I?” and the “self-concept”
represents the totality of an individual’s reflexive view of him/herself at a moment in
time (Rosenberg, 1979, 1981). These self-meanings are at the crux of understanding
identities, and demonstrate how they are constructed through a process of social
interaction with others and with the structure of society (Stryker, 1980).
A basic tenet of identity theories is that identities “matter” (Rosenberg, 1981)
because they both define and are defined by past and current social experiences—a
notion that is key to understanding racial identity in the context of Identity Theory
(IT) (Stryker, 1980). IT as a framework posits that the internal structure of the self
shapes the thoughts and behavior of the person. Rooted in Mead’s conception of
symbolic interactionism (SI), IT takes as its basic premise that we as individuals are
fundamentally shaped by the social structures we are born into and interact within.
These structures are encompassed by networks within which the self develops and is
refined as we interact via a multiplicity of roles and socio-cultural categories. As a
result, the “self reflects society,” since we have as many self-identities as we have
roles and other social category memberships (Mead, 1934).
The concept of “role” is key to Stryker’s (1980) structural symbolic interactionist
formulation, wherein roles represent the key bridge linking individual and society.
Roles stem from positions in social structure and carry behavioral expectations
(1980, p. 57) that both guide interaction and become the basis of identities for the
incumbent. Shared expectations allow those taking roles to coordinate their behavior
(e.g., teacher and student; parent and child; boss and worker) in patterned and
recurrent ways; thus, what we recognize as social structure is sustained and
reproduced by way of individuals internalizing and meeting the shared expectations
of the roles they play (1980, pp. 54–57). For Stryker and colleagues, the self is the set
of role identities, organized into a hierarchy of salience based on patterns of
commitment the person has to others. The salience hierarchy, in turn, allows for
prediction of “which self component, or. . .identity, will take precedence in account-
ing for some self-relevant outcome” (Stryker & Serpe, 1994, p. 18).
In this paper, salience is one of five identity dimensions we focus on. Identity
dimensions can be thought of as elements that, collectively, comprise the overall
“intensity” or scope an identity holds for the person (Hunt, 2020). In this sense,
identities represent multi-dimensional constructs, including salience, prominence,
verification, private self-regard, and public self-regard. Recent work has demon-
strated the empirical utility of a single (multi-item) measure of WRI intensity
(comprising the five dimensions analyzed herein) (Hunt, 2020; Stets & Fares,
2019). However, Hunt (2020) emphasizes that studies analyzing the separate impact
of these identity dimensions remain critically important and rest on strong theoretical
and empirical foundations (see below). Hunt and Reichelmann (2019), for instance,
show that prominence (importance) and public self-regard (pride) are most strongly
associated with whites’ social distance and racial policy stances respectively, while
the three other dimensions showed weaker or no associations with these outcomes.
Identity salience refers to the likelihood of an identity being invoked in everyday
interactions and across situations (Burke & Stets, 2009; Stryker, 1980). Prominence,
a second central identity dimension (see McCall & Simmons, 1978), refers to the
importance a person accords to a given identity (Deaux, 1996). The relationship
between salience and prominence has been the focus of much research, resulting in a
clearer understanding of their distinctiveness (Stryker & Serpe, 1994) and their
causal ordering from prominence to salience (Brenner et al., 2014).
Verification, a third dimension, parallels the logic of reflected appraisals in
symbolic interactionism and refers to the likelihood that others see us through
the same identity categories that we see ourselves. This concept is foundational to
the perceptual control of identities, focusing on internal processes governing the
identity-behavior nexus (Burke & Stets, 2009). Public and private self-regard are the
final two identity dimensions we examine; these constructs have received less
attention within IT, possibly stemming from the inconsistent results found in earlier
work exploring their relationship to one another and to other dimensions of identity
(Gecas, 1982; Sellers et al., 1998). Public self-regard refers to an individual’s
perception of the respect or esteem that others (i.e. non-group members) accord to
their group, while private self-regard refers to a person’s level of pride in an identity/
category membership (Luhtanen & Crocker, 1991).
Much of the past research in IT has focused on roles as the basis of identities
(Stets & Burke, 2000). However, some recent work has sought to apply the IT
framework to identities typically thought of as social (e.g., race, gender) rather than
role-based—particularly since such group-based identities also organize thought and
behavior (Hunt & Reichelmann, 2019; Stets & Fares, 2019).1 Social Identity Theory
(SIT) holds that once a group-based identity becomes part of the self-concept, people
recognize both in- and out-group boundaries (and related distinctions), and are
socialized into the norms of the in-group. Attitudes and preferences are therefore
group-based (Smith & Hogg, 2008), since they become collective when individuals
conform to the normative views of other in-group members. In this way, SIT
resonates with Blumer’s earlier and highly influential work on racial prejudice
stemming from perceived disruptions in the taken-for-granted positional and hierar-
chical arrangement of social groups (1958). Blumer’s (1958) seminal work in this
area focuses on how being a dominant group member in a hierarchical social system
fosters a collective sense/common orientation of group position. When such domi-
nance is challenged, group members defend it with a range of responses including
1
Stets and Burke (2000) allude to this potential overlap between IT and SIT in noting that “role” and
“group,” as concepts, are not easily separable—empirically or analytically—a view echoed by
others calling for greater theoretical synthesis in the interest of a more unified and general view of
the self (see Deaux & Martin, 2003; Hogg et al., 1995; Stryker & Burke, 2000).
racial prejudice. Thus, Blumer’s work demonstrates the role that social/group
identity plays in the construction of prejudice and the reinforcement of the status
quo vis-a-vis views of the self and the in-group.
White Racial Identity and Attitudes
Most social psychological research on racial identity has focused on minority

populations such as African Americans (Phinney, 1996; Rowley et al., 1998; Sellers
et al., 1998) or multiracial individuals (see Khanna, 2010 for thorough discussion).
There has also been substantial interest among social and political psychologists in
white Americans’ out-group attitudes—i.e., how whites’ think about other racial
groups. Relevant work on the latter includes the literatures on traditional or
old-fashioned racism (Kluegel, 1990; Hughes, 1997), racial apathy (Forman &
Lewis, 2006), racial resentment (Feldman & Huddy, 2005; Kinder & Sanders,
1996; Tuch & Hughes, 2011; Wilson & Davis, 2011), racial stratification beliefs
(Hunt, 2007), and laissez-faire/contemporary racism/color-blind racism (Bobo et al.,
1997; Bonilla-Silva, 2014).
In contrast, we know less about white Americans’ in-group attitudes and how
they may impact their attitudes and behaviors. A small pool of literature addresses
how racial socialization affects the development of whites’ racial identity dimen-
sions, though this literature has limitations. Identity is often captured in a cursory
way, explored via a single dimension, such as salience or prominence (Sanders
Thompson, 1999; Knowles & Peng, 2005; Croll, 2007), or through a perceived
group closeness measure (Wong & Cho, 2005). In addition, until recently, theorists
tended to assume that WRI was not salient and had few impacts on persons’
interactions with the social world (Doane, 1997; Rowe et al., 1994). Along these
lines, Wong and Cho (2005) argued for the need to develop a baseline understanding
of WRI before it “seems politically relevant,” in order to understand how it may
change as the demographics of the United States shift (p. 701).
Implicit in these views is the assumption that, because of whites’ historical
dominance in the United States (numerically, politically, and economically), they
do not think of themselves in racial terms (in contrast to how they tend to view
others) or as having clear group-based interests to defend. As this line of thought
goes, whiteness as a category is largely invisible and represents the reference point,
racially-speaking, from which others are compared and racialized. Recent work casts
doubt on the accuracy (or at least permanence) of this viewpoint, showing that
substantial proportions of the white population view their racial identity as important
(Hartmann et al., 2009; Jardina, 2019).
Research also suggests that WRI has been in flux in recent decades owing to
ongoing societal shifts, demonstrating that it can and does serve as a driving factor in
many behavioral, social, and political outcomes (Jardina, 2019; Knowles et al.,
2014). For instance, Jardina (2019) shows convincingly that WRI became increas-
ingly “activated” in recent years (in response to broad forces such as globalization,
demographic change, immigration, and the Obama presidency) and now powerfully
shapes presidential voting and support for social policies seen as benefitting whites’
group-based self-interest (e.g., Social Security; Medicare). Despite these develop-
ments, our knowledge of the nature, strength, and impact of WRI on behavioral and
attitudinal outcomes remains underdeveloped.
Some existing research on WRI has explored its relationship to the production of
intergroup attitudes. This work includes ethnographies exploring how whiteness is
an important part of self-identification for both self-avowed racists (Simi & Futrell,
2010) and anti-racists, though the net result in both instances may be the reinforce-
ment of white hegemony (Hughey, 2012). In another study, Goren and Plaut (2012)
used essay tasks as measures of racial identity, categorizing identities into three
forms—weakly identified, prideful, and power cognizant—finding that prideful and
power-cognizant individuals both have strong WRIs. And, Knowles et al. (2014)
advance a “3D” model of WRI management, stemming from three key strategies
whites utilize to neutralize psychological threats inherent in a privileged group
membership; they can either deny such privilege, distance themselves from white-
ness, or seek to dismantle systems from which they derive benefits. The “deny” and
“distance” strategies are conservative in their implications (fostering insensitivity
and inaction vis-à-vis racial inequality), while the “dismantle” stance seeks to reduce
threat by changing the circumstances from which whites benefit and improve their
group image in the process.
Several quantitative studies have produced inconclusive results regarding the role
of WRI dimensions in shaping intergroup attitudes; some find that these dimensions
have no effect (Sullivan & Ghara, 2015), particularly when examining public and
private self-regard. Others have shown that whites’ private self-regard (sense of
racial pride) is a key determinant of their opposition to race-targeted policies, while
whites’ identity prominence (importance) significantly predicts greater preferred
social distance from African Americans (Hunt & Reichelmann, 2019). Still other
work shows that the effect of WRI on prejudice is maximized when whites are faced
with a realistic loss of status or resources (Lowery et al., 2006; Bobo & Hutchings,
1996). Finally, Croll (2007) observed that WRI prominence has both defensive and
progressive manifestations across outcomes such as racial prejudice and attitudes
toward diversity—findings that align with Knowles’ “deny” and “distance” (con-
servative) versus “dismantle” (progressive) strategies respectively. However, Croll’s
use of a single item measure (prominence) provided only limited understanding of
WRI’s impact on racial attitudes.
The current study builds on Croll’s and others’ recent work by examining
multiple dimensions of WRI (Hunt & Reichelmann, 2019; Hunt, 2020) in the
interest of developing a richer understanding of its defensive and progressive
manifestations. This work advances IT by (1) furthering our understanding of the
relationship between particular dimensions of WRI and important socio-political
attitudes, and (2) demonstrating how IT can advance our knowledge regarding the
role and operation of social identities above and beyond what SIT has demonstrated.
Research on Reparations
Most research and scholarship on reparations lies beyond the social sciences, taking
a philosophical (Darby & Branscombe, 2014; McCarthy, 2004), legal (Cooper,
2012), or purely theoretical lens on questions of reparations for past harm (Lickel
et al., 2004; Balfour, 2014, 2015; McCarthy, 2002). Existing social science contri-
butions on the topic include non-US-based studies of contemporary attitudes toward
reparations, including in post-colonial or recent post-conflict societies such as Chile,
South Africa, and Martinique (Brown et al., 2008; Klandermans et al., 2008;
Armange & Mullet, 2016). The assumption that most white Americans will oppose
reparations is rooted in research on whites’ opposition to other race-targeted policies
including affirmative action (Bobo, 2000; Hughes, 1997; Tuch, 1987), busing for the
purpose of racial integration in public schools (Bobo, 1983), and government aid to
improve the living standards of black Americans (Bobo & Kluegel, 1997). Further,
in a racially-comparative study of support for reparations, Dawson and Popoff
(2004) found that whites were far less likely than blacks to support reparations
(e.g., an apology; financial payments) for both African Americans and Asian
Americans.
Empirical studies focusing on white Americans’ support for reparations demon-
strate effects of question-ordering and of how responsibility for reparations is
presented/framed. Regarding the former, white Americans were shown to be less
sympathetic to a federally-issued apology to African Americans after questions
about reparations to Japanese Americans for their internment during WWII (Dawson
& Popoff, 2004). Regarding framing, whites are more likely to support reparations if
(1) corporations or heirs of slaveholders are listed as the liable providers (versus
whites generally or the government), and (2) education is the form of reparation
versus cash payments (Craemer, 2009a). Relatedly, whites Americans are more
supportive of reparations framed as retributive rather than restorative (Glaser &
Ryan, 2011). Research also demonstrates that white Americans’ status characteris-
tics and political orientations increase support for reparations. These include being
female, supporting non-traditional political candidates, and recognizing the voter
suppression of blacks and an unfair economic system (Dawson & Popoff, 2004). In
addition, stronger feelings of closeness to blacks (Craemer, 2009a, 2009b) and
perception of greater “cost of being black” in the US (Mazzocco et al., 2006) also
increase support for reparations.
While the aforementioned studies provide important insights into white Ameri-
cans’ support for reparations, most of this research casts whites as a homogenous
group, failing to explore sources of variation within this population. In the current
study, we depart from past practice by focusing on how different dimensions of WRI
are associated with support for/opposition to a range of possible forms of reparations
for slavery. As such, our research promises to add important new insights to the
literature on white identity and racial attitudes. In addition, our work advances IT by
demonstrating that “society shapes self shapes social behavior” through a defensive
manifestation: white Americans defend their group’s dominant position in the
stratification order vis-à-vis behavior, attitudes, and policy support/opposition that

reinforce the hierarchical racial structure providing the foundation of their dominant-
group identity (Jardina, 2019). In this way, our work further encourages IT to
interrogate the independent and unique impacts of identity dimensions on the
relationship between WRI and social behavior.
Data & Methods

Sample
The data we utilize in the current study derive from a Survey Sampling International
(SSI)2 panel of non-Hispanic white American respondents, aged 18 years or older.
Our analytic sample consists of 866 respondents who responded to all variables of
interest when completing an online survey fielded in July 2016. The data collection
instrument included a range of measures capturing collective emotions and socio-
political attitudes. These items were embedded in a larger study that utilized
experimental conditions to understand if and how representations of slavery affect
white Americans’ socio-emotional responses. After beginning the survey, respon-
dents were asked to complete five racial identity dimension items (described below)
and were then randomly assigned to one of four condition groups—each exposed to
a different representation of slavery—or a control group which was not exposed to a
representation. The conditions were historically accurate representations of slavery
in the U.S., and varied in two ways: written or visual, and perpetrator-focused or
victim-focused. The four conditions were: written perpetrator-focused, visual
perpetrator-focused, written victim-focused, or visual victim-focused.3
2
SSI develops panels of samples that are tailored to capture the respondents who fit the researcher’s
parameters and they also try to match the sample as closely as possible to the age, gender, and
geographical makeup of the American adult population. Therefore, the sampling method was
proportionally selected. Recent research has found evidence that probability and non-probability
sampling yield similar results on racial attitudes when using particular internet surveying platforms
(Simmons & Bobo, 2015).
3
See Reichelmann (2018) for full details about the experiment including the exact images and
wording of the slavery exposure conditions. In analyses conducted to date, the slavery exposure
experiment has only shown effects on the collective emotion variables (Reichelmann, 2020).
Nonetheless, to be conservative and to ensure full transparency, we include the experimental
conditions as controls in the analyses presented here.
Independent Variables
We utilize responses to questions gauging five dimensions of respondents’ racial

identity. These questions mirror those used in an identity module fielded as part of
the 2014 General Social Survey. Each item used a set of response options ranging
from 0 (“not at all”) to 10 (“completely”). Prominence refers to the respondent’s
self-rated importance of their racial identity and is measured with the question:
“How much is being White an important part of how you see yourself?” Salience
refers to the likelihood that an identity will be invoked in daily life and is measured
with the question: “In general, how much do you find that being White influences or
guides how you behave?” Private self-regard, also known as racial self-esteem,
refers to the respondent’s feelings about their racial group, and is captured using the
question: “How proud are you to be White?” Public self-regard refers to the
perception of how one’s racial group is seen by others and is measured with the
question: “How much do you think people in the U.S. respect Whites?” Verification
refers to how much an individual believes that others see them in the same way they
see themselves, which in this case is focusing on white racial identification, and is
measured with the question: “Think about how you see yourself as White. How
much do you think your friends see you as White?”
Dependent Variables
We examine nine outcomes assessing the extent to which a respondent favors actions
(by the government or themselves) to repair the harm caused by slavery and its
ensuing legacy of racial inequality. We frame four of the outcomes as explicit
reparations since they specifically reference either symbolic or material recognition
of the victims of slavery. We frame four other outcomes as government intervention
on behalf of black Americans. While the latter four items make no explicit mentions
of slavery, they involve reparative policies since they address racial inequalities that
are traceable to the legacy of slavery (even if such links are not reliably made by all
observers). Our final outcome shifts attention away from government action to the
realm of white Americans’ perceived personal responsibility for improving race
relations in the U.S.
The four explicit reparations items focus on reparative action to remedy the
lasting effects of slavery, and range from purely symbolic (apology) to material
(financial payments) forms of reparative action. The first and fourth items below
were adapted from Brown et al. (2008), and the second and third items were
developed by the authors. These items utilized response options ranging from
1 (“strongly favor”) to 7 (“strongly oppose”). The items were:
1. The U.S. government should apologize for all the maltreatment and deprivation
that was caused as a result of slavery.
2. The U.S. government should dedicate a memorial to the victims of slavery.
3. The U.S. government should dedicate a memorial to the victims of slavery on the
National Mall.
4. Black Americans should receive financial payments as a reparation for the
damage that was caused as a result of slavery.
The next four items capture feelings toward the role of government intervention
to ameliorate existing racial inequalities. These items vary in scope from ensuring
fair treatment to the provision of special or preferential treatment for black Amer-
icans. The first was adapted from Brown et al. (2008), and the other three were
adapted from the GSS. These items had response options ranging from 1 (“strongly
favor”) to 7 (“strongly oppose”). The items were:
1. American universities should have special scholarships for Black American
students.
2. Blacks should be given preferential hiring and promotion.
3. The government should see to it that Blacks get fair treatment in jobs.
4. The government should make every effort to improve the social and economic
conditions of Blacks.
A final item was adapted from Powell et al. (2005), and gauges feelings of
personal responsibility for the current relationship between black and white Amer-
icans. This item reads: “I believe I should help repair the damage caused to Black
Americans by White Americans,” with responses ranging from 1 (“strongly agree”)
to 7 (“strongly disagree”).
Control Variables
Our regression models include controls for factors known to correlate with the sorts
of racial policy views under examination. Gender is a binary variable, coded male ¼ 1
and female ¼ 0. Income is measured with ten categories, ranging from “less than
$10,000” to “more than 200,000.” Education is a seven-category variable with the
following options: less than high school, high school graduate, some college,
Associates degree, Bachelor’s degree, some graduate work, and graduate or profes-
sional degree. Region refers to a respondent’s current region of residence, and is
modeled using a series of dummy variables identifying Northeast, Midwest, and
West with South as the reference category in the regressions. Age is an ordinal
variable with the following seven categories: 18 to 21; 22 to 25; 26 to 35; 36 to 45;
46 to 55; 56 to 65; and over 65. Conservative political ideology is measured using a
standard seven-category scale with the following options: extremely liberal (1),
liberal (2), slightly liberal (3), moderate/middle of the road (4), slightly conservative
(5), conservative (6), and extremely conservative (7); thus, higher values indicate
stronger political conservatism. “Condition” refers to the experimental condition that
the respondent was exposed to (written perpetrator-focused, written victim-focused,
visual perpetrator-focused, or visual victim-focused), with no exposure being the

reference group.
Analytic Strategy
The overall goal of our analysis is to acquire a deeper understanding of the interplay
between the individual racial identity dimensions and white Americans’ views on a
range of reparative measures. Given the exploratory nature of this research, and
because other recent empirical work demonstrates that different elements of WRI
have distinct effects on various racial policy attitudes (Hunt & Reichelmann, 2019),
we have elected to model associations using the individual identity dimensions and
reparations items, rather than building indices of each. While the utility of using the
individual identity items has been demonstrated in prior work, the current study
explores these items in relation to previously unexamined outcomes. Regarding
those outcomes, our individual-item approach allows us to attend to the
symbolic vs. material axis among our explicit reparations measures, as well as
differentiating those from the government intervention and personal responsibility
items. As our dependent variables are seven-category ordinal measures, we ran both
ordinary least squares (OLS) and ordinal logistic regression models; these
approaches produced similar results. We present the OLS results, though the logistic
regression results are available upon request.
Results
Table 5.1 reports descriptive statistics for all of the variables used in our final model.
Regarding explicit reparations, whites show more support for building a memorial
for victims of slavery in general (x ¼ 3.53) or on the National Mall (x ¼ 3.71), than
for an apology (x ¼ 4.16) or for financial payments (x ¼ 5.18) as reparations for
slavery. Regarding the government intervention items, whites show more support for
the principle that the government should ensure that black Americans receive fair
treatment in jobs (x ¼ 2.98) and should improve the social and economic conditions
of black Americans (x ¼ 3.63), than they do for special treatment in the form of
scholarships for higher education (x ¼ 4.34) and preferential hiring (x ¼ 4.97) for
blacks.
Overall, the data suggest that white Americans show mixed support for reparative
actions on behalf of blacks, with opposition increasing as reparations move from the
symbolic to the material realm (i.e., apology vs. financial payments) and as govern-
ment policies move from those ensuring equal treatment to those involving per-
ceived special or preferential treatment for blacks (Bobo & Kluegel, 1993). Mixed
levels of support/opposition also describe whites in the realm of personal
Table 5.1 Descriptive statistics of variables (n ¼ 866)

Variable Mean Std.Dev. Min Max
Apology 4.16 2.21 1 7
Memorial 3.53 2.09 1 7
Memorial on the Mall 3.71 2.05 1 7
Financial payments 5.18 2.00 1 7
Special university scholarships 4.34 2.08 1 7
Preferential hiring 4.97 1.94 1 7
Fair treatment 2.98 1.94 1 7
Improve soc. and econ. conditions 3.63 2.01 1 7
Personal feeling responsibility 4.00 2.13 1 7
Prominence 5.90 3.27 0 10
Salience 6.11 2.92 0 10
Private self-regard 6.75 2.98 0 10
Public self-regard 6.90 2.39 0 10
Verification 7.04 2.82 0 10
Male .58 .49 0 1
Income 5.77 2.24 1 10
Education 4.64 1.77 1 7
Region of residence:
Northeast .21 .41 0 1
Midwest .18 .39 0 1
South .39 .49 0 1
West .22 .42 0 1
Age 4.18 1.72 1 7
Conservative 3.78 1.79 1 7
Experimental condition:
Written, perpetrator .19 .39 0 1
Written, victim .20 .40 0 1
Visual, perpetrator .22 .41 0 1
Visual, victim .21 .41 0 1
No representation (control) .19 .39 0 1
responsibility where whites score, on average, at the mid-point (x ¼ 4.00) on the item
asking about their own role in repairing the damage caused as a result of slavery.
Regarding the independent variables, the mean response on all five identity
dimension items is above the midpoint of 5: prominence (x ¼ 5.90), salience (x
¼ 6.11), private self-regard (x ¼ 6.75), public self-regard (x ¼ 6.90), and verification
(x ¼ 7.04). The means of the control variables reveal that the study sample was 58%
male and on average between 36 and 45 years old. Respondents had an average
income of between $50,000 to $75,000 (x ¼ 5.77) and had, on average, completed at
least a Bachelor’s degree (x ¼ 4.64). A plurality of respondents reside in the South
(39%) relative to the other regions (Northeast ¼ 21%; Midwest ¼ 18%; West ¼ 22%).
And, respondents are slightly more liberal than conservative (x ¼ 3.78) on average.
Table 5.2 presents the correlations of all the study variables. Among the 36 cor-
relations between the items measuring support for reparations (broadly understood),
16 are strong (.6 or greater) and 19 are moderate (between .3 and .6). All of these
correlations are expected and explain why the items could either be run as a scale or
independently. Of the ten correlations between the five identity dimensions, four are
strong (.6 or greater) and six are moderate (between .3 and .6), again demonstrating
that they could either be run as a scale or independently. The strong relationship
between prominence and salience is consistent with the findings of previous work
(Brenner et al., 2014; Stryker & Serpe, 1994). The associations of prominence and
salience with verification are part of the broader narrative of whiteness: we seek our
friends’ verification of identities that are important to us, and that we regularly enact.
And, the strong association between prominence and private self-regard supports a
central tenet of SIT: when the group identity becomes central to how we view
ourselves, we are motivated to see that identity as good, and therefore have pride
in it. Of the control variables, no relationships demonstrated a strong correlation.
Conservatism is the only control variable that demonstrates a moderate relationship
with some of the dependent variables, but this relationship was not above .35.
To answer our primary research question, we ran a series of multivariate OLS
regression models designed to assess relationships between the WRI elements and
views on reparations, controlling for respondents’ background characteristics. Given
the strength of the correlations seen in Table 5.2, we examined Variance Inflation
Factor (VIF) and Tolerance statistics associated with all regression coefficients. VIF
values were less than 5 and Tolerance values were less than .2 indicating no evidence
of multicollinearity when using standard decision rules.
Table 5.3 contains the regression results for all nine outcomes. Beginning with
the explicit reparations items (models 1–4), we see that three of the five identity
dimensions have statistically significant associations with at least two outcomes.
First, the salience measure shows inverse associations with both the “apology”
(b ¼ .08; p < .05) and “financial payments” (b ¼ .10; p < .01) outcomes.
Thus, as WRI salience increases, opposition to reparative actions in these two forms
decreases. In contrast, private self-regard shows positive associations with the same
two outcomes. Thus, as whites’ pride in their racial group membership increases, so
does opposition to an apology (b ¼ .12; p < .01) and to financial payments to blacks
(b ¼ .09; p < .01). Finally, public self-regard, like salience, shows significant inverse
associations with these same two outcomes, as well as with the idea of a memorial to
victims of slavery. Thus, the more whites feel that others have a greater respect for
their racial group, they are less opposed to an apology for slavery (b ¼ .12;
p < .01), a memorial to victims of slavery (b ¼ .09; p < .05), and financial
payments to the descendants of enslaved persons (b ¼ .10; p < .01).
These findings indicate that WRI dimensions play differing roles in shaping
views on reparations: white racial pride tends to increase opposition to particular
forms of reparative action, while higher levels of public self-regard, and to a lesser
98
Table 5.2 Correlation matrix of variables (n ¼ 866)

Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
(1) Apology 1.00
(2) Memorial 0.72*** 1.00
(3) Memorial on Mall 0.71*** 0.89*** 1.00
(4) Financ. payments 0.65*** 0.48*** 0.50*** 1.00
(5) Scholarships 0.65*** 0.60*** 0.60*** 0.63*** 1.00
(6) Pref. hiring 0.60*** 0.47*** 0.51*** 0.71*** 0.75*** 1.00
(7) Fair treatment 0.42*** 0.48*** 0.46*** 0.26*** 0.51*** 0.35*** 1.00
(8) Improve cond. 0.60*** 0.59*** 0.61*** 0.49*** 0.72*** 0.60*** 0.67*** 1.00
(9) Personal respons. 0.54*** 0.49*** 0.48*** 0.47*** 0.55*** 0.50*** 0.36*** 0.53*** 1.00
(10) Prominence 0.12*** 0.19*** 0.18*** 0.04 0.06 0.09** 0.16*** 0.09** 0.07* 1.00
(11) Salience 0.03 0.12*** 0.09** 0.16*** 0.04 0.16*** 0.12*** 0.04 0.07* 0.70*** 1.00
(12) Private S.R. 0.19*** 0.20*** 0.20*** 0.06 0.17*** 0.04 0.16*** 0.17*** 0.13*** 0.75*** 0.53*** 1.00
(13) Public S.R. 0.15*** 0.06 0.07* 0.22*** 0.16*** 0.22*** 0.05 0.17*** 0.23*** 0.42*** 0.49*** 0.35*** 1.00
(14) Verification 0.01 0.09** 0.05 0.10*** 0.04 0.13*** 0.04 0.04 0.02 0.61*** 0.66*** 0.53*** 0.45*** 1.00
(15) Male 0.01 0.07* 0.05 0.07* 0.01 0.04 0.08* 0.03 0.06 0.24*** 0.21*** 0.18*** 0.13*** 0.14***
(16) Income 0.01 0.00 0.00 0.03 0.01 0.07* 0.06 0.01 0.01 0.18*** 0.22*** 0.15*** 0.21*** 0.15***
(17) Education 0.07* 0.02 0.04 0.09** 0.07* 0.11*** 0.01 0.05 0.03 0.14*** 0.17*** 0.08* 0.16*** 0.07*
(18) Lives in Northeast 0.04 0.01 0.02 0.00 0.01 0.00 0.01 0.01 0.01 0.02 0.00 0.02 0.01 0.03
(19) Lives in Midwest 0.03 0.00 0.01 0.09** 0.05 0.08* 0.03 0.01 0.05 0.08** 0.07* 0.05 0.07* 0.08*
(20) Lives in South 0.03 0.03 0.04 0.04 0.00 0.02 0.03 0.03 0.01 0.02 0.03 0.05 0.02 0.07*
(21) Lives in West 0.03 0.05 0.03 0.03 0.06 0.06 0.02 0.05 0.03 0.03 0.03 0.01 0.03 0.02
(22) Age 0.23*** 0.21*** 0.23*** 0.28*** 0.20*** 0.23*** 0.04 0.14*** 0.17*** 0.11*** 0.01 0.10*** 0.15*** 0.05
(23) Conservative 0.31*** 0.28*** 0.30*** 0.28*** 0.35*** 0.32*** 0.22*** 0.34*** 0.30*** 0.11*** 0.01 0.23*** 0.13*** 0.01
(24) Written, perp. 0.03 0.03 0.04 0.02 0.01 0.00 0.03 0.01 0.01 0.02 0.01 0.01 0.01 0.05
(25) Written, victim 0.02 0.01 0.01 0.03 0.00 0.03 0.04 0.02 0.01 0.01 0.03 0.04 0.05 0.05
(26) Visual, perp. 0.06 0.06 0.07* 0.01 0.02 0.03 0.06 0.04 0.04 0.06 0.05 0.06 0.06 0.05
(27) Visual, victim 0.07* 0.08* 0.09** 0.04 0.05 0.01 0.05 0.05 0.04 0.02 0.00 0.01 0.04 0.03
(28) Control group 0.02 0.06 0.05 0.00 0.03 0.00 0.06 0.04 0.09** 0.01 0.04 0.02 0.04 0.02
A. V. Reichelmann and M. O. Hunt
Variables (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28)
(15) Male 1.00
(16) Income 0.26*** 1.00
(17) Education 0.23*** 0.44*** 1.00
(18) Lives in 0.06 0.10*** 0.08** 1.00
Northeast
(19) Lives in 0.06 0.14*** 0.08* 0.24*** 1.00
Midwest
(20) Lives in 0.06 0.03 0.11*** 0.41*** 0.37*** 1.00
South
(21) Lives in West 0.06 0.07* 0.12*** 0.28*** 0.25*** 0.43*** 1.00
(22) Age 0.05 0.08** 0.01 0.07* 0.09** 0.05 0.03 1.00
(23) Conservative 0.04 0.00 0.10*** 0.06 0.03 0.11*** 0.11*** 0.16*** 1.00
(24) Written, perp. 0.03 0.05 0.01 0.01 0.01 0.00 0.01 0.02 0.03 1.00
(25) Written, 0.01 0.04 0.09** 0.02 0.05 0.04 0.06 0.03 0.05 0.24*** 1.00
victim
(26) Visual, perp. 0.04 0.01 0.03 0.00 0.01 0.00 0.01 0.01 0.02 0.26*** 0.26*** 1.00
5 White Racial Identity and Reparations for Slavery
(27) Visual, victim 0.01 0.01 0.06 0.02 0.05 0.02 0.00 0.04 0.06 0.25*** 0.25*** 0.27*** 1.00
(28) Control group 0.04 0.03 0.06 0.00 0.00 0.06 0.06 0.04 0.04 0.23*** 0.24*** 0.25*** 0.24*** 1.00
*p < 0.05
**p < 0.01
***p < 0.001
99
degree salience, tend to decrease opposition (i.e., increase support) for reparative
behavior in some fashion. Cast in Croll’s (2007) language, private self-regard fosters
defensive postures in white Americans, while public self-regard and salience appear
to underlie more progressive stances on reparations.
Turning to the set of four items gauging views of government intervention on
behalf of black Americans, we see that all five identity dimensions have at least one
significant association with an outcome we model. Most noteworthy in this set of
findings are the consistent effects of private and public self-regard. Reminiscent of
the defensive and progressive effects just seen, private self-regard is significantly
associated with opposition to special scholarships for American universities
(b ¼ 0.16; p < .001), preferential hiring (b ¼ .11; p < .001), and government efforts
to improve the social and economic conditions of black Americans (b ¼ .11;
p < .01). In contrast, public self-regard is inversely associated with all four forms
of government intervention we examine: special scholarships (b ¼ .11; p < .01),
preferential hiring (b ¼ .08; p < .01), fair treatment (b ¼ .09; p < .01), and
improving social and economic conditions (b ¼ .14; p < .001). The findings
reinforce the earlier impression that white racial pride is a key identity dimension
underlying opposition to reparative actions toward blacks, while whites’ beliefs that
their racial group is respected by others plays the opposite role in fostering support
for such measures.
Regarding other significant effects in models 5 through 8 in Table 5.3, WRI
prominence (importance) shows an inverse association with opposition to preferen-
tial hiring for blacks (b ¼ .09; p < .01). Thus, as whites accord their racial identity
more importance, they are less opposed to race-targeting policy of this sort. Further,
in a reversal of the patterns observed for the explicit reparations items, WRI salience
shows a positive association with the “ensuring fair treatment” outcome (b ¼ .08;
p < .05). Thus, whites who feel that their racial identity more strongly guides their
behavior are also more likely to oppose policy ensuring fair treatment across racial
lines. Lastly, WRI verification shows a significant inverse association with the fair
treatment outcome (b ¼ .07; p < .05). Thus, whites whose racial identities are
verified by friends at higher levels are less opposed to government efforts to ensure
blacks receive fair treatment in the workplace. These prominence and verification
effects represent progressive effects, while the salience effect in this domain repre-
sents a defensive posture—a departure from what we observed for two of the explicit
reparations outcomes.
Taken together, the four explicit reparations and four government intervention
items offer important insights into the sources of white Americans’ support for
national and state-level efforts to make amends for slavery and the legacy of racial
inequality the United States has witnessed in its wake. The final outcome we
examine shifts our focus to white Americans’ reports regarding their own responsi-
bility for addressing racial inequality and race relations. For this outcome, private
and public self-regard show now familiar defensive and progressive implications
respectively: private self-regard predicts disagreement with the statement “I feel I
should help repair the damage caused to Black Americans by White Americans”
(b ¼ .09; p < .05), while public self-regard predicts agreement with this same
Table 5.3 Unstandardized OLS estimates of reparation items (n ¼ 866)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Memorial on Financial Special Prefer. Fair Improve Personal
Apology Memorial Mall payments scholarships hiring treatment conditions respons.
Prominence .08 .06 .07 .01 .03 .09** .06 .05 .06
(.04) (.04) (.04) (.04) (.04) (.03) (.04) (.04) (.04)
Salience .08* .05 .03 .10** .02 .04 .08* .05 .10**
(.04) (.04) (.04) (.03) (.04) (.03) (.03) (.03) (.04)
Private self- .12** .05 .06 .09** .16*** .11*** .03 .11** .09*
regard
(.04) (.04) (.04) (.03) (.04) (.03) (.03) (.03) (.04)
Public self- .12** .09* .07 .10** .11** .08** .09** .14*** .21***
regard
(.04) (.04) (.04) (.03) (.04) (.03) (.03) (.03) (.04)
Verification .03 .02 .04 .02 .05 .02 .07* .04 .03
5 White Racial Identity and Reparations for Slavery
(.03) (.03) (.03) (.03) (.03) (.03) (.03) (.03) (.04)

Male .01 .19 .10 .21 .03 .02 .17 .11 .28*
(.15) (.14) (.14) (.13) (.14) (.13) (.13) (.13) (.14)
Income .03 .02 .01 .06 .03 .01 .04 .03 .04
(.03) (.03) (.03) (.03) (.03) (.03) (.03) (.03) (.03)
Education .05 .01 .04 .06 .04 .06 .02 .03 .01
(.04) (.04) (.04) (.04) (.04) (.04) (.04) (.04) (.04)
Region reference: South
Northeast .06 .11 .01 .27 .23 .24 .01 .09 .10
(.19) (.19) (.18) (.18) (.19) (.17) (.18) (.18) (.19)
Midwest .03 .11 .08 .37* .14 .24 .14 .04 .25
(.19) (.19) (.18) (.17) (.18) (.16) (.18) (.18) (.18)
West .02 .17 .09 .15 .04 .05 .07 .06 .01
101
(continued)
Table 5.3 (continued)
102
(1) (2) (3) (4) (5) (6) (7) (8) (9)

Memorial on Financial Special Prefer. Fair Improve Personal
Apology Memorial Mall payments scholarships hiring treatment conditions respons.
(.19) (.18) (.18) (.16) (.17) (.16) (.17) (.17) (.18)
Age .20*** .18*** .19*** .26*** .16*** .20*** .02 .07 .09*
(.04) (.04) (.04) (.04) (.04) (.03) (.04) (.04) (.04)
Conservative .27*** .25*** .27*** .23*** .30*** .28*** .20*** .30*** .27***
(.04) (.04) (.04) (.04) (.04) (.04) (.04) (.04) (.04)
Condition reference: control group
Written, .07 .28 .27 .23 .10 .14 .30 .08 .31
perpetrator
(.23) (.22) (.21) (.21) (.21) (.20) (.22) (.20) (.21)
Visual, .22 .01 .08 .10 .03 .03 .02 .05 .43*
perpetrator
(.22) (.21) (.21) (.19) (.21) (.19) (.21) (.20) (.22)
Written, victim .07 .24 .13 .17 .05 .21 .37 .17 .27
(.23) (.22) (.21) (.20) (.21) (.19) (.21) (.20) (.21)
Visual, victim .29 .53** .52** .05 .30 .02 .44* .31 .45*
(.22) (.20) (.19) (.20) (.20) (.19) (.19) (.19) (.21)
Constant 2.69*** 1.89*** 1.98*** 3.87*** 2.88*** 3.84*** 2.37*** 2.71*** 3.32***
(.43) (.42) (.40) (.40) (.41) (.38) (.39) (.37) (.43)
Adjusted R2 .17 .13 .15 .18 .17 .18 .08 .16 .16
Robust standard errors in parentheses
*p < 0.05
**p < 0.01
***p < 0.001, two-tailed
A. V. Reichelmann and M. O. Hunt
statement (b ¼ .21; p < .001). We also see a progressive manifestation of salience

in this model (b ¼ .10; p < .01), reminiscent of its association with two of the
explicit reparations items in models 1 and 4 of Table 5.3.
The control variables also fill out the story. Older and more conservative respon-
dents were consistently and significantly more opposed to the examined reparations,
as well as more likely to disavow personal responsibility for repairing black/white
relations in the United States. Exposure to a visual victim-focused representation of
slavery was inversely associated with opposition of a memorial in general
(b ¼ .53; p < .01) and on the Mall (b ¼ .52; p < .01) as well as fair treatment
(b ¼ .44; p < .05), while exposure to any visual representation was inversely
related to disagreement regarding personal responsibility. In other words, respon-
dents who were exposed to visual representations were more likely to support
reparative actions of memorialization and government intervention of fair treatment
as well as agree that they bear a personal responsibility for improving the conditions
of blacks, when compared against those who were exposed to no representation.
Other variables did not show consistent relationships with the outcomes we examine.
Discussion
In this study, we examine the relationship between five dimensions of WRI and
whites’ views on a range of different reparative measures targeting the descendants
of enslaved persons in the United States. Our analyses reveal distinct patterns
between WRI and views on reparations, broadly understood. The most consistent
of these effects involve private and public self-regard. Private self-regard is posi-
tively and significantly associated with six of the nine outcomes we examine, while
public self-regard shows significant inverse associations with eight of them. Sub-
stantively, private self-regard predicts opposition to reparative measures, while
public self-regard fosters support for such measures. Seen alongside selected effects
of the other three identity dimensions (which show mixed and in some cases
inconsistent or no associations with the outcomes we examine), we observe clear
evidence that white Americans’ opinions regarding reparations are directly impacted
by their own racial self-understandings. Our findings also lend additional support to
arguments suggesting that WRI can manifest in both progressive and defensive ways
(Croll, 2007; Knowles et al., 2014).
Our findings demonstrate that private self-regard has defensive implications—
consistent with other recent work (Hunt & Reichelmann, 2019). However, the
current study also provides new insight into the role of public self-regard in fostering
more progressive stances on issues of racial inequality and its possible repair.
Specifically, our results suggest that when whites feel others respect their group,
they are more willing to support reparative steps that assist—materially and sym-
bolically—other groups that have been wronged. That is, when white Americans feel
their racial group is respected, they are more likely to validate the moral claims of
non-white Americans. This pattern holds across all of the outcomes we analyze
except the question gauging support for a memorial to victims of slavery on the
National Mall—the one outcome that was not significantly associated with any of the
identity dimensions. Given that we also observe this same progressive impact of
public self-regard on the outcome gauging feelings of personal responsibility, our
results indicate that whites with higher public self-regard may be more likely to
engage in daily micro-reparative acts, such as those propositioned by
Ocbazghi (2017).
Combined with our previous work (Hunt & Reichelmann, 2019), the consistent
observation of an alignment between white racial pride (private self-regard) and
racial conservatism on the one hand, and between whites’ greater perceived respect
by others (public self-regard) and racially progressive stances, presents a new
challenge to identity scholars. Specifically, our findings strongly signal that two
relatively understudied dimensions of WRI represent key pieces in our understand-
ing of whites’ racial attitudes and politics. WRI may have been relatively invisible,
and its effects underestimated, because key identity dimensions were not being
examined. Prominence is more widely used in research than any other identity
dimension, and is commonly used as a proxy for WRI itself (Croll, 2007; Jardina,
2019). This is potentially problematic, because prominence is a measure of general
self-concept and, as such, may not be reliably related to attitudes about more specific
phenomena and domains. Salience is also very widely used in the identity literature
where, as a measure of identity enactment, it is most likely related to outcomes with a
behavioral referent. This theoretical expectation is borne out in the current study
where higher WRI salience is only significantly linked with outcomes associated
with behaviors such as an apology, financial payments, fair treatment, and the belief
in personal responsibility for improving black/white relations.
Verification’s mean in Table 5.1 (7.04) indicates that, on average, whites do feel
that others see them as white. However, this recognition shows no consistent pattern
with support for reparations. While all its relationships trend mildly negative, only
the relationship between verification and fair treatment is significant (Table 5.3).
This could be due to the current disconnect between racialized political identities and
decades of whites (including some academics) insisting on the relative
unimportance, or even absence, of such an identity. Given that whiteness is still
the reference category in our society, the relationship (or lack thereof) between
verification and the sorts of outcomes examined in the current study may remain
relatively underdeveloped.
In contrast to prominence and salience, the relatively understudied dimensions of
self-regard are consistently and meaningfully tied to the sorts of outcomes examined
in the current study. SIT teaches us that people are motivated to see themselves in a
positive light—a motivation that shapes persons’ perceptions of self and other, and
their interaction with the social world. As such, private self-regard, or pride in being
white, likely underlies opposition to reparations to black Americans (for harms
suffered at the hands of white Americans) because such policies challenge prideful
whites’ self-views as positive and praiseworthy. Public self-regard, on the other
hand, has to do with perceptions of how respected white Americans feel, and as we
have seen, is associated with support for actions that are reparative in nature (the flip-
side of this is, of course, that whites who feel relatively disrespected by others are
more opposed to reparations to black Americans). The logic here is that perception
of respect or esteem from others translates to support for policies that benefit others;
sensing validation of the in-group may increase the likelihood of sympathy to
outgroups.
Future work should seek to more thoroughly theorize and demonstrate whether
the self-regard findings we focus on are generalizable to other outcomes and groups.
Specifically, future work should utilize a racially-comparative framework, to assess
how identity-attitude relationships may vary across major ethno-racial lines in the
United States (Hunt et al., 2013). In so doing, researchers should seek to develop and
implement multiple items tapping each identity dimension to move beyond the
limitations of single-item indicators.
In the years leading up to the 2016 election, scholars focused increasing attention
on whiteness as a potent social force in American life (Abrajano & Hajnal, 2015;
Bonilla-Silva, 2009; Feagin, 2013). In the ensuing years, this fact has become even
clearer with studies demonstrating the impact of WRI on a host of important out-
comes, including racial attitudes (Hunt & Reichelmann, 2019), attitudes toward
immigration, globalization, and redistribution, and on voting behavior (Jardina,
2019). While most of this work highlights the conservative and defensive manifes-
tations of WRI, recent events also demonstrate white Americans’ widespread par-
ticipation in diverse racial coalitions underlying anti-racist social action (e.g., BLM)
in the wake of the murder of George Floyd and other highly-publicized injustices.
Given these facts, identity theorists and researchers should more deeply and criti-
cally examine white Americans’ self-understandings and how such views shape the
way whites conceptualize the legacy of slavery and its current manifestations in the
United States.
Acknowledgements The data collection was completed with funding from the following agen-
cies: the Social Psychology Section of the American Sociological Association (Graduate Student
Investigator Award – 2016); the Fahs-Beck Fund for Research and Experimentation at The
New York Community Trust (Doctoral Dissertation Grant Recipient – 2016); and Northeastern
University’s Provost’s Office (2017), Department of Sociology (2016), and Brudnick Center on
Conflict and Violence (2016).
References
Abrajano, M., & Hajnal, Z. L. (2015). White backlash: immigration, race, and American politics.
Princeton University Press.
Armange, R., & Mullet, E. (2016). Slave descendants’ views regarding national policies on
reparations: a Martinican perspective. Social Science Information, 55(4), 511–530.
Balfour, L. (2014). Unthinking racial realism: a future for reparations? Du Bois Review: Social
Science Research on Race, 11(01), 43–56.
Balfour, L. (2015). Ida B. Wells and ‘color line justice’: Rethinking reparations in feminist terms.
Perspectives on Politics, 13(03), 680–696.
Blumer, H. (1958). Race prejudice as a sense of group position. The Pacific Sociological Review, 1
(1), 3–7.
Bobo, L. D. (1983). Whites’ opposition to busing: symbolic racism or realistic group conflict?
Journal of Personality and Social Psychology, 45(6), 1196–1210.
Bobo, L. D. (2000). Race and beliefs about affirmative action. In D. O. Sears, J. Sidanius, &
L. Bobo (Eds.), Racialized politics: the debate about racism in America (pp. 13–165). Univer-
sity of Chicago Press.
Bobo, L. D., & Hutchings, V. L. (1996). Perceptions of racial group competition: extending
Blumer’s theory of group position to a multiracial social context. American Sociological
Review, 61, 951–972.
Bobo, L. D., & Kluegel, J. R. (1993). Opposition to race-targeting: self-interest, stratification
ideology, or racial attitudes? American Sociological Review, 58(4), 442–464.
Bobo, L. D., & Kluegel, J. R. (1997). Status, ideology, and dimensions of whites’ racial beliefs and
attitudes: progress and stagnation. In S. A. Tuch & J. K. Martin (Eds.), Racial attitudes in the
1990s: Continuity and change (pp. 93–119). Praeger Publishers.
Bobo, L. D., Kluegel, J. R., & Smith, R. A. (1997). Laissez-Faire Racism; The crystallization of a
kinder, gentler, antiblack ideology. In S. A. Tuch & J. K. Martin (Eds.), Racial attitudes in the
1990s: Continuity and change (pp. 15–42). Praeger Publishers.
Bonilla-Silva, E. (2009(2014)). Racism without racists: color-blind racism and the persistence of
racial inequality in America. Lanham: Rowman & Littlefield Publishers.
identity theory: an empirical examination. Social Psychology Quarterly, 77(3), 231–252.
Brown, R., González, R., Zagefka, H., Manzi, J., & Čehajić, S. (2008). Nuestra Culpa: Collective
guilt and shame as predictors of reparation for historical wrongdoing. Journal of Personality
and Social Psychology, 94(1), 75–90.
Cooper, A. D. (2012). From slavery to genocide: the fallacy of debt in reparations discourse.
Journal of Black Studies, 43(2), 107–126.
Craemer, T. (2009a). Framing reparations. Policy Studies Journal, 37(2), 275–298.
Craemer, T. (2009b). Psychological ‘self–other overlap’ and support for slavery reparations. Social
Science Research, 38(3), 668–680.
Croll, P. (2007). Modeling determinants of white racial identity: results from a new national survey.
Social Forces, 86(2), 613–642.
Darby, D., & Branscombe, N. R. (2014). Beyond the sins of the fathers: responsibility for
inequality. Midwest Studies in Philosophy, 38(1), 121–137.
Dawson, M. C., & Popoff, R. (2004). REPARATIONS: Justice and greed in black and white. Du
Bois Review: Social Science Research on Race, 1(01), 47–91.
Deaux, K. (1996). Social identification. In E. T. Higgins & A. W. Kruglanski (Eds.), Social
psychology: Handbook of basic principles (pp. 777–798). Guilford Press.
Deaux, K., & Martin, D. (2003). Interpersonal networks and social categories: specifying levels of
context in identity processes. Social Psychology Quarterly, 66(2), 101–117.
Doane, A. W. (1997). Dominant group identity in the United States: The role of ‘hidden’ ethnicity
in intergroup relations. The Sociological Quarterly, 38(3), 375–397.
Feagin, J. R. (2013). The white racial frame: centuries of racial framing and counter-framing (2nd
ed.). Routledge.
Feldman, S., & Huddy, L. (2005). Racial resentment and white opposition to race-conscious
programs: principles or prejudice? American Journal of Political Science, 49(1), 168–183.
Forman, T. A., & Lewis, A. E. (2006). RACIAL APATHY AND HURRICANE KATRINA: The
social anatomy of prejudice in the post-civil rights era. Du Bois Review: Social Science
Research on Race, 3(01), 175–202.
Gecas, V. (1982). The self-concept. Annual Review of Sociology, 8, 1–33.
Glaser, J. M., & Ryan, T. J. (2011). Remorse, retribution, and restoration: changing minds, if not
hearts, about reparations. Paper presented at the Annual Meeting of the American Political
Science Association, Seattle, WA, Sept. 1–4.
Goren, M. J., & Plaut, V. C. (2012). Identity form matters: white racial identity and attitudes toward
diversity. Self and Identity, 11(2), 237–254.
Hartmann, D., Gerteis, J., & Croll, P. R. (2009). An empirical assessment of whiteness: hidden from
how many? Social Problems, 56(3), 403–424.
Hogg, M. A., Terry, D. J., & White, K. M. (1995). A tale of two theories: a critical comparison of
identity theory with social identity theory. Social Psychology Quarterly, 58(4), 255–269.
Howard, J. A. (2000). Social psychology of identities. Annual Review of Sociology, 26(1), 367–393.
Hughes, M. (1997). Symbolic racism, old-fashioned racism, and whites’ opposition to affirmative
action. In S. A. Tuch & J. K. Martin (Eds.), Racial attitudes in the 1990s: Continuity and change
(pp. 45–75). Praeger.
Hughey, M. W. (2012). White bound: nationalists, anti-racists, and the shared meanings of race.
Stanford University Press.
Hunt, M. O. (2007). African-American, hispanic, and white beliefs about black/white inequality,
1977-2004. American Sociological Review, 72, 390–415.
Hunt, M. O. (2020). Racial identity among white Americans: Structure, antecedents, and conse-
quences. In R. T. Serpe, R. Stryker, & B. Powell (Eds.), Identity and symbolic interaction:
deepening foundations, building bridges (pp. 149–168). Springer Nature.
Hunt, M. O., Jackson, P. B., Kye, S. H., Powell, B., & Steelman, L. C. (2013). Still color-blind? The
treatment of race, ethnicity, intersectionality, and sexuality in sociological social psychology.
Advances in Group Processes, 30, 21–45.
Hunt, M. O., & Reichelmann, A. (2019). Racial identity and racial attitudes among white Amer-
icans. In J. Stets & R. Serpe (Eds.), Identities in everyday life (pp. 217–238). Oxford University
Press.
Jardina, A. (2019). White identity politics. Cambridge University.
Khanna, N. (2010). ‘If You’re Half Black, You’re Just Black’: Reflected appraisals and the
persistence of the one-drop rule. The Sociological Quarterly, 51(1), 96–121.
Kinder, D. R., & Sanders, L. M. (1996). Divided by color: racial politics and democratic ideals.
University of Chicago Press.
Klandermans, B., Werner, M., & van Doorn, M. (2008). Redeeming apartheid’s legacy: collective
guilt, political ideology, and compensation. Political Psychology, 29(3), 331–349.
Kluegel, J. R. (1990). Trends in whites’ explanations of the black-white gap in socioeconomic
status, 1977–1989. American Sociological Review, 55(4), 512–525.
Knowles, E. D., Lowery, B. S., Chow, R. M., & Unzueta, M. M. (2014). Deny, distance, or
dismantle? How white Americans manage a privileged identity. Perspectives on Psychological
Science, 9(6), 594–609.
Knowles, E. D., & Peng, K. (2005). White selves: conceptualizing and measuring a dominant-group
identity. Journal of Personality and Social Psychology, 89(2), 223–241.
Lickel, B., Schmader, T., & Barquissau, M. (2004). The evocation of moral emotions in intergroup
contexts. In N. R. Branscombe & B. Doosje (Eds.), Collective guilt: international perspectives
(pp. 35–55). Cambridge University Press.
Lowery, B. S., Unzueta, M. M., Knowles, E. D., & Goff, P. A. (2006). Concern for the in-group and
opposition to affirmative action. Journal of Personality and Social Psychology, 90(6), 961–974.
Luhtanen, R., & Crocker, J. (1991). Self-esteem and intergroup comparisons: toward a theory of
collective self-esteem. In T. A. Wills (Ed.), Social comparison: contemporary theory and
research (pp. 503–513). Lawrence Erlbaum.
Major, B., Blodorn, A., & Blascovich, G. M. (2018). The threat of increasing diversity: why many
white Americans support Trump in the 2016 Presidential Election. Group Processes &
Intergroup Relations, 21(6), 931–940.
Mazzocco, P. J., Brock, T. C., Brock, G. J., Olson, K. R., & Banaji, M. R. (2006). THE COST OF
BEING BLACK: White Americans’ perceptions and the question of reparations. Du Bois
Review: Social Science Research on Race, 3(02), 261–297.
McCall, G. J., & Simmons, J. L. (1978). Identities and interactions. Free Press.
McCarthy, T. (2002). Vergangenheitsbewältigung in the USA: On the politics of the memory of
slavery. Political Theory, 30(5), 623–648.
McCarthy, T. (2004). Coming to terms with our past, Part II: On the morality and politics of
reparations for slavery. Political Theory, 32(6), 750–772.
Mead, G. H. (1934). Mind, self, and society. University of Chicago Press.
Mutz, D. C. (2018). Status threat, not economic hardship, explains the 2016 Presidential Vote.
Proceedings of the National Academy of Sciences, 115(19), 4330–4339.
Ocbazghi, E. (2017). A Georgetown sociology professor has a radical idea for reparations that
doesn’t involve the government. Business Insider. Retrieved August 24, 2018, from https://
www.businessinsider.com/georgetown-professor-radical-idea-reparations-government-ira-indi
vidual-account-race-slavery-politics-2017-4
Phinney, J. S. (1996). When we talk about American ethnic groups, what do we mean? American
Psychologist, 51, 918–927.
Powell, A. A., Branscombe, N. R., & Schmitt, M. T. (2005). Inequality as ingroup privilege or
outgroup disadvantage: the impact of group focus on collective guilt and interracial attitudes.
Personality and Social Psychology Bulletin, 31(4), 508–521.
Reichelmann, A. V. (2020). Collective threat: conceptualizing Blumer’s threat as a collective
emotion. Sociological Inquiry.
Reichelmann, A. V. (2018). When you are a racist, theoretically speaking: empirically demonstrat-
ing Blumer’s group position theory. In J. H. Michalski (Ed.), Sociological theory, methods, and
perspectives (pp. 31–56). Nova Science Publishers.
Rosenberg, M. (1981). The self-concept: social product and social force. In M. Rosenberg & R. H.
Turner (Eds.), Social psychology: sociological perspectives (pp. 593–624). Basic Books.
Rowe, W., Bennett, S. K., & Atkinson, D. R. (1994). White racial identity models: a critique and
alternative proposal. The Counseling Psychologist, 22(1), 129–146.
Rowley, S. J., Sellers, R. M., Chavous, T. M., & Smith, M. (1998). The relationship between racial
identity and self-esteem in African American college and high school students. Journal of
Sanders Thompson, V. L. (1999). Variables affecting racial-identity salience among African
Americans. The Journal of Social Psychology, 139(6), 748–761.
Sellers, R. M., Smith, M. A., Nicole Shelton, J., Rowley, S. A. J., & Cahvous, T. M. (1998).
Multidimensional model of racial identity: a reconceptualization of African American racial
identity. Personality and Social Psychology Review, 2(1), 18–39.
Simi, P., & Futrell, R. (2010). American Swastika: Inside the white power movement’s hidden
spaces of hate. Rowman & Littlefield.
Simmons, A., & Bobo, L. (2015). Can non-full-probability internet surveys yield useful data? A
comparison with full-probability face-to-face surveys in the domain of race and social inequality
attitudes. Sociological Methodology, 45(1), 357–387.
Smith, J. R., & Hogg, M. A. (2008). Social identity and attitudes. In W. Crano & R. Prislin (Eds.),
Attitudes and attitude change (pp. 337–360). Psychology Press.
Quarterly, 63(3), 284–297.
Stets, J. E., & Fares, P. (2019). The effects of race/ethnicity and racial ethnic identification on
Stryker, S. ([1980]2002). Symbolic interactionism. Menlo Park, CA: The Benjamin/Cummings
Publishing Company, Inc.
Stryker, S., & Burke, P. (2000). The past, present, and future of an identity theory. Social
Stryker, S., & Serpe, R. T. (1994). Identity salience and psychological centrality: equivalent,
Sullivan, J. M., & Ghara, A. (2015). Racial identity and intergroup attitudes: a multiracial youth
analysis. Social Science Quarterly, 96(1), 261–272.
Tuch, S. A. (1987). Urbanism, region, and tolerance revisited: the case of racial prejudice. American
Sociological Review, 52(4), 504–510.
Tuch, S. A., & Hughes, M. (2011). Whites’ racial policy attitudes in the twenty- first century: the
continuing significance of racial resentment. The ANNALS of the American Academy of Political
and Social Science, 634(1), 134–152.
Wilson, D. C., & Davis, D. W. (2011). Reexamining racial resentment: conceptualization and
content. The ANNALS of the American Academy of Political and Social Science, 634(1),
117–133.
Wong, C., & Cho, G. E. (2005). Two-headed coins or kandinskys: white racial identification.
Political Psychology, 26(5), 699–720.
Chapter 6
Public Confederate Monuments and Racial
Identity among White Americans
Ryan D. Talbert and C. André Christie-Mizell
The U.S. landscape remains littered with iconography that romanticizes the short-
lived Confederate States of America. Over 1700 statues, plaques, flags, schools, and
roads paying homage to the Confederacy and its officials occupied public spaces
prior to the 2015 Charleston church shooting in which nine African American
congregants were murdered by a white supremacist (Southern Poverty Law Center
[SPLC], 2019). The placement of monuments also transcends regional boundaries.
More than half of all U.S. states had at least one public Confederate monument prior
to 2015 (SPLC, 2019). Many Confederate monuments were installed as a part of
reconstructing the narrative of the Civil War to commend the lost cause—which is
the myth that the Confederacy’s secession was just and necessary to defend state
autonomy from northern aggression (Blight, 2001). Conversely, white supremacy
was key to the installation of Confederate monuments and remains central to their
defense (Blight, 2001; Coski, 2005; Mask, 2020). Despite the historical nationwide
presence of Confederate monuments, research is only beginning to examine whether
and how Confederate relics are consequential for contemporary social outcomes (see
e.g., O’Connell, 2020). This study extends elements of identity theory to examine
whether the presence of state and/or county Confederate monuments enhances the
racial identity of white Americans.
Identity theory derives from the symbolic interactionist tradition and explains that
society reciprocally impacts the self and shapes social behavior (Stryker, 1980).
Individuals have multiple identities or sets of meanings that define social role
occupancy or group membership (Burke & Stets, 2009; Stryker, 1980; Stryker &
R. D. Talbert (*)
Department of Sociology, University of Connecticut, Storrs, CT, USA
C. A. Christie-Mizell
Department of Sociology, Vanderbilt University, Nashville, TN, USA
112 R. D. Talbert and C. A. Christie-Mizell
Burke, 2000). According to identity theory, identification is the process of classify-

ing or naming the self with implicit and explicit comparisons to other categories
(McCall & Simmons, 1978). The process of identification forms one’s identity, and
racial group identification is one expression of social identity. The present study
relies on the concept of identity prominence—or an individual’s idealized version of
the self (McCall & Simmons, 1978). More prominent identities are important
because they tend to guide social behavior to a greater extent (Brenner et al.,
2014; Stryker & Serpe, 1994). This study draws on identity theory to examine the
prominence of whiteness to white Americans.
The present study contributes to the research literature in at least two important
ways. First, while a strong racial group identity is commonly associated with
positive outcomes for black Americans (Christie-Mizell et al., 2017; Ida &
Christie-Mizell, 2012; LaVeist et al., 2001), fewer studies have examined factors
contributing to racial identity among white Americans. White racial identity is
complex and differs by relation, social context, and situation (Hughey, 2012;
McDermott & Samson, 2005). Research has shown that white individuals are
typically less conscious of their own racial identity (Lewis, 2004; McDermott,
2015). In turn, this invisibility or lack of recognition makes it more difficult for
whites to understand the ways that they are advantaged by being white (Bonilla-
Silva, 2017; Hartmann et al., 2009). Moreover, theoretical examinations of white-
ness generally outnumber empirical analyses of white racial identity (McDermott &
Samson, 2005). We extend the research literature by relying on a nationally repre-
sentative sample of white Americans to investigate variation in the importance that
whites place on their racial identity.
Second, this study fills a gap in the research literature by examining whether
cultural artifacts are associated with the strength of a social identity. Cultural artifacts
include a wide array of physical objects (e.g., statues, buildings, architecture), artistic
outputs (e.g., songs, books, paintings), tools, and technologies (Hunzaker &
Valentino, 2019; Swidler, 1986, 2001; Taylor et al., 2019). Such objects are com-
ponents of the meaning making process that enable individuals to make sense of the
world and become internal markers of identity. These artifacts not only connect
people to their history but also represent constant or current reminders that connect
people to each other. For example, well-known monuments such as the Statue of
Liberty may evoke national and individual pride and a view of the U.S. as the land of
opportunity for all (see e.g., Maddern, 2004). In other words, these perceptions and
thoughts associated with the Statue of Liberty cue an individual’s identity (Hunting-
ton, 2016; Saad et al., 2013; Stovall, 2018). Existing work indicates that Confederate
monuments may also play a powerful role in prompting identity (Forest & Johnson,
2019; Johnson, 2002; Reich, 2020). There is reason to expect that Confederate
monuments factor into white racial identity because they signify state-sanctioned
boundary maintenance and represent observable cues of white supremacy (Ehrlinger
et al., 2011; O’Connell, 2020; Strother et al., 2017; Talbert & Patterson, 2020). In
what follows, we present our theoretical framework and offer a short historical
overview of Confederate monuments.
6 Public Confederate Monuments and Racial Identity among White Americans 113
Background and Theory
Identity Theory and Whiteness
A key assertion of identity theory is that society shapes the individual or self and this
influence is seen through social behavior (Mead, 1934). Individuals have as many
identities as they have organized sets of role relationships to which they are linked
(Burke & Stets, 2009; Stryker, 1980; Stryker & Burke, 2000). One’s identity is
formed via identification or “the process of seeing oneself as a member of a group
and incorporating into the self all the meanings that characterize membership into the
group” (Stets & Burke, 2000, p. 224). Racial identity is one expression of social
group identification and has been the focus of much research for racial minority
groups (Christie-Mizell et al., 2017; Dagadu & Christie-Mizell, 2014; Ida &
Christie-Mizell, 2012; LaVeist et al., 2001). However, less is known about racial
identity among white Americans.
In this research, we focus on factors that shape the degree to which whites view
their racial identity as prominent. Identity theory proposes that identities are
arranged hierarchically into prominence and salience hierarchies. The relative
salience of an identity specifies the probability that it will be situationally invoked
while identity prominence—also referred to as psychological centrality—represents
the relative importance of an identity (Ervin & Stryker, 2001; Rosenberg, 1979; Stets
& Serpe, 2013; Stryker & Serpe, 1994). Individuals are more likely to have identities
high in the prominence hierarchy when they are committed to the identity and when
they receive support or rewards for the identity (Stets & Biga, 2003; Stets & Serpe,
2013). According to Ervin and Stryker (2001), individuals are aware of the level of
an identity’s prominence within their conception of themselves. Examining the
prominence of whiteness is important because prominent identities impact one’s
views, behavior, and awareness of social issues (Bonilla-Silva et al., 2006; Stets &
Biga, 2003).
In the social creation of racial and ethnic groups in the U.S., whiteness is atop the
status hierarchy and is associated with a host of social, political, and economic
benefits (Bonilla-Silva, 2017). Yet, research indicates that whiteness generally
remains an unmarked category, and whites do not often think of themselves as a
racial group or as benefitting from their whiteness (Frankenberg, 1997; Hartmann
et al., 2009; McDermott, 2015). Stets and Fares (2019, p. 3) point out, “[t]he strength
of identification with one’s social group is important for generating feelings of
belonging. How strongly individuals identify with their racial group addresses the
extent to which people’s self-views are intimately tied to their race.”
Helms’ (1997) theory of white racial identity posits that racial identity progresses
from obliviousness of one’s whiteness to the internalization of a “non-racist” belief
system. In other words, Helms theorized that whites generally learn about their
whiteness and the implications of it (Helms, 1990, 2016). Still, white racial identity
is complex, and whites may attach importance to their whiteness for different
reasons. Some whites may place importance on being white due to nationalist
views or antiracist motivations (Croll, 2007; Hughey, 2010, 2012; Lewis, 2004).
In other words, the racialization experience of whites is not identical across individ-
uals and much depends on the process of identification in which people seek to find
ways to positively define their ingroup (Kiecolt & Hughes, 2017; Lewis, 2004; Stets
& Fares, 2019).
Research on identification focuses on the roles that stereotypes, beliefs, semiotics,
and discrimination play in the relationship between social groups (Futrell & Simi,
2004; Kiecolt & Hughes, 2017; Stets & Fares, 2019). White Americans who identify
whiteness as prominent to their identity are more likely to support policies that
defend the inequitable distribution of resources, perceive greater threat from other
racial groups, and express greater distrust of others (Major et al., 2018; Stets & Fares,
2019). Germane to the current study, whites who hold negative views of black
Americans are more likely to approve of the public display of Confederate monu-
ments (Talbert, 2017; Talbert & Patterson, 2020).
Confederate symbols promote intergroup conflict and maintain racial divisions in
the U.S. by representing observable cues of white supremacy and the denigration of
black Americans. Building on prior research, we investigate how cultural artifacts in
the form of public Confederate monuments impact the construction and internalized
importance of racial group identity for white Americans. The following section
provides background on the installation of Confederate monuments and further
explains how monuments may factor into whites’ racial identity according to identity
theory.
Confederate Monuments and the Prominence of Whiteness
Cultural objects represent “shared significance embodied in form” or tangible

expressions of social meaning (Swidler, 1986, p. 3). Cultural objects can take the
form of practices, schemas, and artifacts among a host of other forms (Crane, 2000;
Griswold, 1987; Hunzaker & Valentino, 2019; Taylor et al., 2019; Winchester,
2017). Confederate monuments, as cultural objects, impact racial identity by
reinforcing symbolic group boundaries for white Americans. The dedication of
Confederate monuments peaked primarily in two periods and illustrates boundary
maintenance. The first peak arose during the installation of Jim Crow laws and the
second peak occurred during the modern civil rights movement (Blight, 2001; Mask,
2020). The placement of monuments on public property indicates that the govern-
ment unambiguously supported the racist messages represented by these objects. By
the middle of the twentieth century, every state capitol building in the U.S. South had
at least one monument installed on, within, or surrounding it (Coski, 2005; Equal
Justice Initiative, 2018).
What do these monuments or objects mean for racial group identity? Evidence
indicates that Confederate symbols matter. For instance, exposure to the Confederate
flag leads to whites expressing more stereotypical views of black Americans
(Ehrlinger et al., 2011). Reciprocally, antiblack stereotypes and views are a key
predictor of whites’ support for Confederate monuments such as the Confederate
flag (Clark 1997; Reingold & Wike, 1998; Orey, 2004; Strother et al., 2017; Talbert,
2017; Talbert & Patterson, 2020; Woliver et al., 2001). That is, Confederate mon-
uments serve to enhance the importance that whites place on their whiteness because
these objects and symbols prime antagonism toward black Americans.
Additionally, Confederate monuments impact whites’ sense of self because they
represent observable cues of white supremacy. The Confederate States of America—
the nation memorialized by Confederate monuments—was a collection of eleven
southern states that broke away from the United States between 1860 and 1861 to
defend the institution of slavery (Blight, 2001; Coski, 2005). Scattered across the
U.S. landscape are innumerable monuments honoring the rogue nation and its key
figures. Many of these Confederate monuments remain legally protected (Coski,
2005). The ubiquity of such symbols factors into fictive narratives that the Confed-
eracy fought for “states’ rights” while sanitizing or removing the role of slavery
(Blight, 2001; Coski, 2005; SPLC, 2019). The distorted narrative shifted attention
away from losing the Civil War and instead celebrated the Confederacy’s victory
over Reconstruction (Blight, 2001). Following the Civil War’s end in 1865, states
outside of the South also aided in developing a landscape that perpetuated white
supremacy (Blight, 2001; EJI, 2018; SPLC, 2019). Taken together, Confederate
monuments and the organizations that installed them promoted the lost cause
narrative beyond geographical boundaries. Thus, monuments impact white racial
identity by helping to positively define some whites’ ingroup experiences.
Monument Context
This study focuses on monuments at two levels of geographical space: the county
and the state. We rely on these two levels of residential context for several reasons.
The county provides a relevant space of exposure because monuments sometimes
occupied town squares or county courthouse grounds (Equal Justice Initiative, 2018;
Southern Poverty Law Center, 2019). Counties are a proximate space that individ-
uals interact with regularly. Moreover, prior research on the consequences of public
Confederate monuments for racial inequality uses the county as a space of exposure
(O’Connell, 2020). We follow the literature by examining county information and
expect that county monuments will factor into the prominence of white racial
identity.
We also utilize states as spaces of exposure. State-level monuments may factor
into whites’ identity regardless of the presence of county monuments. We focus on
states since some Confederate monuments such as holidays (e.g., Nathan Bedford
Forrest day in Tennessee), state flags (e.g., Mississippi’s former flag inclusive of the
Confederate Battle Flag), or the option for commemorative license plates (e.g., the
Sons of Confederate Veterans insignia) can only be counted at the state level.
Monuments can also be large enough to span county boundaries (e.g., the naming
of bodies of water, military bases, or public parks). Confederate monuments at the
state level may be more important than county monuments because state
governments played a key role in the opposition to federal desegregation orders

(Coski, 2005). In other words, state monuments may signify state governments’
overt efforts at racial boundary maintenance beyond the potential impact of county
governments. We anticipate that state monuments will enhance whites’ racial
identity.
Residents may also be more readily familiar with their state’s monuments rather
than their county’s monuments given the ranging variability in conspicuousness,
distinctiveness, and grandness of Confederate monuments. For example, while ten of
seventy-nine counties in Kentucky had at least one Confederate monument as of
2018, residents of Kentucky are likely more familiar with the 351-foot monument in
a state park that commemorates Jefferson Davis’ birthplace—the president of the
Confederacy (Equal Justice Initiative [EJI], 2018). In addition, while white Geor-
gians in the fifty counties that have at least one monument may or may not be aware
of their presence, it is likely that residents are aware of the largest Confederate
monument in the U.S.—Stone Mountain—also known as the “Confederate Mt.
Rushmore.” Finally, state monuments could factor more strongly into whites’
identities because residents interact not just with those in their own county but
also in nearby counties. In Florida, for example, the United Daughters of the
Confederacy and Sons of Confederate Veterans created a Flags Across Florida
project that installed prominent Confederate flags alongside interstate highways
(EJI, 2018).
Additional Factors Contributing to Whiteness
Southern states have a greater number of Confederate monuments than other states,
and southern residents are socialized in a context where Confederate symbols are
likely more normalized (Cooper & Knotts, 2006; Talbert, 2017). Hence, socializa-
tion in the South may lead to more positive views of monuments among white
southerners. Indeed, critics of the removal of monuments point to an “erasure of
history.” Yet, the mythologies of the South survived decades because of its sanitized
historical narratives (Blight, 2001). Monuments also represent the denigration of
black Americans, a group that has disproportionately resided in the U.S. South
historically and contemporarily (Talbert & Patterson, 2020). In this study, we are
careful to differentiate between whites residing in former Confederate States and
elsewhere.
In this research, we incorporate other factors such as socioeconomic status (e.g.,
education, income, and employment status), gender, age, type of residential com-
munity (e.g., rural versus urban areas), and political party affiliation because all have
been shown to impact whites’ racial group identity (Croll, 2007; Hartmann et al.,
2009; Hughey, 2012; Torkelson & Hartmann, 2010). White racial identity is expe-
rienced differently based on context and other social status indicators (Hughey,
2010). Whites of lower socioeconomic standing tend to perceive greater economic
and political threat from black Americans and express greater distrust (Stets & Fares,
2019). Consequently, whites with less education and income and who are unem-
ployed place greater importance on their whiteness (Croll, 2007; Stets & Fares,
2019).
Men tend to identify more strongly with whiteness because hyper masculinity and
patriarchal gender role attitudes factor into white racial formation (Ferber, 2004).
Age also factors into racial identity with mixed evidence for the direction of the
relationship (Gallagher, 1995; Grossman & Charmaraman, 2009; McDermott,
2015). Prior research has also shown that whites in rural areas place more importance
on their whiteness than whites in suburban or urban areas due to racial segregation
and the relative likelihood of interracial contact (Croll, 2007). Additionally, we
control for political party affiliation given that Republicans are more likely to
support Confederate symbols and because political views factor into the reasoning
behind whites identifying their whiteness as important (Croll, 2007; Hughey, 2010,
2012; Talbert, 2017).
The present study also includes racial attitude measures to ensure that the
relationship between Confederate monuments and whiteness is not attributable
simply to the antiblack prejudice, perceptions of racial diversity, and white nation-
alism associated with support for Confederate symbols (Cooper & Knotts, 2006;
Ehrlinger et al., 2011; Talbert & Patterson, 2020). Racial attitudes are also important
for white racial identity (Bobo & Johnson, 2000). Specifically, research has shown
that whites’ attitudes toward African Americans factor into the importance that
whites place on their identity (Croll, 2007; Goren & Plaut, 2012). We include
measures that assess the degree to which whites support social assistance for African
Americans and whether whites perceive African Americans’ societal views as
concordant with their own. The former assesses whites’ racial policy attitudes
while the latter is a social distance measure that evaluates whether whites perceive
a shared cultural membership with African Americans (Edgell & Tranby, 2010;
Roos et al., 2019).
Moreover, research shows that white racial identity is intricately tied to whites’
views and perceptions of racial diversity in the U.S. (Goren & Plaut, 2012; Stets &
Fares, 2019). This study incorporates measures that adjust for whites’ perceptions of
racial diversity in their neighborhood as well as whether whites view racial diversity
as a positive good. Finally, the study controls for whites’ views of the United States
because the literature has found that whites may tie their racial identity to pride in the
U.S., support for English language use, and the celebration of national holidays
(Feagin, 2013; Miller, 2017).
Summary and Hypotheses
This study asked: What is the relationship between the presence of public Confed-
erate monuments and white racial identity? To address this question, we examined
the number of public Confederate monuments in one’s residential context and white
identity prominence. Based on our theoretical framework and background research,
Confederate relics factor into whiteness because they visibly cue white supremacy
and demarcate symbolic racial boundaries. We developed two hypotheses. First, we
hypothesized that having a greater number of Confederate monuments in one’s
county of residence would be associated with an increase in the prominence of
whiteness (hypothesis 1). County monuments should impact whiteness because of
their geographical centrality in towns and residents’ frequency of interaction with
them. Second, we hypothesized that there would be a positive association between
the number of Confederate monuments in one’s state and white racial prominence
(hypothesis 2). State monuments should enhance whiteness because residents tie
their identity to conspicuous markers in their state, many homages span local
boundaries, and states have more legal authority to defend monuments.
Data and Methods
Data
Data for this study comes from the American Mosaic Project (AMP), a nationally
representative telephone survey conducted in 2003 with support from the David
Edelstein Family Foundation (Hartmann et al., 2003). The survey is ideal for
answering our research question given that its intent was to identify attitudes about
diversity, racial identity, and discrimination. It has been used in several studies
identifying factors associated with white racial identity (Croll, 2007; Hartmann
et al., 2009; Torkelson & Hartmann, 2010). Additionally, the AMP data are ideal
in part because they were collected prior to several high-profile events that may have
altered whites’ perceptions of Confederate monuments and constructions of self (i.e.,
Charleston in 2015, Charlottesville in 2017, and the police killings of George Floyd
and Breonna Taylor in 2020). Given that our focus is specifically on the relationship
between public Confederate monuments and white racial identity, we limit our
sample to respondents that identified as white and did not describe themselves as
Hispanic/Latino (n ¼ 986).
Information about the location and number of Confederate monuments in
counties and states is derived from a census of public Confederate monuments
collected by the Southern Poverty Law Center (2019). Monument counts are con-
sistent with the year 2003 when the survey data were collected. Monuments removed
prior to 2003 or dedicated after 2003 are not included in counts. The Southern
Poverty Law Center’s (2019) census of monuments has also been used in recent
research (O’Connell, 2020).
Methods and Measures
Our primary analysis utilizes logistic regression to estimate the odds of white
Americans identifying whiteness as a prominent identity (Long & Freese, 2014;
Long & Mustillo, 2018; Mize, 2019). The model incorporates county and state
monuments along with covariates to estimate white racial prominence. The descrip-
tive and multivariable analyses utilize sampling weights that make results general-
izable to noninstitutionalized white adults in the U.S. Descriptive statistics for all
study variables are presented in Table 6.1.
Identity Prominence. After respondents selected a racial classification (e.g., black,
white, or other), they were asked, “How important is this identity to you”?
Responses ranged from 1 (not important at all) to 4 (very important). The question
ascertaining the importance of an identity is consistent with research measuring
identity prominence (Brenner & DeLamater, 2016; Stets & Biga, 2003; Stryker &
Serpe, 1994). Consistent with prior research, we dichotomized the measure so that
whites who identified their whiteness as very important were coded as 1 (Croll,
2007). All others were coded as 0. Approximately 37 percent of respondents
identified their whiteness as very important.
Confederate Monument Presence. Counts of Confederate monuments were
merged with respondents’ county and state of residence. Respondents had approx-
imately 1 (SD ¼ 3.68) monument in their county of residence and on average
44 (SD ¼ 91.27) monuments in their state of residence. In the multivariable model
presented in Table 6.2, we transformed monument counts by taking the natural log to
correct for skewness. Table 6.1 shows that the average number of county monuments
compared across the two categories of racial importance was not statistically differ-
ent. Whites that reported whiteness as very important had on average
54 (SD ¼ 93.25) public monuments in their state of residence, which was signifi-
cantly higher than the number of monuments in the typical person’s state that did not
identify whiteness as very important (mean ¼ 38.86; SD ¼ 88.19; difference
p < .05).
Sociodemographics. We controlled for several factors that are typically associ-
ated with white racial identity. These variables included residence in a former
Confederate state (1 ¼ yes) and socioeconomic measures such as educational
attainment, employment status, and household income. About 24 percent of respon-
dents resided in a former Confederate state (i.e., Alabama, Arkansas, Florida,
Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee,
Texas, and Virginia). Educational attainment ranged from 1 (some high school or
less) to 6 (post graduate degree) and averaged 4 (SD ¼ 2.20), or equivalent to some
college experience. Additionally, almost 70 percent of respondents were employed
(1 ¼ yes), and household income averaged 5.67 (SD ¼ 2.60), or about $50,000 to
$75,000, and ranged from 1 (less than $10,000) to 8 (greater than $100,000). Whites
who identified their whiteness as very important were more likely to reside in a
former Confederate state (30 percent versus 20 percent; difference p < .05), have
Table 6.1 Descriptive statistics for white Americans in the American Mosaic Project, 2003
(n ¼ 986)
Whiteness is very Whiteness is not
Full sample important very important
Mean/ Mean/ Mean/
Measures Proportion SD Proportion SD Proportion SD
Dependent variable
Importance of white 0.37 – – – – –
racial identity (1 ¼ very
important)
Confederate monumentsa
Count of monuments in 1.12 (3.68) 1.22 (3.54) 1.06 (3.74)
one’s county (range
0–39)
Count of monuments in 44.28 (91.27) 53.83* (95.25) 38.66 (88.19)
one’s state (range 0–256)
Sociodemographics
Former Confederate state 0.24 – 0.30* – 0.20 –
(1 ¼ yes)
Educational attainment 3.91 (2.20) 3.64* (1.77) 4.07 (1.70)
(range 1–6)
Employed (1 ¼ yes) 0.70 – 0.66 – 0.71 –
Household income level 5.67 (2.60) 5.43* (2.13) 5.82 (2.05)
(range 1–8)
Age (in years; range 45.05 (24.73) 46.00 (19.14) 44.49 (18.61)
18–91)
Female (1 ¼ yes) 0.48 – 0.49 – 0.47 –
Republican (1 ¼ yes) 0.40 – 0.43 – 0.39 –
Medium city (1 ¼ yes) 0.16 – 0.15 – 0.17 –
Suburb of a large city 0.20 – 0.17 – 0.21 –
(1 ¼ yes)
Large city (1 ¼ yes) 0.10 – 0.11 – 0.10 –
Small town or rural 0.54 – 0.58 – 0.52 –
(1 ¼ yes)
Measures of antiblack prejudice
Assistance to African 5.44 (3.38) 5.13* (2.54) 5.63 (2.53)
Americans (range 2–12,
12 ¼ supportive;
α ¼ .76)
African Americans agree 2.53 (1.16) 2.49 (0.93) 2.55 (0.87)
with one’s societal views
(range 1–4, 4 ¼ agree)
Attitudes about diversity
Diversity is a strength 0.46 – 0.38* – 0.50 –
(1 ¼ yes)
Social/cultural diversity 3.08 (1.56) 3.02 (1.22) 3.12 (1.17)
in hometown (range 1–4,
4 ¼ more)
(continued)

Whiteness is very Whiteness is not
Full sample important very important
Mean/ Mean/ Mean/
Measures Proportion SD Proportion SD Proportion SD
Attitudes about the U.S.
American culture and 3.58 (0.83) 3.68 (0.64) 3.53 (0.66)
values are a strength
(range 1–4, 4 ¼ agree)
Belief that the U.S. is a 0.33 – 0.40* – 0.30 –
white nation (1 ¼ yes)
Sample size 986 372 614
Note: Asterisks denote significant differences between people that identify whiteness as very
important compared to those who do not at p .05 (two-tailed tests)
a
Monument presence information comes from the Southern Poverty Law Center (2019)
lower levels of education on average (3.64 versus 4.07; difference p < .05), and have
lower household incomes (5.43 compared to 5.82; difference p < .05).
We also included controls for age, gender, political affiliation, and the urban/rural
status of one’s community of residence at age 16. Age was measured in years
ranging from 18 to 91 (mean ¼ 45.05; SD ¼ 24.73). Forty-eight percent of
respondents were female (1 ¼ yes), and 40 percent identified as a Republican
(1 ¼ yes). Sixteen percent of respondents resided in a medium city (1 ¼ yes) at
age 16. Twenty and 10 percent of respondents resided in the suburbs or a large city,
respectively. Residence in a small city, on a farm, or rural area was the reference
category (54 percent). There were no significant differences in these measures by
racial importance category.
Measures of Antiblack Prejudice. We incorporated two measures of antiblack
prejudice that are used in the research literature (Croll, 2007; Edgell & Tranby, 2010;
Hartmann et al., 2009; Torkelson & Hartmann, 2010). The first measure was an
index of views of policy intervention labeled Assistance to African Americans. The
measure combined three items assessing how much respondents believe that African
Americans should receive special consideration in job hiring and school admissions,
economic assistance from the government, and help from charities and other
non-profit organizations. The resulting scale ranged from 3 (less supportive) to
12 (more supportive) and had a mean of 5.44 (SD ¼ 3.38). Expressing less support
indicated higher levels of racial prejudice. The alpha reliability was .76. Whites that
identified their whiteness as very important scored an average of 5.13 (SD ¼ 2.54)
compared to those that did not (mean ¼ 5.63; SD ¼ 2.53; difference p < .05). The
second measure asked, “How much do you think [African Americans] agree with
your vision of American society”? This measure was labeled African Americans
agree with one’s societal views because answers ranged from 1 to 4 with 4 indicating
greater belief that African Americans agree with one’s vision of society
(mean ¼ 2.53; SD ¼ 1.16). Identifying greater agreement between one’s views
and the views of African Americans indicated lower levels of antiblack prejudice.
Table 6.2 Prominence of white racial identity regressed onto counts of public confederate
monuments and covariates in the American Mosaic Project, 2003 (n ¼ 986)
Importance of white racial identity
(1 ¼ very important)
Variables Logit SE OR
Confederate monumentsa
County monuments count (log) .23 (.16) .79
State monuments count (log) .24* (.10) 1.27
Sociodemographics
Former Confederate state (1 ¼ yes) .17 (.43) .84
Educational attainment .13* (.07) .88
Employed (1 ¼ yes) .01 (.22) .99
Household income level .07 (.05) .93
Age (in years) .00 (.01) 1.00
Female (1 ¼ yes) .10 (.19) .90
Republican (1 ¼ yes) .15 (.19) 1.17
Medium city (1 ¼ yes) .06 (.26) .95
Suburb of a large city (1 ¼ yes) .05 (.25) 1.05
Large city (1 ¼ yes) .10 (.28) 1.11
Measures of antiblack prejudice
Assistance to African Americans .06 (.05) .94
African Americans agree with one’s societal views .01 (.13) 1.01
Attitudes about diversity
Diversity is a strength (1 ¼ yes) .36* (.18) .70
Social/cultural diversity in hometown .20* (.09) .82
Attitudes about the U.S.
American culture and values are a strength .45* (.18) 1.57
Belief that the U.S. is a white nation (1 ¼ yes) .50** (.19) 1.65
AIC 1672.153
Note: Logged odds, standard errors (SE), and odds ratios (OR) are presented
a
Monument presence information comes from the Southern Poverty Law Center (2019)
*
p .05
**
p .01 (two-tailed tests)
Attitudes about Diversity. Studies have shown that white racial identity is linked
to attitudes about racial diversity (Goren & Plaut, 2012; Stets & Fares, 2019). We
incorporated two measures assessing whites’ views of diversity. First, respondents
answered whether they believe that diversity in the United States is a strength
(1 ¼ yes). Forty six percent of respondents viewed diversity as a strength. In
Table 6.1, a lower percentage of whites in the very important category reported
diversity as a strength (38 percent compared to 50 percent; difference p < .05).
Second, respondents were asked how much social and cultural diversity is present
within their community, and answers ranged from 1 (less diversity) to 4 (more
diversity) (mean ¼ 3.08; SD ¼ 1.56).
Attitudes about the U.S. We included two variables that measured attitudes about
the U.S. (Hartmann et al., 2009; Torkelson & Hartmann, 2010). The first item asked
the degree to which American culture and values are a strength. Answers had a mean
of 3.58 (SD ¼ 0.83) and a range from 1 (strongly disagree) to 4 (strongly agree). We
also measured whether respondents believed that the U.S. is a white nation and
whether this designation is positive (1 ¼ yes). Thirty-three percent of respondents
agreed with the compound statement. A significantly higher percentage of respon-
dents in the whiteness is very important group reported that the U.S. is a white nation
and that this designation is a good thing (40 percent versus 30 percent; difference
p < .05). We turn now to our multivariable results.
Results
Confederate Monuments and Whiteness
Table 6.2 shows results for the importance of white racial identity regressed onto
counts of public Confederate monuments and covariates. Recall that the first
hypothesis was that greater numbers of Confederate monuments in one’s county
of residence would be associated with an increase in the prominence of whiteness.
Results in Table 6.2 did not offer support for hypothesis 1. The second hypothesis
expected that there would be a positive association between the number of state
Confederate monuments and the prominence of one’s whiteness. Findings supported
hypothesis 2. The model illustrates that a unit increase in the number of state
Confederate monuments in one’s state of residence was associated with a 27 percent
increase in the odds of white racial prominence (OR ¼ 1.27; p < .05). We conducted
several sensitivity analyses to determine whether this relationship was simply a
consequence of the variable construction. Findings were substantively similar to
those presented regardless of measurement strategy including the use of quintiles of
Confederate monuments, windsorizing the highest counts, and taking the square root
of counts. We also ran generalized multilevel models with individuals nested within
states to address concerns associated with clustered data, and we found that our
results were substantively identical as those presented.
Moreover, the model demonstrates that respondents with higher levels of educa-
tion had lower odds of identifying whiteness as very important (OR ¼ .88; p < .05).
Results in Table 6.2 also show that viewing diversity as a strength (OR ¼ .70;
p < .05) and reporting greater diversity in one’s hometown (OR ¼ .82; p < .05) were
associated with lower likelihoods of identifying whiteness as very important. Con-
versely, identifying American culture and values as a strength (OR ¼ 1.57; p .05)
and believing the U.S. is a white nation (OR ¼ 1.65; p < .01) increased the odds of
valuing white racial identity.
Figure 6.1 graphs the predicted probabilities of white racial prominence derived
from estimates presented in Table 6.2. Figure 6.1 shows that the probability of
whites reporting their whiteness as important increased as the number of
.8
.7
Probability of White Racial Prominence
.6
.5
.4
.3
.2
.1
0
–2 SD –1 SD Mean +1 SD +2 SD
Number of Confederate Monuments
Fig. 6.1 Predicted probability of reporting while racial identity as very important by number of
state confederate monuments
Confederate state monuments increased. For example, a person at two standard

deviations below the mean number of state Confederate monuments in a state had
a 20 percent probability of identifying whiteness as very important. At the average
number of monuments, the probability of white racial prominence was approxi-
mately 37 percent. Finally, at two standard deviations above the mean number of
monuments, the probability increased to 58 percent.
Discussion
This study examined the association between the presence of county and state
Confederate monuments and the prominence of white racial identity. Findings
indicated that the presence of state Confederate monuments was closely tied to racial
identity for white Americans. Whites are likely most aware of pubic Confederate
monuments at the state level because these monuments are more noticeable, visible,
and more deeply connect place to constructions of the self. For instance, Georgians
are exposed to Stone Mountain, Virginians contend with the multitude of Confed-
erate monuments on Monument Avenue, and South Carolinians observed the Con-
federate flag prominently flying on their capitol grounds for over five decades
(Southern Poverty Law Center, 2019). By contrast, a typical monument in one’s
county of residence may be a generic statue of a Confederate soldier, a street name,
or plaque honoring Confederate veterans. While still impactful (see e.g., O’Connell,
2020), the presence of these symbols does not elicit the same significant affirmation
of white nationalism as the larger, more visible markers that residents identify with
their state of residence.
Additionally, states have a greater amount of resources than local governments to
install Confederate monuments and provide legal protection for them (see e.g.,
Florida [Fla. Stat. § 256:10]; Georgia [Ga. Code. Title 50-3-1]; Louisiana
[La. Rev. Stat. § 14:116]; Mississippi [Miss. Code of Laws. § 97-7-39]; South
Carolina [S.C. Code Ann. § 16-17-210]). State leaders were prominent figures in
the opposition to federal mandates to desegregate. For instance, George Wallace, the
former Governor of Alabama, infamously stood in a doorway to prevent the inte-
gration of the University of Alabama. Consequently, Confederate monuments in the
state of residence that whites reside are of greater importance for their racial identity.
Ultimately, the way that societies commemorate history legitimizes and rein-
forces the identity of some groups and dismisses others. Organizations such as the
United Daughters of the Confederacy and the Sons of Confederate Veterans raised
monuments across the United States in honor of Confederate figures, veterans, and
the lost cause. The underside of this story is that the installation of many of these
monuments coincided with the enacting of Jim Crow laws as well as the mass
movement to abolish them (Blight, 2001; Coski, 2005; Strother et al., 2017; Woliver
et al., 2001). In other words, statues promoted messages beyond the figures they
represented. These markers and symbols were an intentional plan to reconstruct the
narrative surrounding the Confederacy’s efforts in the Civil War. All the while,
Confederate monuments painted a backdrop to the further dehumanization and
disenfranchisement of African Americans. Identity theory proposes that individuals
seek ways to define their ingroup positively compared to an outgroup (Kiecolt &
Hughes, 2017; Stets & Fares, 2019). Foremost, these monuments perpetuate white
supremacy and shore up the importance of white racial identity for those who admire
and identify with symbols of white nationalism. Additionally, Confederate monu-
ments legitimize ideologies of black inferiority and promote the notion that slavery
was secondary to the Confederacy’s efforts. Together, these factors (i.e., white
supremacy and black denigration) buttress a positive group identity among some
whites.
Future research will be able to extend findings from the present study to
problematize why and how whites come to place importance on their whiteness
(see e.g., Hughey, 2010, 2012). For example, future research should investigate
potential differences in the sorts of Confederate monuments that evoke whiteness in
various ways compared to others (e.g., Confederate flags versus schools named after
Confederate officials). Moreover, the measure of public Confederate monuments
used in this study undercounts the true number of monuments across the United
States. Markers installed on private property as well as public tributes on battlefields
or in museums and cemeteries, for instance, were not included in the Southern
Poverty Law Center (2019) data. Including these additional markers in future
research could reveal an even stronger association between Confederate monuments
and identity. Finally, the survey data used in the present study were collected in
2003. Future research should investigate the role of Confederate monuments and
whiteness in the aftermath of the Charleston massacre, deadly Charlottesville pro-

tests, and police killings of Breonna Taylor and George Floyd.
Conclusion
McDermott and Samson (2005) posit that scholars cannot fully understand the
existence of racism without studying the formation, maintenance, and consequences
of white racial identity. This study shows that whites are aware of their residential
context and particularly the presence of public Confederate monuments at the state
level. In turn, the presence of Confederate monuments affirms a particular form of
whiteness; one that simultaneously celebrates the “lost cause” and denies the
existence of racism (Bonilla-Silva, 2017; McDermott, 2015). The findings presented
in this study have broad implications given the tie between racial identity and self-
concept. Some whites have tended to root their identity in a narrative of distorted
American history memorialized by Confederate monuments. Consequently, violent
backlashes occur when these monuments are threatened, banned, or removed (e.g.,
the Charlottesville protests and NASCAR’s ban in 2020). Findings from this study
also extend identity theory by showing that cultural artifacts such as statues, mon-
uments, and the commemorative naming of public spaces factor into the strength of a
social identity. Ultimately, this study shows that Confederate monuments celebrate
an ahistorical narrative around the Civil War and operate as a symbolic boundary
between whites and blacks, which serves to support intergroup conflict and increase
the importance that whites place on their whiteness.
References
Blight, D. W. (2001). Race and reunion: The Civil War in American memory. Harvard University
Press.
Bobo, L., & Johnson, D. (2000). Racial attitudes in the prismatic metropolis: Identity, stereotypes,
and perceived group competition in Los Angeles. In L. D. Bobo, M. L. Oliver, J. H. Johnson, &
A. Valenzuela (Eds.), Prismatic Metropolis: Inequality in Los Angeles (pp. 83–166). Russell
Sage Foundation.
Bonilla-Silva, E. (2017). Racism without racists: Color-blind racism and the persistence of racial
inequality in America. Rowman & Littlefield Publishers.
Bonilla-Silva, E., Goar, C., & Embrick, D. G. (2006). When whites flock together: The social
psychology of white habitus. Critical Sociology, 32, 229–253.
Brenner, P. S., & DeLamater, J. (2016). Lies, damned lies, and survey self-reports? Identity as a
cause of measurement bias. Social Psychology Quarterly, 79(4), 333–354.
Christie-Mizell, C. A., Leslie, E. T. A., & Hearne, B. N. (2017). Self-rated health, discrimination,
and racial group identity: The consequences of ethnicity and nativity among black Americans.
Journal of African American Studies, 21(4), 643–664.
Clark, J. A. (1997). Explaining elite attitudes on the Georgia flag. American Politics Quarterly, 25
(4), 482–496.
Cooper, C. A., & Knotts, H. G. (2006). Region, race, and support for the South Carolina
Confederate flag. Social Science Quarterly, 87(1), 142–154.
Coski, J. M. (2005). The Confederate battle flag: America’s most embattled symbol. Harvard
University Press.
Crane, D. (2000). Fashion and its social agendas: Class, gender, and identity in clothing.
University of Chicago Press.
Croll, P. R. (2007). Modeling determinants of white racial identity: Results from a new national
survey. Social Forces, 86(2), 613–642.
Dagadu, H. E., & Christie-Mizell, C. A. (2014). Heart trouble and racial group identity: Exploring
ethnic heterogeneity among black Americans. Race and Social Problems, 6(2), 143–160.
Edgell, P., & Tranby, E. (2010). Shared visions? Diversity and cultural membership in American
life. Social Problems, 57(2), 175–204.
Ehrlinger, J., Ashby Plant, E., Eibach, R. P., Columb, C. J., Goplen, J. L., Kunstman, J. W., & Butz,
D. A. (2011). How exposure to the Confederate flag affects willingness to vote for Barack
Obama. Political Psychology, 32(1), 131–146.
Equal Justice Initiative [EJI]. (2018). Segregation in America. Retrieved February 9, 2020, from
https://segregationinamerica.eji.org/
T. J. Owens, S. Stryker, & N. Goodman (Eds.), Extending self-esteem theory and research:
Sociological and psychological currents (pp. 29–55). Cambridge University Press.
Feagin, J. R. (2013). The white racial frame: Centuries of racial framing and counter-framing (2nd
ed.). Routledge Press.
Ferber, A. L. (2004). Home-grown hate: Gender and organized racism. Routledge Press.
Fla. Stat. § 256:10. “Mutilation of or Disrespect for Confederate Flags or Replicas.”
Forest, B., & Johnson, J. (2019). Confederate monuments and the problem of forgetting. Cultural
Geographies, 26(1), 127–131.
Frankenberg, R. (Ed.). (1997). Displacing whiteness: Essays in social and cultural criticism. Duke
University Press.
Futrell, R., & Simi, P. (2004). Free spaces, collective identity, and the persistence of U.S. white
power activism. Social Problems, 51(1), 16–42.
Ga. Code. Title 50-3-1. “Description of State Flag.”
Gallagher, C. A. (1995). White reconstruction in the university. Socialist Review, 24, 165–187.
Goren, M. J., & Plaut, V. C. (2012). Identity form matters: White racial identity and attitudes toward
diversity. Self and Identity, 11(2), 237–254.
Griswold, W. (1987). A methodological framework for the sociology of culture. Sociological
Methodology, 17, 1–35.
Grossman, J. M., & Charmaraman, L. (2009). Race, context, and privilege: White adolescents’
explanations of racial-ethnic centrality. Journal of Youth and Adolescence, 38(2), 139–152.
Hartmann, D., Gerteis, J., & Croll, P. R. (2009). An empirical assessment of whiteness theory:
Hidden from how many? Social Problems, 56(3), 403–424.
Hartmann, D., Gerteis, J., & Edgell, P. (2003). American Mosaic Project Survey, 2003. Inter-
university Consortium for Political and Social Research. Retrieved January 16, 2020, from
https://doi.org/10.3886/ICPSR28821.v1
Helms, J. E. (Ed.). (1990). Black and white racial identity: Theory, research, and practice.
Greenwood Press.
Helms, J. E. (1997). Toward a model of white racial identity development. In K. D. Arnold, P. G.
Altbach, & I. C. King (Eds.), College student development and academic life: Psychological,
intellectual, social and moral issues (pp. 33–47). Routledge Press.
Helms, J. E. (2016). Toward a theoretical explanation of the effects of race on counseling a black
and white model. The Counseling Psychologist, 12(4), 153–165.
Hughey, M. W. (2012). White bound: Nationalists, antiracists, and the shared meanings of race.
Stanford University Press.
Hughey, M. W. (2010). The (dis)similarities of white racial identities: The conceptual framework of
‘hegemonic whiteness’. Ethnic and Racial Studies, 33(8), 1289–1309.
Huntington, S. P. (2016). The crisis of national identity. Military Review, 96(6), 10–22.
Hunzaker, M. B. F., & Valentino, L. (2019). Mapping cultural schemas: From theory to method.
Ida, A. K., & Christie-Mizell, C. A. (2012). Racial group identity, psychosocial resources, and
depressive symptoms: Exploring ethnic heterogeneity among black Americans. Sociological
Focus, 45(1), 41–62.
Johnson, N. C. (2002). Mapping monuments: The shaping of public space and cultural identities.
Visual Communication, 1(3), 293–298.
Kiecolt, J. K., & Hughes, M. (2017). Racial identity and the quality of life among blacks and whites
in the U.S. Social Science Research, 67, 59–71.
La. Rev. Stat. § 14:116. “Flag Desecration.”
LaVeist, T. A., Sellers, R., & Neighbors, H. W. (2001). Perceived racism and self and system blame
attribution: Consequences for longevity. Ethnicity and Disease, 11(4), 711–721.
Lewis, A. E. (2004). ‘What group?’ Studying whites and whiteness in the era of ‘color-blindness’.
Sociological Theory, 22(4), 623–646.
Long, J. S., & Freese, J. (2014). Regression models for categorical dependent variables using stata.
Stata Press.
Long, J. S., & Mustillo, S. A. (2018). Using predictions and marginal effects to compare groups in
regression models for binary outcomes. Sociological Methods and Research, 1–37.
Maddern, J. (2004). Huddled masses yearning to buy postcards: The politics of producing heritage
at the Statue of Liberty-Ellis Island National Monument. Current Issues in Tourism, 7(4–5),
303–314.
Major, B., Blodorn, A., & Blascovich, G. M. (2018). The threat of increasing diversity: Why many
white Americans support Trump in the 2016 Presidential Election. Group Processes &
Intergroup Relations, 21(6), 931–940.
Mask, D. (2020). The address book: What street addresses reveal about identity, race, wealth, and
power. St. Martin’s Publishing Group.
McDermott, M. (2015). Color-blind and color-visible identity among American whites. American
Behavioral Scientist, 59(11), 1452–1473.
McDermott, M., & Samson, F. L. (2005). White racial and ethnic identity in the United States.
Annual Review of Sociology, 31(1), 245–261.
Mead, G. H. (1934). Mind, self, and society: From the standpoint of a social behaviorist. University
of Chicago Press.
Miller, P. K. (2017). How white college students enact whiteness: An investigation of the attitudi-
nal, behavioural and cultural components that comprise white racial identity at a historically
white college. Whiteness and Education, 2(2), 92–111.
Miss. Code of Laws. § 97-7-39. “Flags: Desecration of National or State Flag Prohibited.”
Mize, T. (2019). Best practices for estimating, interpreting, and presenting nonlinear interaction
effects. Sociological Science, 6, 81–117.
O’Connell, H. A. (2020). Monuments outlive history: Confederate monuments, the legacy of
slavery, and black-white inequality. Ethnic and Racial Studies, 43(3), 460–478.
Orey, B. D. (2004). White racial attitudes and support for the Mississippi state flag. American
Politics Research, 32(1), 102–116.
Reich, G. A. (2020). Monumental refraction: Monuments, identity, and historical consciousness.
Historical Encounters, 7(1), 1–23.
Reingold, B., & Wike, R. S. (1998). Confederate symbols, southern identity, and racial attitudes:
The case of the Georgia state flag. Social Science Quarterly, 79(3), 568–580.
Roos, J. M., Hughes, M., & Reichelmann, A. V. (2019). A puzzle of racial attitudes: A measure-
ment analysis of racial attitudes and policy indicators. Socius, 5, 1–14.
S.C. Code Ann. § 16-17-210. “Desecration or Mutilation of Flags.”
Saad, C. S., Damian, R. I., Benet-Martínez, V., Moons, W. G., & Robins, R. W. (2013). Multicul-
turalism and creativity: Effects of cultural context, bicultural identity, and ideational fluency.
Social Psychological and Personality Science, 4(3), 369–375.
Southern Poverty Law Center [SPLC]. (2019). Whose heritage? Public symbols of the confederacy.
Retrieved January 16, 2020, from https://www.splcenter.org/20190201/whose-heritage-public-
symbols-confederacy
Stets, J. E., & Biga, C. F. (2003). Bringing identity theory into environmental sociology. Socio-
logical Theory, 21(4), 398–423.
Quarterly, 63(3), 224–237.
Stets, J. E., & Fares, P. (2019). The effects of race/ethnicity and racial/ethnic identification on
social psychology (pp. 31–60). Springer Netherlands.
Stovall, T. (2018). White freedom and the lady of liberty. The American Historical Review, 123(1),
1–27.
Strother, L., Piston, S., & Ogorzalek, T. (2017). Pride or prejudice? Racial prejudice, southern
heritage, and white support for the Confederate battle flag. Du Bois Review: Social Science
Research on Race, 14(1), 295–323.
Stryker, S. (1980). Symbolic interactionism: A social structural version. Benjamin/Cummings
Publishing Company.
Stryker, S., & Burke, P. J. (2000). The past, present, and future of an identity theory. Social
Swidler, A. (1986). Culture in action: Symbols and strategies. American Sociological Review, 51
(2), 273–286.
Swidler, A. (2001). Talk of love: How culture matters. University of Chicago Press.
Talbert, R. D. (2017). Culture and the Confederate flag: Attitudes toward a divisive symbol.
Sociology Compass, 11(2), 1–10.
Talbert, R. D., & Patterson, E. J. (2020). Racial stratification and the Confederate flag: Comparing
four perspectives to explain flag support. Race and Social Problems, 12(3), 233–245.
Taylor, M. A., Stoltz, D. S., & McDonnell, T. E. (2019). Binding significance to form: Cultural
objects, neural binding, and cultural change. Poetics, 73, 1–16.
Torkelson, J., & Hartmann, D. (2010). White ethnicity in twenty-first-century America: Findings
from a new national survey. Ethnic and Racial Studies, 33(8), 1310–1331.
Winchester, D. (2017). “A part of who I am”: Material objects as “plot devices” in the formation of
religious selves. Journal for the Scientific Study of Religion, 56(1), 83–103.
Woliver, L. R., Ledford, A. D., & Dolan, C. J. (2001). The South Carolina Confederate flag: The
politics of race and citizenship. Politics & Policy, 29(4), 708–730.
Part II
Family, Religion, and Work
Chapter 7
The Mother Identity and Self-Esteem
Jan E. Stets and Juha Lee
The Mother Identity and Self-Esteem
Self-esteem is the evaluation of oneself in positive or negative terms (Rosenberg,

1979). For half a century, scholars have studied self-esteem and found that high self-
esteem produces positive outcomes such as greater sociability and assertiveness, a
willingness to help others, and feeling valued in interpersonal relationships, while,
low self-esteem generates negative outcomes such as a greater likelihood of
experiencing negative emotions and increased involvement in antisocial behaviors
such as substance abuse, delinquency, and crime (Leary & MacDonald, 2003). In
this study, we theoretically advance research on self-esteem by examining whether
verification of the mother identity may be related to different dimensions of self-
esteem: efficacy and worth. Specifically, we replicate and expand on earlier research
by examining whether the verification of an identity may be related to multiple
dimensions of self-esteem, relying on the idea that an identity and corresponding
self-esteem dimensions may share meanings (Stets & Burke, 2014b).
Researchers have made two arguments on the relationship between the verifica-
tion of identities and self-esteem (Stets & Burke, 2014b). First, the verification of
different identity bases result in different self-esteem outcomes. One basis of iden-
tities, role identities, emphasize meanings related to one’s performance in a role.
When the meaning of people’s behavior corresponds to internalized role identity
meanings, individuals feel a sense of competence in successfully performing their
role. Thus, when role identities are verified, it increases efficacy-based esteem.
J. E. Stets (*) · J. Lee

134 J. E. Stets and J. Lee
Another basis of identities, group or social identities that offer acceptance and
inclusion, communicate that individuals are valued and worthwhile members in
society. When group and social identities are verified, it increases worth-based
esteem. Second, while different identity bases are related to different self-esteem
dimensions, what ultimately connects an identity to a self-esteem dimension is the
meanings that are shared with the identity and the self-esteem dimension. An identity
base may be related to another self-esteem dimension if the meanings of the identity
are shared with that other esteem dimension.
To illustrate the above two points, we use the mother identity as an example. The
mother role identity contains meanings of “care” and “nurturance” (Arendell, 2000;
Simon, 1997), and behaviors that reflect these meanings would include feeding,
clothing, educating, and spending leisure time with one’s children. When children
communicate that they feel cared for and nurtured by their mothers, this raises
mothers’ feelings of competence, and mothers thus experience efficacy-based
esteem.
The mother identity also carries the meaning of being valued in society. Indeed,
research reveals that most women, including those across different social classes and
racial/ethnic groups, have positive attitudes toward being a mother (Tichenor et al.,
2017). Since group and social identities share meanings of being socially accepted in
society, when mothers join a “neighborhood of moms” to offer and receive childcare
and share in the joys and struggles of parenthood, they likely share the meaning of
being socially valued and accepted in society given their status as mothers, this gets
reaffirmed in the group, and both should generate feelings of self-worth. Thus, the
mother identity shares meanings not only with the efficacy-based dimension of
esteem of doing care and nurturance, but also the worth-based dimension of esteem
of being a valued member in society.
We test these two arguments by examining the mother identity for a primarily
Hispanic, low-income sample of women. This moves the research beyond the
college samples of earlier research (Stets & Burke, 2014b) and examines an identity
that is important to women in society (McQuillan et al., 2008). Hispanics are an
important group to examine because they are one of the fastest growing minority
groups in the U.S.; by 2060, nearly a third of U.S. residents will be Hispanic
compared to 16% today (United States Census Bureau, 2012). Though this sample
is of low income, primarily minority women, it provides insight as to how mothering
is experienced among the most disadvantaged in our society who rest at the
intersection of a system of inequality based on gender, class, and race (Ridgeway
& Kricheli-Katz, 2013).
In identity theory, because the mother identity has been studied as a role identity
(Cast, 2004; Ellestad & Stets, 1998; Nuttbrock & Freudiger, 1991; Simon, 1992), we
expect that verification of the mother identity is associated with efficacy-based
esteem. If a woman performs her mother role in a way that carries meanings
consistent with her mother identity meanings, she should be more likely to experi-
ence identity verification and feel competent and efficacious.
However, verification of the mother identity simultaneously may shore up feel-
ings of acceptance since it is a valued status shared in society (McQuillan et al.,
2008), including in one’s family (Stets & Tsushima, 2001). Being a mother also
7 The Mother Identity and Self-Esteem 135
carries value because it is associated with being female and feminine (Gillespie,
2003). Thus, verification of the mother identity at the micro level may help integrate
women at the meso level (in families) and at the macro level (as a female) in society
(Turner, 2016) and be associated with worth-based esteem. Essentially, mothers may
derive positive feelings about themselves based not only on what they do but who
they are.
While we investigate the verification of the mother role identity in terms of
efficacy-based esteem and worth-based esteem, we also examine other dimensions
of an identity that may be related to self-esteem including the prominence and
salience of the mother identity. Prior research has established a positive relationship
between identity prominence, or the importance of an identity and self-esteem
(Rosenberg et al., 1995), even establishing that multiple prominent role identities
are related to both efficacy and worth-based esteem (Serpe et al., 2019). Recently,
researchers examined the simultaneous influence of prominence, salience, and
verification of the parent identity on these different bases of esteem (Markowski &
Serpe, 2018). While they found prominence to be positively related to worth and
efficacy-based esteem, only salience was positively associated with worth-based
esteem. However, the parent identity was not measured as a set of meanings, but how
positive respondents viewed themselves as a parent. Thus, the identity verification
results cannot be compared to the current results.
We use a sample of mothers from an Early Head Start program in a southwestern
town during 2017–18. There were 100 mothers who participated in a survey.
They were queried on a number of issues such as what kind of mother they thought
they were on the basis of different meanings of the mother role that were presented to
them, how they thought others saw them as a mother, how prominent and salient was
being a mother to them, and how they felt as a mother. We first discuss the different
dimensions of self-esteem, and then we link them to the mother identity.
Theory
The idea that self-esteem is comprised of at least two dimensions, a self-efficacy and
self-worth dimension, is not new (Cast & Burke, 2002; Franks & Marolla, 1976;
Gecas & Schwalbe, 1986; Stets & Burke, 2014b). Self-efficacy is the extent to which
people think they have an effect on their environment (Gecas, 1989); it is analogous
to having mastery or control over the forces that affect people’s lives (Pearlin et al.,
1981). Self-worth is the degree to which individuals feel they are valued and
accepted (Rosenberg et al., 1995).
Both self-efficacy and self-worth are rooted in different motives. Self-efficacy is
rooted in the desire to have agency (known as the agency motive (Swann & Bosson,
2010)) or being effective in the world (identified as the self-efficacy motive (Gecas,
1991)). Individuals want to feel competent or see that they can bring about particular
outcomes (Bosson & Swann, 1999; Tafarodi & Swann, 1995). Self-worth is rooted
in the desire for people to feel good about themselves (recognized as the self-esteem
motive (Gecas, 1991) or self-enhancement motive (Leary, 2007)). Part of this

positive feeling stems from people feeling that they have social worth: that others
accept them, and they belong. This satisfies what some have labeled people’s desire
for communion and connection with others (Swann & Bosson, 2010), prompting a
feeling of self-liking or positive evaluation of people’s personal worth given others’
responses to them (Bosson & Swann, 1999; Tafarodi and Swann, 1995).
In identity theory, self-esteem is an outcome of the identity verification process.
The identity verification occurs when individuals act in ways so that others in the
situation see them in the same way that they see themselves. When others react to
how individuals are behaving in a situation, individuals perceive these meanings.
This is the reflected appraisal process or how individuals think that others see them in
a situation. Individuals compare the reflected appraisal meanings with the meanings
set in their identity, and they determine whether there is a discrepancy.
A large discrepancy between self-in-situation meanings and identity meanings is
identity non-verification. People will experience negative emotions (Burke &
Cerven, 2019; Stets & Burke, 2014a; Stets et al., 2018; Stets & Carter, 2011,
2012), and there will be a decline in self-esteem (Burke & Harrod, 2005; Burke &
Stets, 1999; Cast, 2004; Cast & Burke, 2002; Stets & Harrod, 2004). Individuals
respond by making adjustments in how they behave and how they see themselves,
and behavioral and cognitive responses work in concert to reduce the discrepancy in
meanings to a minimum or near zero difference (Stets et al., 2020). The goal is to
achieve identity verification or a match between reflected appraisal meanings and
identity meanings. In identity verification, people experience positive emotions and
an increase in self-esteem.
Recently, researchers theorized and showed that the verification of different bases
of an identity are associated with different dimensions of self-esteem (Stets & Burke,
2014b). One identity base is the set of meanings that people attribute to themselves
when they take on roles that are attached to positions in society such as friend,
worker, spouse, and parent (Burke & Harrod, 2005; Burke & Stets, 1999; Cast,
2004; Cast & Burke, 2002; Stets & Harrod, 2004). In any role-based identity such as
the mother identity, individuals act to fulfill the expectations of the role, and they
coordinate their role as it relates to other roles in the group (Burke & Stets, 2009).
Thus, in the mother role, women coordinate this role as it relates to other roles in the
family such as the roles of child and father.
The role identity of mother is the meanings women attribute to themselves while
in the mother role such as being “involved,” “giving love” and “guiding and
teaching” (Simon, 1997). These meanings are consistent with the prevailing ideol-
ogy of motherhood in American culture, which entails intensive mothering that is
“child-centered, expert-guided, emotionally absorbing, labor-intensive, and finan-
cially expensive” (Hays, 1996, p. 8). The idea of intensiveness means that mothers
are spending an extraordinary amount of time, energy, and money in meeting a
child’s needs. Hays admits that while not all mothers practice intensive mothering, it
is the hegemonic standard for raising a child. It is the dominant, normative standard
by which mothering is judged in society.
Others have echoed Hays ideology (Douglas & Michaels, 2004; Freely, 2000;
Warner, 2006). For example, some discuss the “new momism” as an “insistence that
no woman is truly complete or fulfilled unless she has kids, that women remain the
best primary caretakers of children, and that to be a remotely decent mother, a
woman has to devote her entire physical, psychological, emotional, and intellectual
being, 24/7, to her children” (Douglas & Michaels, 2004, p. 4). Douglas and
Michaels recognize that the cultural image of motherhood is ideal, and though it is
impossible to achieve, it nevertheless guides how women view their mother identity.
Intensive parenting becomes a cultural script which mothers are expected to play
out to the best of their ability. The focus is on “doing” or engaging the maternal
practice of care and devotion to their offspring that fosters children’s intellectual,
social, emotional, and moral development (Arendell, 2000). While mothers’ con-
cerns for their children may differ by class, for example, middle and upper-class
mothers may worry that their children will not enjoy a middle and upper-class status
given rising income inequality, and lower-class mothers may focus on protecting
their children and ensuring they are safe from harmful influences (Nomaguchi &
Milkie, 2020), all mothers share the all-encompassing expectation of being highly
accessible, responsible, and devoted to their children. Intensive mothering is con-
sistent with a family model in which the husband is the full-time wage earner, and
the wife is the stay-at-home caretaker. Working in the labor force sets up an inherent
contradiction, and working women resolve the contradiction by altering their moth-
ering ideology to reinforce their work decision (Johnston & Swanson, 2006).
Another identity base in which we consider motherhood is as a group identity.
While the mother identity is about “doing,” enacting the role behaviors that carry
meanings associated with meeting the needs of offspring, the mother identity also is
a group identity and is about “being.” For mothers, group membership occurs in
meso level groups such as one’s family, as well as membership in more distal, macro
level categories such as being a woman in society.
The family typically is characterized by close bonds, solidarity, and intimacy. It is
a domain that serves as a deep source of meaning for individuals; it is a primary
source of happiness, especially the interpersonal bonds with one’s partner and
children (Delle Fave et al., 2011). Family meanings often entail “commitment,”
“support,” and “loyalty.” Parents seek approval from their children that they are
good parents, and children seek approval from their parents that they are worthy of
their parents love and attention (Stets & Tsushima, 2001). Through love and
devotion that mothers both give and receive, they may feel that they belong, and
that they are valued. This should foster feeling accepted and feeling worthwhile.
Motherhood also is a valued status in our society, and it is tied to the categorical
membership of being a woman. As a valued status, for example, when researchers
examine the meanings of different roles in society, the meaning of mother, on scale
that runs from +3 to –3, is very high (2.80) on how it is evaluated in society (Francis
& Heis, 2006).
Motherhood is linked to femininity (Chodorow, 1989). To be feminine is to be a
mother. In fact, a women’s gender identity as female is strengthened through
mothering (McMahon, 1995). For example, research reveals that the birth of a
child leads women to view themselves as more feminine (Burke & Cast, 1997).
More recent work finds that with the birth of a child, women become more support-
ive of traditional gender role stereotypes, and their attitudes reflect more traditional
gendered behavior such as spending more time in childcare and less time in the labor
force (Endendijk et al., 2018). Following the birth of a child, women become more
supportive of motherhood as women’s most important role in life (Baxter
et al., 2015).
Thus, belongingness may not only be felt within one’s immediate family, but it
also may be an identity that helps women feel integrated in society. They are
fulfilling the cultural expectations as to what it means to be a woman—to also be a
mother. In this way, verification of the mother identity also may be associated with
worth-based esteem.
We cannot easily separate a role from group membership either analytically or
empirically because one who occupies a role may simultaneously belong to a group
within which that role is defined (Burke & Stets, 2009; Stets & Burke, 2000). Roles
are performed within groups, and groups flourish because tasks within the group are
accomplished. For example, women “do mothering” within the family, thus the
mother identity carries meanings that signify the various roles that they enact, but it
also carries meanings that represent being a member of a family or being a woman in
society. Thus, role identities and group and social identities may be relevant at the
same time in situations, influencing the meanings of behavior and impacting multi-
ple bases of self-esteem in response to the identity verification process.
As discussed in earlier theorizing (Stets & Burke, 2000), identities like the mother
identity that are both role-based and group-based activate the organic and mechan-
ical forms of social integration (Durkheim, 1893 (1984)). Organic integration,
common in modern societies, emerges from a division of labor in which individuals
are dependent upon each other to perform specific roles and tasks (Burke & Stets,
2009; Stets & Burke, 2000). In the family, mothers often perform roles that are
distinct from what fathers do such as spending more time in demanding activities
like childcare, cooking and cleaning, while fathers spend more time with children in
play and leisure activities (Musick et al., 2016). Roles focus on the individual that
carries out specific performances (Thoits & Virshup, 1997).
Mechanical integration, common in pre-industrial societies such as hunting and
gathering societies, binds individuals to each other based on similarities and shared
experiences. Differences among individuals are minimized. Individuals act in a
parallel fashion, and each person often is interchangeable in accomplishing the
groups’ goals. As applied to the family, members are biologically similar, and
they encounter a set of common experiences living within the same household.
All work in a similar manner to advance their family and a feeling of a “we”
develops (Thoits & Virshup, 1997). Ultimately, both the organic/role and mechan-
ical/group forms of integration that are tied to different identities link the self to
society in different but related ways. Role-based identities form the integrative
function of providing distinctiveness for individuals in groups, given the different
positions and corresponding roles that each holds in the group. Group-based
identities provide the integrative function of inclusion, ensuring that people do not
feel isolated.
The mother identity is a good example of an identity that inherently provides
organic and mechanical integration. Whether we attend to the identity as role-based
or group-based depends upon which psychological state gets activated in the situa-
tion (Thoits & Virshup, 1997). If the meanings in the situation predominately signal
“doing” the mother identity, as when a child cries because he is hungry, role identity
meanings will get activated. If the situational meanings primarily cue mothers as part
of a family unit, as when she introduces to her local minister at her church, her
family, group meanings associated with the mother identity will get activated.
Alternatively, meanings in the situation may alternate between activation of the
mother role identity and mother group identity. Still yet, because the mother role
identity is embedded within the mother group identity, role-based and group-based
meanings may get activated simultaneously. Given that meanings associated with
the mother role identity and mother group identity often may co-occur in situations,
the verification of the mother identity may be associated with both efficacy-based
self-esteem and worth-based esteem.
Finally, we explore whether identity verification of the mother identity is the
critical identity mechanism that is associated with feelings of efficacy and worth-
based esteem, or whether other identity mechanisms, such as the prominence and
salience of a mother identity also are associated with the different bases of self-
esteem. While a prominent identity is about the centrality or value of an identity to an
individual, a salient identity is the probability of an identity being invoked in
situations (Stets & Serpe, 2013). While prominence is about how an identity is
internalized, salience is about the degree to which an identity will be expressed in a
situation. While identity prominence and salience are sometimes overlapping con-
structs such that an identity can be both prominent and salient, they also may be
distinct as when an identity is important but not salient or salient but not important.
Given their distinctive character, we examine how each is associated with the
different bases of esteem.
Early research linked a salient parent identity to greater parental strains and, in
turn, psychological distress (Simon, 1992). This would imply that a salient parent
identity would reduce feelings of efficacy and worth. Other early research theorized
that rather than importance and salience influencing efficacy and worth-based
esteem, efficacy and worth-based esteem influences identity importance and salience
(Ervin & Stryker, 2001). Presumably, the more effective and worthwhile one feels in
a role, the more prominent and salient will be the role identity. In the current
research, we model identity prominence and salience as a social force rather than a
social product of self-esteem (Brenner et al., 2018; Rosenberg, 1990).
Given that our data is cross-sectional, we cannot test whether identity prominence
or identity salience, or even identity verification is a cause or consequence of
self-esteem. Given that identity processes and self-feelings are ongoing within and
across situations, it is likely that the different dimensions of an identity are both a
source and outcome of self-esteem (Brenner et al., 2018). In the current study, we
simply examine whether a relationship exists.
Method
A survey was administered to mothers from an Early Head Start program in a

southwestern town between 2017 and 2018. The broader, Head Start Program was
established in 1975 as part of the government’s welfare policy to fight against
poverty. The Early Head Start program emerged in 1999 to create healthier attach-
ments between parents and children from low-income families. The program serves
mothers through age three. It offers services to parents that foster the cognitive,
social, and emotional development of children. Services include in-home visits that
promote the parent’s support of their child’s development, classroom courses, and
education and child development services by licensed family childcare providers that
are delivered to the home.
Mothers from the Early Head Start program with infants from 3 months to
36 months were eligible to participate in the current study. Mothers with children
younger than 3 months are excluded given the brief duration of the mother role, and
the thought that participation would be too stressful given their already taxing
responsibility of a newborn. A total of 100 mothers volunteered to fill out the survey.
The survey took about 20 min to complete. Respondents were queried as to their
level of self-efficacy and self-worth. They also were asked questions surrounding the
mother identity including what kind of mother they thought they were given
the different meanings of the mother role that were presented to them, how they
thought others saw them as a mother, and how prominent and salient the mother
identity was to them. Respondents also were asked information on their background
such as their race, age, marital status, and employment.
Measures
Dependent Variables
Efficacy-based Esteem. A 10-item efficacy scale was administered to respondents

(Luszczynska et al., 2005; Scholz et al., 2002; Schwarzer & Jerusalem, 1995).
Respondents were asked how they approach goals, tasks, and challenges. Items
included, for example, “I can always manage to solve difficult problems if I try hard
enough,” “If someone opposed me, I can find the means and ways to get what I
want,” and “It is easy for me to stick to my aims and accomplish my goals.”
Responses were “Not at all true,” “Hardly true,” “Moderately true,” and “Exactly
true” (coded 1–4). The items formed a single factor with an omega reliability of .95.
We summed the items with a high score representing high self-efficacy.
Worth-based Esteem. Self-worth was measured using seven items from the
Rosenberg Self-Esteem Scale (1979), which has been used in prior research (Cast
Table 7.1 Factor analysis of efficacy and worth-based esteem (N ¼ 100)

Factor
loading
Items 1 2
I can always manage to solve difficult problems if I try hard enough. .67 .04
If someone opposed me, I can find the means and ways to get what I want. .41 .15
It is easy for me to stick to my aims and accomplish my goals. .72 .01
I am confident that I could deal efficiently with unexpected events. .71 .02
Thanks to my resourcefulness, I know how to handle unforeseen situations. .77 .07
I can solve most problems if I invest the necessary effort. .70 .09
I can remain calm when facing difficulties because I can rely on my coping abilities. .74 .11
When I am confronted with a problem, I can usually find several solutions. .83 .01
If I am in trouble, I can usually think of a solution. .83 .01
I can usually handle whatever comes my way. .68 .01
On the whole, I am satisfied with myself. .13 .55
At times, I think I am no good at all. .07 .51
I feel that I have a number of good qualities. .08 .91
I feel I do not have much to be proud of. .04 .57
I feel that I am a person of worth. .02 .86
I wish I could have more respect for myself. .31 .38
I take a positive attitude toward myself. .09 .75
& Burke, 2002; Rosenberg et al., 1995; Serpe et al., 2019; Stets & Burke, 2014b).
Responses included “Strongly disagree,” “Disagree,” “Agree,” and “Strongly agree”
(coded 1–4) with each statement such as “On the whole, I am satisfied with myself,”
“At times I think I am no good at all,” and “I feel that I have a number of good
qualities.” The items formed a single factor with an omega reliability of .85. We
summed the items with a high score representing high self-esteem.
We performed a factor analysis (with promax rotation) to examine whether the
items that made up efficacy-based esteem were distinct from the items that made up
worth-based esteem. The results are in Table 7.1. We find that the two constructs
represent two distinct, underlying dimensions given that the items representing
efficacy fall on the first factor loading, and the items representing worth fall on the
second factor loading.
Independent Variables
Mother Identity Discrepancy. We obtained a measure of the mother identity and the
reflected appraisals of the mother identity, as both variables are needed to construct
identity discrepancy or non-verification of the mother identity. For the mother
identity measure, respondents were given a list of six bipolar characteristics based
on meanings that may be involved in being a mother (Collins, 2007; Glenn, 1994;
Table 7.2 Factor analysis of Items Loading

the mother identity (N ¼ 100)
Selfless .49
Forgiving .68
Involved with my child(ren) .75
Physically strong .69
Moral .65
Financial provider .55
Ω .79
Hays, 2007; Ruddick, 2007; Simon, 1997; Thurer, 2007). These included “Very
selfish/Not selfish,” “Very forgiving/Not forgiving,” “Not involved with my child
(ren)/Very involved with my child(ren),” “Very physically strong/Not physically
strong,” “Not moral/Very moral,” and “Not a financial provider/Financial provider.”
We see these meanings as capturing different dimensions of the prevailing ideology
of intensive mothering that is, for example, child-focused, labor-intensive, and
financially expensive.
Early in the survey, respondents were asked how they saw themselves with
respect to each characteristic and identified where they placed themselves between
each bipolar characteristic. Responses ranged from 1 to 7 where 1 reflected agree-
ment with one bipolar characteristic, 7 reflected agreement with the other bipolar
characteristic, and 4 placed the respondents in between the two bipolar characteris-
tics. A factor analysis of the items revealed one factor with an omega reliability of
.79. The results are presented in Table 7.2. Negatively worded characteristics were
reverse coded with a high score reflecting positive characteristics associated with the
mother identity.
Conceptually, reflected appraisals of the mother identity is how mothers think
others close to them see them as a mother. To measure this, respondents were given
the same characteristics as those identified above for the mother identity, and they
responded with how they thought close others saw them on each. Responses again
ranged from 1 to 7 where 1 reflected agreement with one bipolar characteristic,
7 reflected agreement with the other bipolar characteristic, and 4 placed them in
between the two bipolar characteristics. Again, a factor analysis of the items revealed
one factor with an omega reliability of .81. The results are presented in Table 7.3.
Negatively worded characteristics were reverse coded with a high score representing
a high mother identity reflected appraisal.
To create the mother identity discrepancy measure, we first subtracted each
mother identity item from each mother identity reflected appraisal item. We then
squared each difference term, and then summed all the squared differences for the
total mother identity discrepancy score. The squared difference was used because
according to identity theory, a departure of the difference from zero in either a
positive or negative direction will increase negative feelings associated with identity
Table 7.3 Factor analysis of Items Loading

the mother identity reflected
Selfless .55
appraisals (N ¼ 100)
Forgiving .66
Involved with my child(ren) .71
Physically strong .82
Moral .75
Financial provider .49
Ω .81
Table 7.4 Factor analysis of salience of the mother identity (N ¼ 100)

Items Loading
Meeting a person of the same sex for the first time .90
Meeting a friend of a close friend for the first time .92
Meeting a friend of a family member for the first time .89
Meeting a stranger for the first time .79
Ω .92
non-verification (Stets & Burke, 2014a). A high score on the identity discrepancy
measure indicated a high level of non-verification of the mother identity.
Identity Prominence. Respondents reported how much they agreed with three
statements having to do with the importance of being of mother: (1) “Being a mother
is an important part of how I see myself,” (2) “Being a mother is an important
reflection of who I am,” and (3) “I have come to think of myself as a mother”
(Brenner et al., 2014, 2018). Response categories included “Strongly agree,”
“Agree,” “Disagree,” and “Strongly disagree” (coded 1–4). The items were reverse
coded and summed to create a scale, with a higher value indicating a high promi-
nence of the mother identity. The omega reliability of this scale was .91.
Identity Salience. The respondents reported how certain they were that they
would tell the following people who they met for the first time that they were a
mother: (1) a person of the same sex, (2) a friend of a close friend, (3) a friend of a
family member, and (4) a stranger” (Brenner et al., 2014, 2018). Response categories
included “Almost certainly would not,” “Probably would not,” “Probably would,”
and “Almost certainly would” (coded 1–4). The items formed a single factor with an
omega reliability of .92 as shown in Table 7.4. A high score represented high
salience of the mother identity.
Background Characteristics. The background characteristics of race, age, marital
status, and employment were captured. Since most respondents were Hispanic, race
was coded 1 for Hispanic, and 0 for Whites, African Americans, Asian, and Native
Americans. Respondents’ age was calculated in years. Marital status was coded 1 for
married, and 0 for never married, separated, divorced, or widowed. Employment
status included those who currently were full-time workers, part-time workers, and
not working. We created two dummy variables. The variable “part-time and full-
time employment” compared the effects of part-time employment to being
unemployed. The variable “full-time employment” compared the effects of full-time

employment to part-time employment.
Results
Table 7.5 presents the means and standard deviations of the variables used in our
analysis. Demographically, most respondents are Latina/Hispanic (66%). The
remaining respondents are white (12%) Asian (10%), African American (7%),
Native American (1%) and other (3%). The average age is 31. Most respondents
are married (70%), with the remaining never married (21%), divorced (6%), sepa-
rated (1%), or widowed (1%). Few respondents are employed full-time (20%).
Most respondents report relatively high levels of efficacy-based esteem (X
¼ 3.34) and worth-based esteem (X ¼ 3.20). The women also report strong
meanings associated with the mother identity (X ¼ 5.39). The mean value of the
mother identity discrepancy is low (X ¼ 3.04).
We explored which meanings in the mother identity generated the greatest
amount of discrepancy for respondents. We find that two meanings generate the
highest discrepancy: being physically strong and being a financial provider. Respon-
dents see themselves physically stronger compared to how they think that close
others see them. However, they think that close others see them as more of a financial
Table 7.5 Means, standard deviations, and range of variables (N ¼ 100)

Mean
Variables (SD) Minimum Maximum
Efficacy-based esteem 3.34 1 4
(.49)
Worth-based esteem 3.20 1.71 4
(.46)
Mother identity 5.39 1.33 7
(1.25)
Mother identity discrepancy 3.04 0 24.83
(3.78)
Mother identity prominence 3.68 1 4
(.58)
Mother identity salience 3.49 1 4
(.64)
Hispanic .66 0 1
(.47)
Age 31.23 22 54
(6.00)
Married .70 0 1
(.46)
Full-time employment .20 0 1
(.40)
provider compared to how they see themselves. The meaning that generates the least
amount of discrepancy is being moral.
The importance of the mother identity is rather high (X ¼ 3.68) as is its salience
(X ¼ 3.49). Thus, not only do these women see the mother identity as a central
identity to them, but they also report frequently invoking the mother identity in
situations.
Table 7.6 presents the correlations of the variables. Not unexpectedly, efficacy
and worth-based esteem are positively related (r ¼ .38, p .01). Higher efficacy-
based esteem also is positively associated with the mother identity (r ¼ .26, p .01).
As anticipated, as the mother identity discrepancy increases, efficacy-based esteem
decreases (r ¼ .30, p .01). Like efficacy-based esteem, worth-based esteem is
positively related to the mother identity (r ¼ .32, p .01) and a prominent mother
identity (r ¼ .24, p .05). Worth-based esteem is negatively related to the mother
identity discrepancy (r ¼ .25, p .05). The other factor that is positively
associated with worth-based esteem is full-time employment (r ¼ .25, p .05).
Finally, a prominent mother identity is positively associated with strong meanings in
the mother identity (r ¼ .27, p .01) and a more salient mother identity (r ¼ .32,
p .01).
To examine the factors associated with efficacy and worth-based esteem, con-
trolling for the effects of other potential relationships, we use structural equation
modeling in which a full information maximum likelihood procedure allows us to
estimate the coefficients using all possible data. The procedure avoids dropping
cases with partial data or imputing data to missing values, which alternative pro-
cedures are biased and less efficient (Enders & Bandalos, 2001). The standardized
estimates of the coefficients are provided in Table 7.7.
Importantly, we find that a discrepancy in the mother identity is negatively
associated with efficacy-based esteem (β ¼ .32, p .01) and worth-based esteem
(β ¼ .21, p .05). Thus, the mother identity carries meanings associated not only
with the role of mothering or “doing” mothering, but it also carries meanings
associated with group membership in a close unit, the family, and “being” accepted
and valued. When these meanings are not verified, women experience a reduction in
feelings of competence, belongingness, and feeling worthy. Neither the prominence
nor salience of the mother identity is associated with efficacy or worth-based esteem.
Rather, it is the verification of the mother identity that is most relevant for this
sample of women.
The results also reveal that full-time employment, and not part-time employment,
is positively associated with worth-based esteem (β ¼ .25, p .05). Apparently, full-
time work offers the women in this sample a feeling of worth beyond being a mother.
Periodic work does not offer the benefits that full-time work offers in increasing
feelings of worth, showing the importance of work in our culture.
146
Table 7.6 Correlations among the variables (N ¼ 100)

1 2 3 4 5 6 7 8 9 10
1. Efficacy-based esteem 1.00
2. Worth-based esteem .38** 1.00
3. Mother identity .26** .32** 1.00
4. Mother identity discrepancy .30** .25* .39** 1.00
5. Mother identity prominence .17 .24* .27** .18 1.00
6. Mother identity salience .15 .01 .01 .02 .32** 1.00
7. Hispanic .03 .03 .11 .15 .08 .01 1.00
8. Age .08 .01 .02 .04 .01 .03 .26** 1.00
9. Married .15 .04 .23* .11 .20* .12 .31** .05 1.00
10. Full-time employment .01 .25* .25* .09 .17 .12 .01 .02 .01 1.00
*p .05
**p .01
J. E. Stets and J. Lee
Table 7.7 Structural equation models standardized estimates for esteem dimensions (N ¼ 100)
Efficacy-based esteem Worth-based esteem
Independent variables β β
Mother identity discrepancy .32** .21*
Mother identity prominence .05 .17
Mother identity salience .10 .05
Hispanic .07 .04
Age .03 .03
Married .18 .06
Part-time and full-time employment .09 .08
Full-time employment .07 .24*
R2 .15 .15
RMSEA ¼ 0.00 for both efficacy and worth-based models
*p .05
**p .01
Discussion
The findings reveal that it is the verification of the mother identity and not the
prominence or salience of the mother identity that raises feelings of competence and
belongingness for mothers. While recent research established a positive relationship
between a prominent and salient identity and efficacy and worth-based esteem across
multiple role identities, neither a salient nor verified identity was included in the
analysis (Markowski & Serpe, 2018; Serpe et al., 2019). When we simultaneously
examine the effect of multiple dimensions of an identity (a prominent, salient, and
verified identity) on self-esteem, we find that only the verification process plays a
significant role.
The lack of effect of prominence and salience on efficacy and worth-based esteem
is surprising given that the women in this sample reported that the mother identity is
important to them and frequently invoked across situations. We would expect that if
the mother identity were highly valued and often enacted, it would be positively
associated with feelings of competence and worthiness. The high prominence and
salience that these women place on the mother identity may be because most
respondents are Hispanic. Indeed, one of the most important roles of Hispanic
women is to be a wife and mother (Galanti, 2003). Some have argued that while
the role of wife is important, motherhood is seen as more rewarding; when mothers
make sacrifices for their children in Hispanic families, they are glorified and admired
(Preto, 1990). However, other research reveals that white mothers show higher
importance on motherhood than Hispanic mothers, suggesting that white women
may be more susceptible to the ideology of the cult of domesticity than women from
other groups (McQuillan et al., 2008). Because there are not enough white women in
this sample to draw a comparison, future work is needed on the prominence and
salience of the mother identity across different racial/ethnic groups.
It is possible that the effect of a prominent or salient identity on esteem is

contingent on the verification process. If a high prominent or salient identity is
verified, it may raise feelings of efficacy and belongingness compared to a low
prominent or salient identity. Indeed, as discussed earlier, having one’s identity
verified is what is especially relevant in generating positive self-esteem (Burke &
Harrod, 2005; Burke & Stets, 1999; Cast, 2004; Cast & Burke, 2002; Stets &
Harrod, 2004). We cannot examine these relationships in the current research
given that the sample is small. Future work with a larger sample would produce
more reliable results.
Whether the effect of identity verification and efficacy and worth-based esteem in
this study is unique to this group of people who rest at the intersection of multiple
categories of disadvantage: gender, race, and class, or whether it is a more general
pattern also requires further research. Research reveals that Hispanics, who comprise
most of the respondents in this study, tend to have a collectivistic orientation such
that they consider close others an integral part of who they are, and they want to
belong and enjoy being part of groups (Oyserman et al., 2002). These very charac-
teristics may make them more sensitive to the views of others, and how they think
close others’ view them as a mother is an important basis of our identity discrepancy
measure in this study. Research also reveals that lower status actors are more
sensitive to their interaction partners’ perspectives, and they take them into account
when acting (Love & Davis, 2014). Lacking the structural resources to take the
initiative or lead in situations, lower status actors rely more on interpersonal
resources including the views of others to guide their behavior in situations. Accom-
modating to others’ views helps ensure that they are accepted by others. This may be
what is occurring in the mother identity in this study.
If the verification of the mother identity is associated with efficacy-based esteem
and worth-based esteem for mothers, it is important to consider how role-based
identity meanings might influence group-based identity meanings and vice versa. As
discussed elsewhere, if the verification of a group-based identity fosters integration
in society and makes one feel valuable, this may allow one, in turn, to take risks to
pursue verification of a role-based identity (Stets & Burke, 2014b). For the mother
identity, if the family provides support and love, feeling worthwhile can encourage
mothers to take on the “motherhood mandate” of intensive mothering meanings even
more, with the goal of being seen as competent and effective in the mother role.
Alternatively, feelings of competence that are generated when maternal practices
register meanings of being “very involved” to both self and others, may facilitate
stronger meanings of devotion and commitment by mothers toward the family unit.
While the set of meanings that we used to operationalize the mother identity in
this study factored into one underlying construct with a high reliability, these
meanings provide only a preliminary set of meanings that may be tied to mother-
hood. We do not claim that they capture the breath or depth of meanings that exist for
mothers. While the goal was to operationalize meanings that reflected “intensive
mothering,” further research is needed that examines additional meanings than those
used here. And, whether the meanings of the mother identity differ by race and class
needs to be examined.
There is reason to suspect that like other racial/ethnic groups, Hispanics adopt the
intensive ideology view of motherhood. There is diversity among Hispanic mothers
in terms of their cultural/national background, degree of acculturation, and economic
circumstances making adherence to the intensive mothering mandate somewhat
uneven. For example, among Puerto Rican Americans, female-headed households
are more common than among Mexican Americans (Umana-Taylor & Alfaro, 2006),
making intensive mothering difficult if they are the sole financial provider. Another
variation is that compared to more acculturated Hispanics, less acculturated His-
panics are more likely to endorse more traditional, sex-role differentiation in the
family and foster interaction with extended kin (Negy & Woods, 1992), making
intensive mothering more likely, although perhaps shared with grandmothers, god-
mothers, and aunts. Irrespective of within variation, Hispanics are the most likely to
report that they are stay-at-home moms (Livingston, 2014) than other racial/ethnic
groups, which is consistent with the intensive mothering ideology. Additionally,
across different racial/ethnic groups including Hispanics, women tend to have a
positive attitude toward being a mother, thus suggesting that the ideology of
intensive mothering continues to be influential in their orientation toward mother-
hood (Tichenor et al., 2017).
What may be different is that some meanings that relate to intensive mothering
may take on more significance than other meanings for different racial/ethnic groups.
For example, it is possible that the meaning of being a financial provider is more
relevant for lower income, minority women, than higher income, non-minority
mothers. The higher fertility of Hispanic women compared to white women may
block Hispanic women’s opportunities for full-time work, and if they obtain full-
time work, they may experience discrimination by being placed in lower tiered jobs.
These circumstances will have a bearing on being a successful financial provider.
The results suggest that mothering is not simply about doing—about competence
in the mother role activities—but, it is also about being—about belonging to
something larger than oneself whether that be a family, or a larger unit as in
assuming a valued status associated with being female in society. It is not surprising
that an increase in identity discrepancy is associated with a reduction in efficacy-
based esteem. The intensive mothering ideology is rather pervasive in society,
representing a “hegemonic form of mothering” (Tichenor et al., 2017), and it is
easy to shore up the role expectations and corresponding meanings that mothers
strive to confirm.
The fact that a discrepancy in the mother identity also is associated with worth-
based esteem is important to highlight. A reduction in a feeling of worthiness and
acceptance is experienced when the mother identity is not verified. However, future
research will want to examine whether this finding is just as strong in other races/
ethnicities than the largely Hispanic group of women in this study. The family is the
primary unit in the Hispanic culture (Galanti, 2003). Hispanics value familism
including large families, family solidarity, and close bonds with extended kin
(Segura & Pierce, 1993; Skogrand et al., 2009). Consequently, a lack of feeling
that one belongs may deeply affect them when their mother identity is not verified
compared to other groups in society.
There is another base of identity that we have not discussed in this research: the
desire to be who people generally feel they are. This is the experience of authenticity,
which is expressed in person-based identities (Burke & Stets, 2009; Stets & Burke,
2014b). Person identities are culturally recognized characteristics that individuals
internalize as core to themselves such as being more or less “moral” (Stets & Carter,
2011, 2012) or “dominant” (Stets et al., 2020) that define them in distinct ways from
others in society. These person identities are conceptualized as always with individ-
uals within and across situations, and they influence the meanings held in role and
group identities. Person identities represent the “me” as opposed to the “we” of
group or social identities. Person identities are not about “doing” as in role identities
but about expressing the “real” or “true” self.
Some role identities may share meanings not only with group and social identities
but also person identities. The mother identity may be one such role identity that
does this. It is possible that meanings that characteristic the mother role identity such
asbeing “selfless” and “moral” overlap with the meanings that define these women as
persons. Thus, when their mother identity is verified, it also may raise authenticity-
based self-esteem (Stets & Burke, 2014b). Future work will want to examine this.
More generally, the verification of role-based and group-based identities as
revealed in the mother identity, or confirming doing and being, satisfies people’s
needs to live their lives in distinctive ways with others, while at the same time in
unity with others. Our role identities guide us in the individual contribution we can
make to something larger than ourselves, while we simultaneously see ourselves as a
part of that greater good. Neither doing nor being is more “essential.” Both bases of
identities reference different aspects of the social structure (roles within positions
and positions within groups) that make society possible.
Acknowledgements We thank members of the social psychology seminar at the University of

California, Riverside, for the comments on an earlier version of this paper.
References
Arendell, T. (2000). Conceiving and investigating motherhood: The decade’s scholarship. Journal
of Marriage & the Family, 62(4), 1192–1207.
Baxter, J., Buchler, S., Perales, F., & Western, M. (2015). A life-changing event: First births and
men’s and women’s attitudes to mothering and gender divisions of labor. Social Forces, 93(3),
989–1014.
Bosson, J. K., & Swann, W. B. (1999). Self-liking, self-competence, and the quest for self-
verification. Personality and Social Psychology Bulletin, 23(10), 1230–1241.
Brenner, P. S., Serpe, R. T., & Stryker, S. (2018). Role-specific self-efficacy as precedent and
Burke, P. J., & Cast, A. D. (1997). Stability and change in the gender identities of newly married
couples. Social Psychology Quarterly, 60(4), 277–290.
Burke, P. J., & Harrod, M. M. (2005). Too much of a good thing? Social Psychology Quarterly, 68
(4), 359–374.
Burke, P. J., & Cerven, C. D. (2019). Identity accumulation, verification, and well-being. In J. E.
Stets & R. T. Serpe (Eds.), Identities in everyday life (pp. 17–33). Oxford University Press.
Cast, A. D. (2004). Well-being and the transition to parenthood: An identity theory approach.
Sociological Perspectives, 47(1), 55–78.
Chodorow, N. (1989). Feminism and psychoanaytic theory. Yale University Press.
Collins, P. H. (2007). Shifting the center: Race, class, and feminist theorizing about motherhood. In
A. O’Reilly (Ed.), Maternal theory: Essential readings (pp. 311–330). Demeter Press.
Delle Fave, A., Brdar, I., Freire, T., Vella-Brodrick, D., & Wissing, M. P. (2011). The eudaimonic
and hedonic components of happiness: Quallitative and quantitative findings. Social Indicators
Research, 100(2), 185–207.
Douglas, S. J., & Michaels, M. W. (2004). The mommy myth: The idealizatin of motherhood and
how it has undermined all women. Free Press.
Durkheim, E. (1893 (1984)). The division of labor (G. Simpson, Trans.). The Free Press.
Ellestad, J., & Stets, J. E. (1998). Jealousy and parenting: Predicting emotions from identity theory.
Endendijk, J. J., Derks, B., & Mesman, J. (2018). Does parenthood change implicit gender-role
stereotypes and behaviors? Journal of Marriage & the Family, 80(1), 61–79.
Enders, C. K., & Bandalos, D. L. (2001). The relative perforance of full information maximum
likelihood estimation for missing data in structural equation models. Structural Equation
Modeling, 8(3), 430–457.
T. J. Owens & S. Stryker (Eds.), Extending self-esteem theory and research: Sociological and
psychological currents (pp. 29–55). Cambridge University Press.
Franks, D. D., & Marolla, J. (1976). Efficacious action and social approval as interacting dimen-
sions of self-esteem: A tentative formulation through construct validation. Sociometry, 39(4),
324–341.
Francis, C., & Heise, D. R. (2006). Mean affective rations of 1,500 concepts by indiana university
undergraduates in 2002–3.
Freely, M. (2000). The parent trap: Children, families, and the new morality. Virago.
Galanti, G.-A. (2003). The hispanic family and male-female relationships. Journal of Transcultural
Nursing, 4(3), 180–185.
Gecas, V. (1989). The social psychology of self-efficacy. Annual Review of Sociology, 15, 291–
316.
Gecas, V. (1991). The self-concept as a basis for a theory of motivation. In J. A. Howard & P. L.
Callero (Eds.), The self society interface: Cognition, emotion and action (pp. 171–188).
Cambridge University Press.
Gecas, V., & Schwalbe, M. L. (1986). Parental behavior and adolescent self-esteem. Journal of
Marriage and the Family, 48(1), 37–46.
Gillespie, R. (2003). Childfree and feminine: Understanding the gender identity of voluntarily
childless women. Gender & Society, 17(1), 122–136.
Glenn, E. N. (1994). Social constructions of mothering: A thematic overview. In E. N. Glenn, G.
Chang, & L. R. Forcey (Eds.), Mothering: Ideology, experience, and agency (pp. 1–32).
Routledge.
Hays, S. (1996). The cultural contadictions of motherhood. Yale University Press.
Hays, S. (2007). Why can’t a mother be more like a businessman? In A. O’Reilly (Ed.), Maternal
theory: Essential readings (pp. 408–430). Demeter Press.
Johnston, D. D., & Swanson, D. H. (2006). Constructing the ‘good mother’: The experience of
mothering ideologies by work status. Sex Roles, 54(7), 509–519.
Leary, M. R. (2007). Motivational and emotional aspects of the self. Annual Review of Psychology,
58, 317–344.
Leary, M. R., & MacDonald, G. (2003). Individual differences in self-esteem: A review and
theoretical integration. In M. R. Leary & J. P. Tangney (Eds.), Handbook of self and identity
(pp. 401–418). The Guilford Press.
Livingston, G. (2014). Among hispanics, immigrants more likely to be stay-at-home moms and to
believe that’s best for kids.
Love, T. P., & Davis, J. L. (2014). The effect of status on role-taking accuracy. American
Luszczynska, A., Scholz, U., & Schwarzer, R. (2005). The general self-efficacy scale: Multicultural
validation studies. The Journal of Psychology, 139(5), 439–457.
of prominence and salience in the verification and self-esteem relationship. Advances in Group
McMahon, M. (1995). Engendering motherhood: Identity and self-transformation in women’s
lives. The Guilford Press.
McQuillan, J., Greil, A. L., Shreffler, K. M., & Tichenor, V. (2008). The importance of motherhood
among women in the contemporary United States. Gender & Society, 22(4), 477–496.
Musick, K., Meier, A., & Flood, S. (2016). How parents fare: Mothers’ and fathers’ subjective well-
being in time with children. American Sociological Review, 81(5), 1069–1095.
Negy, C., & Woods, D. J. (1992). The importance of acculturation in understanding research with
hispanic-americans. Hispanic Journal of Behavioral Sciences, 14(2), 224–247.
Nomaguchi, K. M., & Milkie, M. A. (2020). Parenthood and well-being: A decade in review.
Journal of Marriage and the Family, 82(1), 198–223.
Nuttbrock, L., & Freudiger, P. (1991). Identity salience and motherhood: A test of stryker’s theory.
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectiv-
ism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 129(1),
3–72.
Pearlin, L. I., Lieberman, M. A., Menaghan, E. G., & Mullan, J. T. (1981). The stress process.
Journal of Health and Social Behavior, 22(4), 337–356.
Preto, N. G. (1990). Hispanic mothers. Journal of Feminist Family Therapy, 2(2), 15–32.
Ridgeway, C. L., & Kricheli-Katz, T. (2013). Intersecting cultural beliefs in social relations:
Gender, race, and class binds and freedoms. Gender & Society, 27(3), 294–318.
Rosenberg, M. (1990). The self-concept: Social product and social force. In M. Rosenberg & R. H.
Turner (Eds.), Social psychology: Sociological perspectives (pp. 593–624). Transaction
Publishers.
(1), 141–156.
Ruddick, S. (2007). Maternal thinking. In A. O’Reilly (Ed.), Maternal theory: Essential readings
(pp. 96–113). Demeter Press.
Scholz, U., Dona, B. G., Sud, S., & Schwarzer, R. (2002). Is general self-efficacy a universal
construct? European Journal of Psychological Assessment, 18(3), 242–251.
Schwarzer, R., & Jerusalem, M. (1995). Causal and control beliefs. In M. Johnston, S. Wright, & J.
Weinman (Eds.), Measures in health psychology: A user’s portfolio (pp. 35–37). NFER-Nelson.
Segura, D. A., & Pierce, J. L. (1993). Chicana/o family structure and gender personality: Chodorow,
familism, and psychoanalytic sociology revisited. Signs, 19(1), 62–91.
Serpe, R. T., Yarrison, F., Stets, J. E., & Stryker, S. (2019). Multiple identities and self-esteem. In J.
E. Stets & R. T. Serpe (Eds.), Identities in everyday life (pp. 53–71). Oxford University Press.
Simon, R. W. (1992). Parental role strains, salience of parental identity, and gender differences in
psychological distress. Journal of Health and Social Behavior, 33(1), 25–35.
Simon, R. W. (1997). The meanings individuals attach to role identities and their implications for
mental health. Journal of Health and Social Behavior, 38(3), 256–274.
Skogrand, L., Barrios-Bell, A., & Higginbotham, B. (2009). Stepfamily education for latino
families: Implications for practice. Journal of Couple & Relationship Therapy, 8(2), 113–128.
Quarterly, 63(3), 224–237.
Stets, J. E., & Burke, P. J. (2014a). Emotions and identity nonverification. Social Psychology
Quarterly, 77(4), 387–410.
Stets, J. E., & Burke, P. J. (2014b). Self-esteem and identities. Sociological Perspectives, 57(4), 1–
25.
Stets, J. E., Burke, P. J., & Savage, S. V. (2018). Exchange, identity verification, and social bonds.
Stets, J. E., & Carter, M. J. (2011). The moral self: Applying identity theory. Social Psychology
Quarterly, 74(2), 192–215.
Stets, J. E., & Carter, M. J. (2012). A theory of the self for the sociology of morality. American
Stets, J. E., Savage, S. V., Burke, P. J., & Fares, P. (2020). Cognitive and behavioral responses to
the identity verification process. In R. T. Serpe, R. Stryker, & B. Powell (Eds.), Identity and
symbolic interaction: Deepening foundations, building bridges (pp. 65–88). Springer.
Stets, J. E., & Tsushima, T. (2001). Negative emotion and coping responses within identity control
theory. Social Psychology Quarterly, 64(3), 283–295.
Swann, W. B., & Bosson, J. K. (2010). Self and identity. In S. T. Fiske, D. T. Gilbert, & G. Lindzey
(Eds.), Handbook of social psychology (pp. 589–628). John Wiley and Sons.
Tafarodi, R. W., & Swann, W. B., Jr. (1995). Self-liking and self-competence as dimensions of
global self esteem: Initial validation of a measure. Journal of Personality Assessment, 65(2),
322–342.
Thoits, P. A., & Virshup, L. K. (1997). Me’s and we’s: Forms and functions of social identities. In
R. D. Ashmore & L. J. Jussim (Eds.), Self and identity: Fundamental issues (pp. 106–133).
Thurer, S. L. (2007). The myths of motherhood. In A. O’Reilly (Ed.), Maternal theory: Essential
readings (pp. 331–344). Demeter Press.
Tichenor, V., McQuillan, J., Greil, A. L., Bedrous, A. V., Clark, A., & Shreffler, K. M. (2017).
Variation in attitudes toward being a mother by race/ethnicity and education among women in
the United States. Sociological Perspectives, 60(3), 600–619.
Turner, J. H. (2016). The macro and meso basis of the micro social order. In S. Abrutyn (Ed.),
Handbook of contemporary sociological theory (pp. 123–148). Springer.
Umana-Taylor, A. J., & Alfaro, E. C. (2006). Divorce and relationship dissolution among latino
populations in the United States. In M. A. Fine & J. H. Harvey (Eds.), Handbook of divorce and
relationship dissolution (pp. 515–530). Taylor and Francis.
United States Census Bureau. (2012). Percent of the project population by race and hispanic origin
for the United States: 2015 to 2016. United States Census Bureau Report #np2012-t6.
Warner, J. (2006). Perfect madness: Motherhood in the age of anxiety. Vermilion.
Chapter 8
From Stoke to Swaddle: Surfer-Dads,
Voluntary and Obligatory Identities,
and Identity Change
Andrew Quinn
Using ethnography and in-depth interviewing, I examine the identities of surfers

who become new fathers. Specifically, this research utilizes the extended case
method (Burawoy, 1998). The extended case method is immersive by definition;
therefore, theory is necessary to ground the research. Theory provides the impetus
for the research because it serves as a point of departure for the research, allowing the
data to elaborate that theory (Burawoy, 1998). This research is firmly rooted in the
well-established framework of Identity Theory (Burke & Stets, 2009).
Identities refer to roles in the social structure and help provide purpose and
meaning for individuals’ lives by connecting us to society (Burke & Stets, 2009).
For example, the father identity connects an individual to a family structure and to
other family members. Identities are comprised of meanings, such as being a
provider in the father identity, and we feel positive emotions when reflected
appraisals from others reaffirm our identity meanings (Burke & Stets, 2009).
When we perceive that how others view us is consistent with how we view
ourselves, it contributes to a stable sense of self and high self-esteem (Cast &
Burke, 2004). Thus, identities are meanings that are used to define the self, and
how we view ourselves in these roles have important implications for self-develop-
ment and psychological well-being. The impact of the surfer and father identities on
participants’ lives is examined in this research.
Becoming a father is a stressful life event and a culturally important social role
(Baxter et al., 2015; Bleidorn et al., 2016; Cast, 2004). In general, fathers are
becoming increasingly more involved in traditional childrearing activities; addition-
ally, fathers view their role as more important to their identity than in previous
generations (Livingston & Parker, 2019). Despite this, some fathers are more
tangibly involved than others (Collett et al., 2015). From the perspective of identity
A. Quinn (*)
University of California, Riverside, CA, USA
156 A. Quinn
theory, becoming a father is a source of an abrupt identity change through the

process of acquiring a new identity (Burke & Stets, 2009). As a culturally important
identity, becoming a father also contains the potential to produce identity change by
changing how we view ourselves in other identities. This research examines how
surfer-dads navigate this transition to fatherhood.
Conceptually, identity theory distinguishes between obligatory and voluntary
identities (Thoits, 2003). This reflects the idea that some identities are more fre-
quently and compulsory occupied (obligatory), while others are less common and
more freely chosen (voluntary). However, there is still much that is unknown about
obligatory and voluntary identities (Stets & Serpe, 2013). Therefore, this study uses
the ethnographic case of surfer-dads to observe a voluntary and obligatory identity,
respectively. The surfer identity provides insight into what draws people to choose a
voluntary identity, and how it can interact with other identities. The father identity
provides insight as to why some men choose to take on an obligatory identity and
demonstrates how fatherhood is a potent source for identity change.
In the following sections, I first detail the methods of data collection and the
analytical strategy. Included is my personal positionality in relation to the research.
Then, I detail the relevant theoretical concepts by conceptualizing an identity and the
process of identity verification, followed by obligatory and voluntary identities. The
surfer identity and father identity are contextualized by examining recent research on
surfing and fatherhood, situating this research in current literature. Results detail the
salience and prominence of the surfer identity, as well as its effect on self-develop-
ment, through impacting the worker and spouse identities. Then, the father identity is
discussed as a potent force for identity change. This occurs through new identity
acquisition, as well as through changing the salience and prominence of other
identities. Participants frequently refer to a juxtaposition between being a surfer
and father, describing the former as “selfish” and the latter as “selfless,” revealing a
characteristic of identities that may be important to balance. The significance of this
research is detailed in the discussion.
Data Collection Method and Analytical Strategy
This research utilizes ethnographic field methods, specifically the extended case
method (Burawoy, 1998) and in-depth interviews. The extended case method uses a
specific ethnographic case to elaborate theoretical concepts. Using identity theory,
surfer-dads are representative of a voluntary and obligatory identity, respectively.
Fieldwork is conducted by actively participating in the lives of those being studied. I
have spent time surfing with participants, in some cases have met participants’
children, and am a father, myself.
The extended case method is a fluid, explorative, and iterative process. It is fluid
in that data collection and analysis is a continuous process with a broad scope
defining the boundaries of the field. It is explorative because the researcher partic-
ipates in the everyday lives of participants to explore theory. My experiences with
8 From Stoke to Swaddle: Surfer-Dads, Voluntary and Obligatory Identities,. . . 157
each identity provide context for the research and facilitates a rapport where partic-
ipants are able to articulate nuances about surfing and fatherhood. Because the
research process is continuous and guided by theory, data collection and analysis
benefit from being able to reflect on theory throughout. This makes it an iterative
process.
Entry into the field occurred in the summer of 2012 when I moved to Southern
California to pursue an M.A. in Sociology. I had my first surf experience with a
professor who invited my graduate student cohort to surf one weekend. A few of us
showed up. He brought surf boards, taught the basics of surfing, and he was patient
enough to come into the water to help. In the hourlong session, I was thoroughly
thrashed around by the ocean, and never came close to making it past “the break”
and into “the lineup.” This is surfer nomenclature for making it past the point where
the waves break to position oneself to actually surf a wave.
That day, I gained a deep appreciation for how difficult surfing is. We went out
several more times during my tenure as a graduate student. I managed to become
proficient at getting past the break. Once, I even caught a wave. I recall involuntarily
belting out a loud “WHOOOP!” of sorts. My professor later commented that he had
seen “the twinkle in my eye” as I successfully stood on the board for the first time.
The genuine joy in my yell, and the accompanying twinkle in my eye, are emblem-
atic of what surfer’s call “the stoke.” After graduating, the same professor gifted me
one of his old surfboards. At the time, I lived mere blocks from the ocean, and all he
wanted was for me to continue chasing “the stoke.”
I worked in the service industry for over seven years in Southern California.
Coworkers often came into work chatting about the morning surf session, tomor-
row’s surf report, and upcoming plans of surf trips. I would go surfing with
coworkers every now and then. One of my coworkers even gave me a new wetsuit
for the occasions (and later became my first research participant). My experiences
with my coworkers and my professor were my introduction to surfing and the surf
culture, a familiarity cultivated over the last seven years. Importantly, I would not
consider myself a surfer, making me an outsider to the culture and to the surfer
identity.
I witnessed two of my fellow servers, both of whom were dedicated surfers,
become fathers for the first time while working in the service industry. This led both
of them to move inland, away from their beach community abodes. Simultaneously,
their time allotted to surfing decreased drastically. I was curious how the lack of
surfing might affect their identity as a surfer. Soon after, I became a father myself.
This was just two short months before embarking on the arduous journey through a
Ph.D. program. The experience of becoming (and being) a father was life-altering. It
is especially surreal watching this research on fatherhood evolve while simulta-
neously watching my daughter grow. In this sense, I am an insider to my research as I
am part of the shared nuances of fatherhood.
I share this background information because I believe my positionality is integral
to this research. I am uniquely situated as an insider/outsider. As one participant
remarked about fatherhood, “only a parent knows,” and with a wide smile shared
between the two of us, no further explanation was needed. Interestingly, a later
158 A. Quinn
participant remarked the same about surfing and the feeling of catching a wave:
“Only a surfer knows.” This time, knowing that I was not a surfer, the participant felt
the need to elaborate as best as he could about what it is that “only a surfer knows.”
The message expressed in these sentiments is that there is something unique to the
experience of being fathers or surfers that act as something of a bond because “only
they know.” As an insider to fatherhood, I am privy to the insights and experiences
of fatherhood that “only a parent knows.” But I am an outsider to the insights and
experiences that “only a surfer knows.”
The time spent in the field surfing and being a father provides tacit knowledge that
can only be gained through experience (Burawoy, 1998). This is important because
it provides insight into the identities being studied, allowing the researcher to engage
with theory with the benefit of these insights. This experience helps inform the
context and questions of the interview itself, as well as supplement the interview
data. Though I had some tacit knowledge about surfing to inform the interview, my
outsider status allowed participants to be experts in their field.
Fifteen Interviews1 were conducted over a two-year period, beginning in January
2017, and concluding in December of 2018. Participants were first recruited via
convenience sampling through personal networks. Additional participants were
recruited via snowball sampling. The duration of the semi-structured interviews
lasted between an hour and an hour and a half. The first half of the interview focused
on the surfer identity and the second half focused on the father identity.
Detailed analytic research memos were compiled throughout the data collection
process (Saldaña, 2013). Memos were written immediately following interviews to
compile and elaborate on notes taken during the interview as well as capture first
impressions of the interview. Memos also were written after transcribing each
interview that focused on first impressions of what stood out or surprised me in
the interview; what was similar or different relative to previous interviews; and to
reflect on my interview technique. Additionally, memos were written after coding
each interview, focusing on making meaning out of the emerging codes. Further,
memos were written after conversations with mentors (Saldaña, 2013), attempting to
make meaning of emerging results. Finally, memos were written after “member
checks” in which I shared emerging results with participants to improve the validity
of the results (Emerson et al., 1995; Saldaña, 2013).
Interviews were transcribed and coded manually to assist in generating familiarity
with the data. A combination of first and second cycle coding strategies was
implemented throughout the data collection and analysis process. First cycle coding
refers to the initial rounds of coding, while the second cycle reorganizes codes into
collapsed categories (Saldaña, 2013). The coding strategies implemented were first
structural coding and then focused coding (Saldaña, 2013). Structural coding is a
first cycle coding strategy that creates codes based on the research questions that
frame the interview. This organizes codes in a structure that mirrors the interview,
making coded segments of interview data easy to compare across interviews.
1
Detailed demographic information about interviewees included in the Appendix.
Segments that are similarly coded are identified for further coding and analysis.
Focused coding is a second cycle coding strategy that collapses the codes into fewer
categories that emphasize elaborating on theoretical processes.
In structural coding, one is attentive to specific research questions. This coding is
recommended in standardized or semi-structured data collection methods, particu-
larly interview transcripts. The researcher examines comparable sections of many
interviews to assess commonalities, differences, and relationships as related to the
research questions. This approach was used to make a distinction between the
obligatory identities (such as the father identity) and the voluntary identity (the
surfer identity). Interviews were organized around a discussion of the surfer identity
first, and then the father identity.
The focused coding process is a more concentrated and detailed analysis of one’s
research questions with an emphasis on theoretical processes. First, I identified
identity salience and prominence in the surfer identity. Then, I identified how the
voluntary surfer identity effected other aspects of the self, including obligatory
identities like the worker and spouse identity. In the obligatory father identity, I
focused on identity change. Identity change occurs first through the acquisition of
the father identity, and then through changing the salience and prominence of other
identities. Finally, participants describe each identity as integral to their self-devel-
opment by virtue of being “selfish” and “selfless” identities, respectively.
Theory
Identities
People have as many identities as they have roles, groups, and different ways in
which they define themselves (James, 1890). Specifically, an identity is defined as
the meanings attached to the roles people hold in society, the groups or social
categories they belong to, and the unique ways they see themselves as individuals.
(Burke & Stets, 2009). Given this, there are three distinct bases of an identity: role
identities, group/social identities, and person identities. The identities observed in
this research, the surfer and father, are role identities.
A role identity is the meaning tied to a role that is attached to a position in the
social structure (Burke, 1980; McCall & Simmons, 1978). Examples of role identi-
ties include meanings attached to family roles (such as parent, spouse, or sibling) as
well as other roles such as friend, worker, or student. Role identities are defined, in
part, by their relationship to counter-role identities in the social structure. For
example, the role identity of father is defined in relation to the role identity of
child and mother. This means that internalizing and enacting the meanings associ-
ated with the father role occur with complimentary meanings associated with the
child role or mother role that also are enacted.
It is important to note that it is the meanings in an identity that guide an
individual’s behavior. The process begins when an identity is activated in a situation.
160 A. Quinn
Identity activation occurs when the meanings perceived in a situation correspond

with the meanings in an identity. Through a continuous feedback loop, individuals
control their perceptions in a situation so that how they see themselves in an identity
matches how they think others see them. This is identity verification (Burke & Stets,
2009). When identity verification occurs, individuals will feel good about them-
selves and experience positive emotion. When identity non-verification occurs,
individuals will feel bad, and they will work to better control perceptions to obtain
identity verification.
Thoits (2003) has made a theoretical distinction between obligatory and volun-
tary identities. Recent research demonstrates that people tend to view their identities
as either obligatory or voluntary in the ways that identity theory has conceptualized
(Gallagher, 2016). However, little research has examined this distinction further
(Stets & Serpe, 2013). The purpose of this research is to examine both an obligatory
and voluntary identity occupied by the same individual.
The obligatory identity of being a father and the voluntary identity of being a
surfer are examined. The distinction between obligatory and voluntary identities
reflects the idea that some identities are expected to be occupied across the life
course (Turner, 1978). For example, being a student, worker, spouse, and parent are
identities that individuals are obliged to assume as they age. Other identities may be
tied to individual interests or activities that may not be as commonly occupied
(Turner, 1978). For example, being a dancer (Kushida, 2019) or a volunteer (Thoits,
2012) are identities that individuals are more likely to choose rather than assume out
of obligation. These are voluntary identities.
Voluntary identities contribute more to psychological well-being, relative to
obligatory identities (Thoits, 2003, 2012). Well-being is measured by symptoms of
distress, such as anxiety, and substance use, with less symptoms and less substance
use reflecting greater well-being. Voluntary identities are associated with greater
well-being because these are identities that individuals are freer to choose to occupy
and abandon compared to obligatory identities. Obligatory identities that produce
stress due to the societal expectations tied to them are more difficult to abandon
because of the strong structural and often emotional ties to counter identities (Thoits,
2003). Voluntary identities that are routinely verified will produce positive emotion
and will continue to be occupied, while those that do not are easily abandoned. Using
a volunteer role to represent a voluntary identity, Thoits (2012) illustrates that more
time spent volunteering (identity salience) was directly associated with how impor-
tant people viewed the identity (identity prominence). In turn, the salience and
importance of the identity provides meaning and purpose for the individual, and
therefore contributes more to one’s well-being. More generally, a combination of
voluntary and obligatory identities is important for individuals’ well-being (Thoits,
1992).
Much is still unknown about the nature of voluntary and obligatory identity
formation (Stets & Serpe, 2013). For example, despite obligatory identities being
normative, not all individuals assume them. Not all men become fathers. Moreover,
those who do take on the father identity do so with varying degrees of involvement.
Some fathers are more tangibly involved than others (Collett et al., 2015).
Additionally, we do not know enough about the relationship between obligatory and
voluntary identities. For example, do individuals occupy voluntary identities to
fulfill parts of the self that are not satisfied by obligatory identities (Stets & Serpe,
2013)? This research provides some insight to these gaps. By examining a voluntary
and obligatory identity occupied by the same individual, the dynamics between these
distinct identities can be uncovered. Using qualitative, interpretive methods may be
particularly useful in this context as participants describe in detail what these
identities mean to them.
The surfer identity represents a voluntary identity. All the surfers interviewed had
been surfing for a minimum of five years (oftentimes much longer), indicating they
began surfing in childhood, adolescence, or young adulthood. Before having chil-
dren, all participants consistently surfed a minimum of four times a week for several
years at a time. They also indicated that being a surfer had a significant impact on
how they saw themselves as individuals. This means that participants had suffi-
ciently developed the surfer identity as both a salient (frequently enacted) and
prominent (important) identity.
The father identity represents an obligatory identity. It is a culturally important
adult identity that many experience (Baxter et al., 2015; Bleidorn et al., 2016; Cast,
2004). As noted, not all men become fathers, and some fathers are more involved
than others. All participants assumed the father identity and indicated that it currently
was a significant aspect as to how they saw themselves. The father identity
represented a salient and prominent obligatory identity.
Thus, this study focuses on the development of a salient and prominent voluntary
identity. It also focuses on acquiring a new salient and prominent obligatory identity.
Identity change also is considered as participants reflect on acquiring a new identity.
Understanding how surfers navigate the transition to fatherhood may provide some
insight into the relationship between voluntary and obligatory identities.
The Surfer Identity
There is little research that sociologically examines surfing, and even less that
considers one’s identity as a surfer, making this research a contribution to the
sociological and social psychological literatures. Some macro-sociological
approaches to surfing conceptualize surfing as a counterculture that eschews the
mundane rationalization of everyday life in favor of communion with nature
(Lawler, 2011; Smith, 2010; Stranger, 2011). Stranger (2011) views recreational
(as opposed to competitive) surfing as a postmodern culture. It is an identifiable
subculture, but one without formal organization. Surfing is viewed as a critique of
the increasing rationalization of the modern world, and capitalism, in particular
(Lawler, 2011). From a Marxist perspective, surfing is not only a resistance to
commodity fetishism (Lawler, 2011), but also a fundamental medium to mitigate
one’s alienation from self and nature (Smith, 2010).
162 A. Quinn
Other macro-sociological approaches observe the effects of modernity, or the

increasing inter-connectedness of modern lives, on the surf culture (Daskalos, 2007;
Roberts & Ponting, 2020). Daskalos (2007) finds that “old school” surfers in
“localized” surf breaks2 in Southern California resent the increasing inclusivity of
surfing, which contributes to less barriers to entry into the identity. They view this as
an affront to their “old school” surfing identity. Likewise, others (Roberts & Ponting,
2020) lament the decreasing (geographical) barriers to entry to surfing in their
evaluation of manmade “wave pools.”3 Additionally, by replicating waves, there is
an inherent lack of an authentic connection to nature.
Other approaches to surfing focus on masculinity and reveal a gender paradox
embedded in surf culture. Typically stereotyped as laidback “bro’s,” there is an
element of toxic masculinity in surf culture (Booth, 2001). Tracing the history of
gender and surfing, a cultural shift occurred that redefined gender roles in surfing
predominately in terms of male physical prowess and female objectification. This
has a negative effect on female surfers, a minority among surfers, and male surfers
who do not conform to aggressive masculinity. Others acknowledge the aggressive
masculinity present in surfing culture while providing an analysis that adds com-
plexity (Evers, 2004, 2006). The range of movement, feelings, and emotions that
men experience while surfing makes more complex the masculinity of surfers. Thus,
men learn the unwritten cultural rules of both surfing and masculinity in the water,
becoming socialized into gender and the surf culture simultaneously. The current
research is situated in this conversation by focusing on the kinds of surfers who go
on to become fathers, thus expounding on issues of masculinity within surfing.
Taken together, the culture of surfing, and therefore the identity of surfers, has
undergone change over the last several decades (Daskalos, 2007). Moreover, similar
to how the father identity is susceptible to vague identity meanings in the wake of
“new fatherhood” ideals (Collett et al., 2015), the surfer identity appears to be under
threat by the interconnectedness ushered in by modernity (Daskalos, 2007).
The Father Identity
Recent research by the Pew Research Center (Livingston & Parker, 2019) shows that
beyond the primary breadwinner role, fathers are becoming more involved in
childrearing and caretaking. Additionally, more fathers view parenting as extremely
important to their identity than previous generations (Livingston & Parker, 2019).
Nevertheless, McLaughlin and Muldoon (2014) find significant tension between
new fatherhood ideals and actual practices. For example, contemporary fatherhood
2
A “localized” surf break refers to a specific stretch of beach, and therefore ocean, that is known to
be surfed by “locals only.”
3
“Wave-pools” are indoor facilities where waves are manually produced to mimic the effect of
natural waves found in outdoor surf breaks.
tends to include more “home-making” efforts, yet they also find the “breadwinning”
aspect does not necessarily wane in proportion. Thus, while contemporary fathers
are embracing the childrearing aspects of fatherhood, they are still internalizing more
traditional aspects of fatherhood such as the breadwinner role. Although this allows
for more variability in (and therefore verification of) the father identity, it also
contains a high potential for role conflict (McLaughlin & Muldoon, 2014).
Collett et al. (2015) find that the meanings tied to the father identity are variable
and sometimes not clear. For example, in their study, they find that fathers professed
to “being there” for their children as requisite for being a good father. Yet, when
asked to elaborate on how they accomplished this, some fathers could not articulate
what “being there” meant, nor could they provide tangible examples. Thus, some
men held vague father identity meanings (Collett et al., 2015). Though their father
identity meanings were vague, they still perceived themselves as being a “good
father” despite the lack of reports that they were verified in the father identity.
Conversely, men who provided specific examples of “being there” for their children
reported high levels of verification of the father identity (Collett et al., 2015). Thus,
being verified and supported in the father identity was related to a close correspon-
dence between identity meanings and meaningful behavior or involvement as a
father.
The meanings and behaviors associated with the father identity may be influenced
by significant others. For example, Adamsons (2010) discusses the effects of
parental “gatekeeping” on parental behavior. Adamsons research on parental gate-
keeping shows that mothers who believe fathers should be more involved tend to
behave in ways that positively promote father involvement. She suggests that
mothers’ expectations of fathers are part of the set of meanings contained in their
mother identity. Thus, parents have their own expectations in parenting, but each
parent also influences the expectations of the other for parenting. In a similar vein,
Collett et al. (2019) demonstrate the role of significant others (one’s parents and
especially one’s partner) in the development of meanings associated with the father
identity. Further, Cast (2004) finds that role congruence, or agreement and under-
standing of one’s own role as well as one’s partner’s role, facilitates identity
verification among new parents, thus promoting well-being amidst a difficult life
transition.
While Adamsons’ (2010) work reveals how one parent can influence the behavior
of the other parent, and while Collett and her associates (2019) extend this influence
to the meanings that men hold for the father identity, Cast’s (2004) research shows
that partners agreement on parenting engenders a feeling of marital well-being.
Thus, a partner’s expectations matter in how one becomes a parent, but congruent
expectations allow parents to thrive. For surfer-dads in this research, then, it is
important for them that their spouses understand the importance of the surfer identity
on their lives prior to becoming fathers. Should this be the case, surfer-dads may be
more willing to be involved as fathers because of the mutual understanding of what
is important between spouses.
164 A. Quinn
Results
The Surfer Identity
“The stoke” was a common phrase used by surfers. It described the feeling of
catching a wave. The stoke also involved the positive feelings that emerged follow-
ing a surf session. And so, I begin with a discussion of this phrase as it was a
common refrain among surfers and seemed to capture the essence of the surfing
experience. Then, I discuss the salience and prominence of the surfer identity. This is
followed by a discussion of how the surfer identity helped shape other aspects of the
self, most notably by influencing participants’ worker and spouse identities—two
obligatory identities.
The Stoke
Stoke is something like the feeling of excitement that resonates through the body. I think
there is a couple different kinds. There is the stoke of the moment. Like this morning I caught
a nice wave, came out of the barrel, and felt like “YEAH!!!” This stoke is for a couple
minutes. But then, the bigger part of it is you get out of the water, you feel kind of exhausted,
your body feels good, and your mind feels replenished. Maybe it’s the exercise in combi-
nation with the ocean that elevates the experience. I truly find that it makes me a better
person.
Jeff referred to the “bigger part” of the stoke as the more lasting feeling that
resonated even after the surf experience was over. It is the combination of all these
alluring elements that may draw surfers to develop the surfer identity. Jeff found that
the surf experience carried him through the day and helped “make [him] a better
person.” This suggests that the surf experience had a positive effect on the develop-
ment of Jeff. Likewise, Jason shared the fact that when he was young, he was
hyperactive and struggled with school, such that even a short surf session channeled
and quelled his energy so that “for some reason that time [in the water] made me feel
like I could do anything else,” including homework. Currently, he thought it might
help in being a competent father.
This “bigger part” of the stoke that resonated after the surf session reflected the
meditative or therapeutic aspect of surfing that was appealing to surfers. Once
proficient at the activity, the surfer experience generated positive feelings. José,
for example, said that “it makes me feel really calm; like It really clears my head. I
cannot explain it.” Likewise, John explicitly stated that “it’s like the best therapy
there is.” David admitted, “I don’t meditate, but I’m basically meditating [when I
surf],” which included stretching and breathing exercises intended to calm and relax
him. He compared surfing to other physical activities he did when he was younger
and said surfing was set apart because he felt like it “cleanses [his] soul.” “Maybe it’s
the exercise in combination with the ocean that elevates the experience,” as Jeff
suggested. Thus, the feeling of “being stoked” in the moment or after the surf session
reflected a positive experience that spilled over to feeling good about oneself.
A Salient and Prominent Surfer Identity
Identity salience refers to the likelihood of an identity being invoked in a situation

(Burke & Stets, 2009). The surfer identity was salient to surfers. For example, all
participants remarked that at one point in their lives they were surfing at least several
days a week, for several years in a row. John described this aspect of being a surfer,
reflecting its salience:
I lived with my cousin for about three years, and we would surf five days a week. If the
waves were breaking, we were surfing. . . . It has to be a constant running theme throughout
your daily and weekly life. It’s something that you have to be doing constantly and is always
on your mind that you’re obsessing over.
Similarly, Sean recalled that “when I wanted to learn, I wanted to learn. So, I
started going almost every day, definitely for a couple years at least.” José also
recalled that when he first started surfing, he would go every day before work, and
“say it was my day off, though, I was surfing 7 in the morning to 5 PM.” This was a
consistent remark from participants, that is, that they would go surfing every day or
many times a week, for years at time.
Every participant I interviewed reflected the necessity of these criteria to be a
surfer. Yet, at the same time, no one wanted to be the gatekeeper of who could and
could not call themselves a surfer. In their own mind, however, there were still
definitive criteria that they had to meet before they considered themselves a surfer.
John described this dynamic between his working definition of a surfer, while also
being careful not to be a gatekeeper to the identity:
You have to go through a period of your life where you put in that drive, and you go five
times a week, and you go double sessions, and you progress and reach that next level. If you
don’t ever leave that like “oh I’m having fun on a longboard” level, then you know you’re
having fun. That’s great. . . . [But] you know it in yourself when you reach that next level.
What seemed to set surfers apart from people who were just out there having fun
was their dedication to surfing. In other words, what they felt set them apart from
others who might surf more casually was the salience of their surfer identity. Colin
described in more depth why surfing regularly, or maintaining a salient surfer
identity, was important to him being a surfer:
A surfer surfs, or at least did at one point. That’s one criterion is that you surfed regularly at
one point. You have a fundamental understanding of the ocean current and tides, of wave
shape, wave dynamics, the general unwritten rules, the culture, and how to respect other
surfers.
Stating that surfing regularly was a criterion for being a surfer, he explained why
it was important. It had to do with understanding ocean currents and tides, wave
dynamics, and the unwritten rules of the culture, which included respect. There were
other parts of being a surfer, and those things could only be learned through repeated
166 A. Quinn
experience. Like other participants, he staked his definition of himself as a surfer on

its salience. He later stated that his definition of a surfer applied only to himself,
explicitly stating that he “did not want to be a gatekeeper” to the surfer identity. It
was striking that across participants, their definition of a surfer was someone who
prioritized that identity for years. Yet, at the same time, they were careful to say that
it was just what it meant to them personally.
In addition to the surfer identity being a salient identity to surfers, it also was a
prominent identity. Whereas salience refers to the likelihood of playing out an
identity within and across situations, identity prominence refers to how important
that identity is to an individual (Burke & Stets, 2009). Every participant revealed that
the surfer identity was not only salient but also prominent. Daniel described the
surfer identity prominence:
There is something about surfing that draws up from within you, becomes a part of your life
where it is almost necessary. When I catch a good wave, and I haven’t surfed in a while, it
feels like I just remembered a fundamental thing to my own existence that I’m back in
touch with.
Earlier, participants described the necessity of going several days a week for
years. However, various life circumstances could prohibit surfers from doing this
their whole lives, both before and after becoming a father. It also should be noted that
several participants stated that they would never move away from the beach or stop
surfing, even after becoming a dad. What Daniel was describing was that there were
points in his life when he was not surfing as much, yet he still described it as
something necessary in his life.
Similarly, Jeff recalled when he was working on his doctoral dissertation that
he could not surf for a few years. His dissertation was about surfing, however, so he
recalled being “mentally immersed but not physically immersed” in surfing. As he
reflected, he recalled feeling “untethered from himself” because of this. His words of
feeling “untethered from himself” due to a lack of surfing, echoed Daniel’s words of
surfing being “a fundamental thing to [his] existence” that he remembered when he
surfed again.
These kinds of poetic expressions were common among participants. Brian stated
that being a surfer “is woven within the fabric of my being. It’s a big part of who I
am, definitely.” Mike stated that being a surfer was “when the stoke becomes a part
of you,” reflecting how important, or prominent, the identity becomes. José said that
“being a surfer is just part of who you are, every surfer just feels it.” John agreed,
saying that “I felt almost instantly that [surfing] was what I was looking for, I just felt
so stoked. I assume that everyone else feels the same thing.” For surfers, the surfer
identity becomes a defining aspect of who they are, demonstrating how prominent
the identity is for them.
In a similar way to how participants described the salience of the surfer identity,
they described its prominence. And, in a similar way to not wanting the salience of
the surfer identity to act as a form of gatekeeping, they did not want the prominence
of the surfer identity to be a form of gatekeeping. Joel described this dynamic,
refusing to define himself as a surfer “until it becomes a part of your life, like
essential to your being, that’s what being a surfer means.” However, he was quick to
point out, “that’s for me, though. Other people will tell you different, and that’s OK.”
Joel described how prominent the identity was to him, but was quick to add the
caveat, demonstrating that he did not want to be a gatekeeper to the identity. It is
striking that he conceded that “other people will tell you different,” yet none of the
participants actually did tell me something different.
For these surfers, the surfer identity was both salient and prominent aspects of
their lives. The salience was reflected in the frequent enactment of the surfer identity.
Each participant remarked that at one point in their life they were surfing several
times a week. While some surfers were able to maintain this salience longer than
others, its prominence did not wane. This was reflected in the comments that
demonstrated how important the identity became to these individuals. Because the
identity was both salient and prominent, it had an effect on some of their obligatory
identities. The surfer identity helped shaped the self in important ways.
The Influence of the Voluntary Surfer Identity on Obligatory Identities
Because the surfer identity was salient and prominent to these surfers, it became a
potent force for shaping the self. This is interesting because as a voluntary identity, it
helped shaped several of their obligatory identities. For participants, the surfer
identity had a noticeable impact on their worker identity and their spouse identity,
which are both obligatory identities. Every participant remarked that at some point in
their life, they explicitly chose their job based on proximity to the ocean and flexible
working hours to facilitate their surfing habit.
Gary, for example, stated that he lived on the West Coast because of surfing, and
explicitly chose to live in a community where he could work in the service industry
in the evening: “I live in PB so it’s really more of a party community than a surf
community. Which is actually great because then I can bartend in the evening and
surf during the day.” Likewise, Sean talked about his job as a way of “shirking
responsibility and being able to surf as much as I wanted to. I had the perfect job for
that.” Sean also shaped his work identity in another way as a result of his surfing
identity. When asked how surfing has impacted his life, he reflected:
If it wasn’t for surfing, I’d probably still be with that chick I was with in the Navy. Probably
would have stayed in the Navy [pause]. . .which I would be retired right now [laughs]. Some
buddies I was with in the Navy were stationed out here at one point, and because they knew
me, were like, “just go, you got to go check out OB.” So, I was like, alright. Checked it out
and have lived here ever since for like 15 years.
In addition to describing his service industry job as a way to facilitate surfing as

much as possible, he recalled that surfing was the impetus for him to leave his former
job in the Navy. It also is noteworthy that surfing not only drove him away from his
worker identity in the Navy, but it also drove him away from his girlfriend at the
time, demonstrating that the surfer identity had a profound effect on these obligatory
identities.
168 A. Quinn
Some participants surfer identity shaped their worker identity in even more
profound ways. Brad, Daniel, Jeff, and Mark all pursued and obtained a doctorate
degree with the explicit goal of researching surfing or other things aquatic. Brad and
Daniel both became marine biologists because of their love for the ocean and to
make a career based on proximity to the ocean. Mark is a professor. When pursuing
his doctorate, he told me that he only applied to schools near the ocean. Surfing first
shaped his student identity, and he wrote his master’s thesis and doctoral dissertation
on surfing related topics. Although he no longer researches surfing for various
reasons, he told me that his job as a professor provided extreme flexibility to allow
him to surf, which was the main reason he wanted to work in a university. Arguably,
the surfer identity had the most impact on Jeff’s worker identity:
Now I get paid to think about surfing; it’s my job. I had to invent my job as a field of study,
looking at surf tourism. But that’s the kind of depth of effect it’s had on my life. It’s pretty
much defined and dictated what has happened to me since 1996.
Jeff still works as a professor at a university that is close to the ocean, and remarks
that his job flexibility facilitates his surfer identity.
In addition to the surfer identity effecting the worker identity, many participants
recalled that their surfer identity also had an effect on shaping their spouse identity.
For some, surfing was explicitly how they met their spouse. For others, their spouse
understood that the surfer identity was and is important in their lives. Most partic-
ipants remarked that, had their spouse not understood that part of them, their
relationship would not have worked out. Even though they worked together, Rick
recalled how it was a surfing trip through Australia that helped him meet his spouse:
She was really excited when she found out I surfed. It turned into this thing, like I wasn’t just
this annoying creepy guy to her anymore [laughs]. She surfs too, but it wasn’t even like that
was her main thing, like she doesn’t even surf much. She’s not bad though, and that caught
me hook, line, and sinker.
Rick recalled that “surfing has given [him] a lot,” including many of his current
friends and his eventual wife. They bonded over their shared identity as surfers. Sean
similarly admitted that surfing “shaped the direction of [his] life completely,” and
that he would never have met his wife, and therefore not have his son, if not for
surfing. These kinds of statements from participants were not uncommon, as they
emphasized the importance of their spouse’s understanding of their surfer identity.
Mark reflected that surfing was a “restorative” activity for him that had the
capacity to “transform [his] mood from grouchy to happy.” He concluded that his
“wife understands this and is very supportive of [him] surfing as frequently as
[he] can manage to do it.” David likewise reflected that surfing “is so important to
my mental and physical health and my wife understands that.” He then provided an
example of a stressful day, and his wife “stopped everything” and told him that he
needed to leave and go surfing. Upon coming home from surfing, they “started over
fresh, and felt great [laughs].” José told me that although his wife did not surf, he did
wish at one point while they were dating that she would have tried to get into
it. Although he was disappointed that she did not want to, he said it was “fine as long
as she understands that it’s something important to me.” Recalling this time in their
lives, he admitted that if she did not understand that part of him then “it would not
have worked out between us.”
While some surfers explicitly met their spouses through surfing, each participant
at the very least admitted that the surfer identity was such an integral aspect of their
lives that it was important that their spouse understood that part of them. In some
way, the surfer identity had an effect on their spouse identity. By extension, this
indirectly effects their father identity.
Because the surfer identity was salient, at least at some point in their lives, and
because it remained prominent, it shaped other identities or other aspects of them-
selves. This is particularly remarkable because, as a voluntary identity, surfing is not
an identity that holds the same societal significance as a worker or spouse identity.
Yet, the surfer identity for participants was such a potent force in their lives that it
had observable effects on their worker and spouse identities in myriad ways. In many
ways, their surfer identity took priority over and shaped their worker and spouse
identity. These are powerful examples of how a voluntary identity can interact with
obligatory identities and shape the self. In the following section, participants
described the effect the obligatory identity of father had on them.
The Father Identity
At the time of the interview, no participant’s child was more than five years old.
Eleven participant’s children were only three years old or younger. For each
participant, then, they had spent significantly more of their lives enacting the surfer
identity than they had the father identity. As a result, participants were still internal-
izing what being a father meant to them. However, these men, without exception,
expressed how important and salient the father identity was for their lives. Although
they were still making sense of being a father, becoming a father was an impetus for
profound identity change. Identity change can be observed in the restructuring of
their identities. The father identity was prioritized over the surfer identity. The surfer
identity prominence, however, did not wane.
Identity Change: The Effect of the Father Identity on the Surfer Identity
Colin described the unique dynamic of becoming a parent, explicitly referring to

how the identity was obtained almost instantaneously:
Having a kid; it’s all of a sudden. You were never a father before. It was zero part of your
identity, and now all of a sudden, it’s your main thing. Not through years of practice, like
surfing, not through years of building a career, but instantly. You’ve got nine months to prep
for it, but still you’re not prepared for it and instantly it becomes your number one thing, by
necessity. I mean, not for all fathers maybe, but for me it was.
Jeff described changes within himself:

170 A. Quinn
It has re-defined who I am and what it’s all about. . . . I think I’m reasonably good at it, I pride
myself on it a bit. I never miss a doctor appointment or dentist appointment or any
opportunity to do stuff with them. I think that it is cool. So, it is a very major part of my
identity these days, and one that I relish.
These tangible expressions of acting out the father identity revealed its salience.
Calling it a “major part of [his] identity,” similar to how Colin referred to fatherhood
as his “number one thing,” reflected the father identity being prominent, as well.
Gary presented an interesting case among participants because of his parental
status. He shared custody with his son’s mother and did not regularly enact the father
identity until a few months into his son’s life. However, he quickly found that he was
“stoked” on being father:
It was not something that I was looking forward to at first because his mom and I were never
a couple. It was three or four months before I actually met him, so I was able to ease my way
into it, and now I get to spend a lot more time with him. I’m a lot more stoked on it. . . . On
the days that I have him I don’t surf, which is fine because I’m stoked on just spending time
with him. It’s definitely more of an addition than a burden.
Because Gary was not married, he could not claim the spouse identity, nor did he
enact his father identity with as much regularity as other participants. He also was
able to continue to surf more frequently than other participants. Nonetheless, he
found himself “stoked” on being a father when he had custody. On those days, his
father identity was salient. He found that it added more to his life than it took away.
As a point of contrast with Gary, Mike was a full-time, single father. Because of
his status as a single father, Mike surfed less than other fathers in this research. He
reflected on how this dynamic changed his identity and his outlook on life in general:
Being a dad really helps me to be able to look at everybody in the world as the same as her.
We are all just trying to find love, trying to find belonging, all just trying to find our own
stoke. . . . For the first time in my life there is something that’s more important than surfing.
[Pause] . . . That’s what has changed. Everything is still the same, surfing just isn’t the most
important thing in my life anymore.
Mike’s perspective was important because he surfed less than other fathers in this
research, and due to his status as a single father, there could certainly be some
resentment that he felt at not being able to surf very often anymore. Instead, he
seemed to be even more introspective about what being a father meant, and how it
changed him. As a single father, his father identity was salient, but he also acknowl-
edged that being a father was more important or prominent than being a surfer.
Part of what it meant to be a surfer changed for these fathers. Additionally, many
participants reflected on the influence becoming a father had on their surfer identity
in another way. They were aware that their priorities changed; the salience of their
identities changed. Rick, for example, said this:
I judge myself more now for my actions, whereas maybe before I didn’t care as much. The
days of getting, like, blackout drunk are gone. Now I’m on call 24 hours a day. The days of
just disappearing, like on a surf trip, those days are over. . . . I’ve never been away, not even
one night since he was born. I look forward to the routine that we have every day and night. I
get so into it. I feed him, give him a bath, read him a story, get him to bed.
Not only did Rick enact the father identity in tangible ways, he seemed to prefer it
and found joy in it in a similar way that Jeff “relished” being a father. The fathers in
this research were acutely aware of the identity change that occurred after becoming
a father.
Joel made a similar realization:
My identity as a surfer takes a lower spot, it’s not as big a part of my life as it was before. It’s
still very important, very fundamental to what I am, but something else has taken its place.
Like I can’t surf whenever I want. I would like to, but I can’t. If I have to pick up my kid from
daycare, that’s what I have to do. I can text my wife and be like, “Hey the surf is really great
today, can we switch, and you pick him up today?” And if you can maintain that balance,
that’s great. If the answer is no, then I pick him up. Of course, I’m going to pick him up, and
it’s great.
As discussed earlier, the surfer identity was prioritized in these individual’s lives
for several years. It was salient. It also was prominent. Because of this, it had the
capacity to influence other identities that were obligatory. Once these individuals
became fathers, however, the father identity was prioritized over the surfer identity,
which demonstrated identity change in two ways. First was the acquisition of a new
identity, the father identity. Second, the father identity became more salient com-
pared to the surfer identity. However, the surfer identity retained its high promi-
nence. The father identity produced another change in these surfer-dads. This was
seen in the effect that the father identity had on another obligatory identity, the
worker identity.
Identity Change: The Effect of the Father Identity on the Worker Identity
Whereas earlier we saw the effect that the surfer identity could have on the worker
identity, the father identity likewise had an effect on the worker identity. The surfer
identity had the capacity to shape the worker identity. However, the prominence and
salience of the father identity was potent enough to not only influence these
individual’s surfer identity, but also their worker identity. The change in the worker
identity was not consistent among participants. There was a marked difference in the
how the worker identity changed as result of the father identity, and this difference
was a function of the participant’s socioeconomic status when they became a father.
Fathers in lower socioeconomic positions tended to re-prioritize their worker iden-
tity, while fathers in higher socioeconomic positions tended to de-prioritize their
worker identity.
Brad, a marine biologist, exemplified a father in a higher socioeconomic status:
I kind of pause to reflect on what I’ve down with my life, my career, and ask was it
meaningful? Did I want to do more with [my career]? That stuff matters a little less because
you’ve got little awesome humans that you’re charged with sculpting into good big humans,
and that’s the biggest job I’ll ever do; the most important one.
172 A. Quinn
Because of the prominence of his father identity, Brad found that his worker
identity became less prominent. Likewise, Jeff, a university professor, experienced a
similar shift in his worker identity:
Some of my professional ambitions have toned down a little bit. There is another part of my
life for self-satisfaction and self-esteem to come from. So, I am definitely less invested in my
job than I used to be. I view myself as a dad now, and not an academic, and that is probably
the biggest source of identity change. If I meet someone, I’m showing them pictures of my
kid on a surfboard; I’m that guy. [shows me photos].
In contrast to Brad and Jeff, Daniel and Mark, another marine biologist and
professor, respectively, never provided any indication that their worker identities
were ever all that important to them. They both relayed that their interest in their
careers had more to do with placating their surfer identity by ensuring flexible work
schedules and proximity to water. After they became fathers, however, their worker
identity did not increase in prominence. They also represented higher socioeconomic
status fathers.
Mark recalled that part of the reason he wanted to work as a professor was for a
flexible work schedule and less obligations to be on campus during the summer. This
was for the purpose of being able to travel to “fun or exotic surfing destinations.”
However, now that he was a father, he admitted to doing “less of that, now that time
off is spent with my family.” I got the sense that his worker identity was never all that
prominent for him, and after becoming a father it still was not prominent. The
biggest change was reflected more in how it affected his surfer identity, as he used
his time off and flexible schedule to spend time with his wife and child. Similarly,
Daniel refers to work being a “respite” from the fun but exhausting responsibilities
of fatherhood that dominate evenings and weekends.
Another participant, Jason, transitioned to a career in real estate shortly before
becoming a father. However, his wife stayed at home with the children. He relayed
how important his job was for him and his family then, as the sole provider. In
clearer examples of how their socioeconomic status interacted with their father
identity to increase the prominence of their worker identity, those in service and
hospitality found their jobs to increase substantially in prominence. Sean, for
example, who earlier reflected that he would be retired by now if he had stayed in
the Navy, said this about his job in hospitality:
The most important thing I want to do is when he gets in school is to make some more
goddamn money, because I’ve been making the same amount of money over time. [Pause]
. . . I might not make that much money, but childcare is crazy. I have the exact schedule I
need to reflect my wife’s schedule so that we don’t pay for childcare.
The father identity not only produced an identity change through decreasing the
salience of their surfer identity, but it also produced identity change through chang-
ing the prominence of their worker identity. For some fathers, their worker identity
prominence decreased or stayed low in prominence. This was a function of their
relatively higher socioeconomic status before becoming fathers. By contrast, some
fathers found that their worker identity prominence increased. This was a function of
their relatively lower socioeconomic status before becoming fathers.
Stoke by Proxy: Selfishness vs. Selflessness
During the interviews, I was continually struck by how involved these fathers were
in the lives of their children. This did not occur begrudgingly, either, as participants
frequently used words like “stoked” to describe how they felt about being fathers.
Participants seemed genuinely engaged in the lives of their children. They reflected
that being a father entailed being selfless. They contrasted this with being a surfer,
which to them was selfish.
Because they were able to be selfish in their surfer identity, they likewise were
able to embrace the selflessness that fatherhood demanded. This dynamic revealed
itself in the common language of surfing being a selfish activity and fatherhood
being selfless. Brad gave some insight into why surfing was selfish:
It’s a selfish sport because it’s a competition for a limited resource, waves are a limited
resource. Have you ever heard snowboarders or skiers say, “no friends on a powder day?”
It’s like that. I don’t even go with friends when the surf is good.
Along the same lines, Joel said this:

“Surfing sucks, don’t try it” [laughs]. Ever seen that bumper sticker? I hope to get bit by the
bug again, but whenever I would have the chance to surf, my wife and I would both have to
be off [work]. But then that’s so rare that I’d rather go get lunch together. Surfing is a selfish
thing. Now that I have a family, it’s even more selfish.
Colin found that part of the selfishness had to do with leaving his worries behind
when he surfed:
Surfing gives you the ability to switch off of normal life. There’s a bit of a selfish element to
it where you’re not worried about other things or people or this or that. Now it’s different
when I surf.
Colin found that now that he was a father, the selfishness of surfing changed
because of his son. He was not able to “switch off” because his son was back on
shore now. Jason also called surfing selfish because of how much it “requires time
and involvement.” He admitted that “fatherhood makes that more difficult,” to find
the time. He was okay with giving up the selfishness of surfing for fatherhood. In a
way, he positioned the identities, surfer and father, as opposites.
Sean made the connection between fatherhood and surfing more explicit:
No, they are [completely] opposite. One is super selfish, and the other is totally selfless. I had
always played team sports, too, and surfing is the one, true selfish thing I’ve ever had. I see
them very opposite.
Similarly, Rick reflected on the type of father he would like to be and referenced
his own father as an inspiration. He referred to his father as “a very honest, selfless
person,” and concluded that this was what he “would like to be for [his] son.” Rick
elaborated on some of the aspects of fatherhood that made it selfless:
It means to me that I have to always do everything right, so there is no alternative to doing
the right thing every time, right now. I have to give him a good example. I have to make sure
that whatever needs he has come before mine, and I have to make sure that he’s always OK. I
174 A. Quinn
think it’s very important to do the right thing every single time right now. Whereas at an
earlier point in my life, maybe it wasn’t as important.
Jeff is able to articulate this dynamic as well:

They [surfing and fatherhood] are both satisfying, but in pretty different ways. Like it’s a
good escape from the day-to-day, or it’s a catalyst for similar kind of self-reflections. Aside
from that, surfing is pretty selfish.
However, Jeff identifies one exception where surfing is not selfish:

Except when I go surfing with my daughter, then it isn’t [selfish] at all. I’m not even surfing
most of the time; I’m out in the water on my feet helping her. That has been an interesting
endeavor because I think I actually enjoy it more. It’s like rediscovering the sport from the
ground up. Seeing her improve and knowing how amazing it is to stand up and ride a wave.
Seeing her do that for the first time, I’m stoked by proxy.
Jeff demonstrates what happens when the two identities collide, or when both
identities are highly salient at the same time: the father identity prevails between the
two. And, what’s more, Jeff “actually enjoy[s] it more” than when he surfs by
himself. So, while surfing provides participants with an experience of the stoke,
part of what the father identity brings in return is a “stoke by proxy.” The surfer
identity enhances the self because it is selfish. The father identity enhances the self
precisely because it is selfless.
Discussion
The activity of surfing is an objectively difficult activity. Because voluntary identi-

ties are easily abandoned, surfing requires a certain amount of willpower to stay with
it and achieve proficiency. The “stoke” that surfers felt, encouraged a salient surfer
identity. Although participants did not want to be gatekeepers to the surfer identity, a
salient surfer identity and dedicating themselves to be skillful surfers appeared to be
criteria they imposed on themselves to be considered true “surfers.” Because of the
benefits they received in the surfer identity, it also became a prominent identity. It
was an important and integral part of themselves. The surfer identity was both a
salient and prominent force in surfers’ lives, and therefore had the capacity to shape
other aspects of the self. This was most clear in participants’ descriptions of the
effect being a surfer had on their worker identity and their spouse identity, both
obligatory identities. In some instances, participants prioritized their surfer identity
relative to their worker identity and spouse identity. At the very least, their surfer
identity helped shape and guide their worker and spouse identities.
Becoming a father is experiencing identity change due to the acquisition of a new
identity (Burke & Stets, 2009). As participants described it, this was a sudden and
important acquisition. When the surfers became fathers, I was continually struck by
their dedication and enthusiasm towards their father identity. Frequently, they
referred to being “stoked” about being a dad or used similar language to describe
how they felt as fathers that they had used as a surfer. They conveyed a sense of
pride, responsibility, and importance in the father identity. Thus, the father identity
was prominent. Moreover, these surfer-dads routinely expressed the tangible
involvement they had in their children’s lives. They frequently enacted the father
identity in their daily lives and were heavily involved in the raising and caring of
their children. As research by Collett et al. (2015) emphasize, it is crucial for fathers
to elaborate beyond wanting to “be there” for the kids by articulating how they are
involved in their children’s lives. The surfer-dads in this research routinely articu-
lated frequent involvement in the lives of their children. Thus, the father identity also
was salient to them.
In addition to identity acquisition, becoming a father was the source of another
kind of identity change. The father identity influenced a noticeable reduction in the
salience of the surfer identity. While participants indicated that being a surfer was
still a prominent part of their identity, they found that being a father was more
important. Essentially, the father identity was prioritized over the surfer identity.
Rather than dwelling on the “loss” of their surfer identity, participants tended to
relish the addition of their father identity and the accompanying change.
One area that participants differed was how the acquisition of the father identity
changed their worker identity. For those who enjoyed a relatively higher socioeco-
nomic status, they found that their worker identity either declined or stayed relatively
lower in importance. By contrast, those who were relatively lower in socioeconomic
status tended to increase the importance of their worker identity. This is an under-
standable trend. Those fathers who were more financially secure and stable prior to
becoming fathers did not have to worry as much about their families’ financial
future. Fathers who were not as financially secure or stable became more concerned
with their worker identity. This was a function of becoming a father and the sense of
responsibility they felt. Their worker identity was no longer just about themselves, or
even themselves and their spouse. The importance of their father identity coupled
with their relative financial instability made their worker identity more important to
them. The change in their worker identity, in either case, was a function of the
acquisition of their father identity as it interacted with their socioeconomic status.
The two identities, the voluntary surfer identity and obligatory father identity,
complemented each other. As the surfer-dads in this research indicated, the identities
complimented each other with each fulfilling different, yet important needs for
individuals. The essential difference between the two was that one met one’s own
needs and thus was seen as “selfish,” while the other focused on the needs of another,
one’s child (and often spouse by extension), and therefore was perceived as “self-
less.” The distinction between selfishness and selflessness should not be generalized
as a distinction between voluntary and obligatory identities. This distinction is
simply relevant to the surfer and father identities. The language of selfishness and
selflessness to describe the surfer and father identities, respectively, came from
participants, themselves.
By experiencing selfishness through the surfer identity and for several years,
individuals may have been ready to take on the rigors of a more selfless identity such
as being a father. The selfishness of one identity is complimented and balanced by
the selflessness of another identity. The essence of both words (selflessness and
176 A. Quinn
selfishness), like the essence of both identities, is the “self.” Participants expressed
the importance each identity had for their personal development. The implications of
the balance between selfishness prior to fatherhood and selflessness in fatherhood is
significant. As a society, we should be concerned with the involvement of fathers in
childrearing. To increase father involvement, we might want to encourage self-
development through activities that contribute in meaningful ways to one’s sense
of self prior to becoming fathers. Although this research focuses on the surfer
identity, this activity need not be surfing.
It has been suggested that voluntary identities may fulfill aspects of the self that
are not fulfilled by obligatory identities (Stets & Serpe, 2013). I do not think the
selfish/selfless distinction applies uniformly to the voluntary/obligatory distinction.
However, the idea that some identities are more likely to meet one’s own needs while
other identities are more likely to meet the needs of others may provide insight into
what it is that identities provide for individuals and society, more generally. It is
important to act in the service of others, but we also need to ensure that we have
taken care of our own needs. At issue is how we can balance the two carefully and
responsibly.
While this study only examined two identities, one voluntary and one obligatory,
there could be other identity processes or experiences intrinsic to other unexamined
identities. Moreover, although qualitative research provides the benefit of eliciting
rich narratives and allows for detailed probing questions, the nature of a relatively
small sample size (N ¼ 15) limits the generalizability of this research. Notwith-
standing these limitations, the dynamic between voluntary and obligatory identities,
including how they interact with and shape each other, should continue to be
examined to better understand these identities, and how they shape the individual
and others.
Acknowledgements I would like to acknowledge the UC Riverside Social Psychology Seminar,

and Dr. Jan Stets, in particular, for comments on earlier drafts. I would like to dedicate this chapter
to my daughter and wife. This research would not be possible without them.
Appendix: Participant Demographics
Age of
child Parental Years
Respondents Race/Ethnicity Age Occupation Education (ren) status surfed
David White/Caucasian 33 Service High 3 years; Married 18
School 1.5
years;
1 month
Brian White/Caucasian 34 Construction Bachelor’s 1 year Married 16
Mike White/Caucasian 36 Service High 5 years Single 30
School Father
(continued)
Age of
child Parental Years
Respondents Race/Ethnicity Age Occupation Education (ren) status surfed
Jose Latino/Mexican 26 Service Some 3.5 years Married 6
College
John White/Caucasian 26 Service Bachelor’s 1 year Married 11
Colin White/Caucasian 29 Real Estate Master’s 2.5 years Married 25
Sean White/Caucasian 42 Service Bachelor’s 4 years Married 24
Rick White/Caucasian 36 Real Estate Some 1.5 years Married 18
College
Jeff White/Australian 45 Professor Doctorate 5; 3.5 Married 40
years
Brad White/Hawaiian 48 Marine Bio. Doctorate 2.5 years Married 35
Jason White/Caucasian 45 Real Estate Bachelor’s 5; Married 30
3 years
Daniel White/ 36 Marine Bio. Master’s 1 year Married 15
Norwegian
Gary White/Caucasian 38 Service Bachelor’s 2 years Single, 20
Shared
Custody
Mark White/Caucasian 38 Professor Doctorate 1 year Married 30
Joel Latino/Mexican 28 Service Some 3 years Married 8
College
References
Adamsons, K. (2010). Using identity theory to develop a midrange model of parental gatekeeping
and parental behavior. Journal of Family Theory & Review, 2(1), 137–148.
Baxter, J., Buchler, S., Perales, F., & Western, M. (2015). A life-changing event: First births and
men’s and women’s attitudes to mothering and gender divisions of labor. Social Forces, 93(3),
989–1014.
Bleidorn, W., Buyukcan-Tetik, A., Schwaba, T., van Scheppingen, M. A., Denissen, J. J. A., &
Finkenauer, C. (2016). Stability and change in self-esteem during the transition to parenthood.
Social Psychology and Personality Science, 7(6), 560–569.
Booth, D. (2001). From bikinis to board-shorts: “wahines” and the paradoxes of surfing culture.
Journal of Sport History, 28(1), 3–22.
Burawoy, M. (1998). The extended case method. Sociological Theory, 16(1), 4–33.
Burke, P. J. (1980). The self: Measurement requirements from an interactionist perspective. Social
Cast, A. D. (2004). Well-being and the transition to parenthood: An identity theory approach.
Collett, J. L., Vercel, K., & Boykin, O. (2015). Using identity processes to understand persistent
inequality in parenting. Social Psychology Quarterly, 78(4), 345–364.
Collett, J. L., Vercel, K., & Pierce, K. D. R. (2019). The role of significant others in shaping fathers’
identities and behavior. In J. E. Stets & R. T. Serpe (Eds.), Identities in everyday life
(pp. 359–380). Oxford University Press.
178 A. Quinn
Daskalos, C. T. (2007). Locals only! The impact of modernity on a local surfing context. Socio-
logical Perspectives, 50(1), 155–173.
Emerson, R. M., Fretz, R. I., & Shaw, L. L. (1995). Writing ethnographic fieldnotes. University of
Chicago Press.
Evers, C. (2004). Men who surf. Cultural Studies Review, 10(1), 27–41.
Evers, C. (2006). How to surf. Journal of Sport and Social Issues, 30(3), 229–243.
Gallagher, M. (2016). Distinguishing obligatory and voluntary identities. In J. E. Stets & R. T.
Press.
James, W. (1890). The principles of psychology: Volume I. Dover Publications.
Kushida, M. (2019). The nature of the ballet and hip-hop dancer identity. In J. E. Stets & R. T. Serpe
(Eds.), Identities in everyday life (pp. 331–358). Oxford University Press.
Lawler, K. (2011). The American surfer: Radical culture and capitalism. Routledge.
Livingston, G., & Parker, K. (2019). Eight facts about American dads. Pew Research Center.
McLaughlin, K., & Muldoon, O. (2014). Father identity, involvement and work-family balance: An
in-depth interview study. Community and Applied Social Psychology., 24(5), 439–452.
Roberts, M. J., & Ponting, J. (2020). Waves of simulation: Arguing authenticity in an era of surfing
the hyperreal. International Review for the Sociology of Sport, 55(2), 229–245.
Saldaña, J. (2013). The coding manual for qualitative researchers (2nd ed.). Sage Publications Inc..
Smith, C. J. (2010). Surfing and social theory: The significance of surfing and its social contexts.
M.A. thesis, Department of Sociology, San Diego State University.
Stets, J. E., & Serpe, R. T. (2013). Identity theory. In J. DeLamater & A. Ward (Eds.), The
handbook of social psychology (pp. 31–60). Spring and Dordrecht.
Stranger, M. (2011). Surfing life: Surface, substructure and the commodification of the sublime.
Ashgate Publishing Company.
Thoits, P. A. (1992). Identity structures and psychological well-being: Gender and marital status
comparisons. American Sociological Association, 55(3), 236–256.
Thoits, P. A. (2003). Personal agency in the accumulation of multiple role-identities. In P. J. Burke,
T. J. Owens, R. T. Serpe, & P. A. Thoits (Eds.), Advances in identity theory and research
(pp. 179–194). Kluwer Academic/Plenum.
Thoits, P. A. (2012). Role-identity salience, purpose and meaning in life, and well-being among
volunteers. Social Psychology Quarterly, 75(4), 360–384.
Turner, R. H. (1978). The role and the person. American Journal of Sociology, 84, 1–23.
Chapter 9
Religious Identity in a Proximate Social
Structure: Mothers, Fathers,
and the Religious Socialization of Their
Children
Philip S. Brenner
Introduction
How do parents pass on who they are—their beliefs and values, norms and habits,
and the social groups with which they identify—to their children? That is, when
children grow up to be like their parents, to a greater or lesser extent and for better or
worse, how does it happen? Sociologists and social psychologists call this complex
process “socialization” (Grusec, 2011). Although the process is far from determin-
istic, parents raise their children to resemble themselves and children tend to grow to
resemble their parents, attending to the same norms, identifying with the same
groups, and sharing their parents’ beliefs, values, and habits (Bengtson et al.,
2013; Liefbroer & Elzinga, 2012; Min et al., 2012).
Some part of the socialization process is attributable directly to the actions of
parents, rewarding some behaviors and attributes and penalizing others (Grusec,
2011). But other parts of the socialization process are indirect, routed through and
rooted in the community in which the family lives (Martin et al., 2003). Parents and
their children live in communities comprised of networks of people who tend to be
alike, sharing the associations and values of the parents (Fletcher et al., 2000). These
communities of like individuals, called the proximate social structure, include
extended kin, members of the family’s religious group, schoolteachers and so on
that influence the socialization of children in alignment with the values of the
community (Bengtson et al., 2009; Gunnoe & Moore, 2002; Silverstein & Bengtson,
2018; Vedder et al., 2009). In transmitting these norms and affiliations to their
P. S. Brenner (*)
Center for Survey Research, University of Massachusetts Boston, Boston, MA, USA
180 P. S. Brenner
children, parents also socialize their children into important identities—links to

others through personal connections and group memberships that give their lives
meaning. An important function of these identities is to link parents and their
children to their communities (Stryker & Serpe, 1983).
Religiosity is one of many important identities that parents pass on to their
children (Bengtson et al., 2013). By instructing them in forms of religious worship,
telling them their community’s foundational stories, and teaching them the key
beliefs of their religious tradition (Dollahite & Thatcher, 2008; Marks, 2004),
parents attempt to mold their children into fellow believers and co-religionists who
share their religious identity (Myers, 1996; Sherkat, 1998, 2003; Snell Herzog &
Mitchell, 2016; Storm & Voas, 2012). Moreover, passing one’s religious identity on
to one’s offspring is important beyond the wishes and desires of parents. The
maintenance and growth of religious communities depends on the children of
adherents. Some religious communities do not evangelize and admit few, if any,
new members (DellaPergola, 2005; Lerer & Mayer, 1989). These religious commu-
nities depend solely on their children and would die out quickly without the
successful intergenerational transference of religious identities. Indeed, religious
history includes examples of religious communities whose members did not have
enough, or any, offspring to carry forward their faiths (Bainbridge, 1982; Brewer,
1984). Other religious communities, even those focused on evangelism, rely heavily
on the successful religious socialization of their own children (and a relatively high
birthrate of members) for the maintenance and growth of the community (Hout et al.,
2001; McQuillan, 2004; Perry & Schleifer, 2019).
The transference of religious identities from parents to children, or religious
socialization, has conventionally and traditionally been the domain of mothers
(Acock & Bengtson, 1978; Hayes & Pittelkow, 1993). Mothers read religious stories
to their children, show them how to celebrate religious holidays, and take them to
religious services (Boyatzis & Janicki, 2003). In so doing, mothers’ religious
identities influence those of their children (Pearce & Thornton, 2007). The role of
fathers in the intergenerational transmission of religious identities, however, remains
under-investigated and the extent and nature of fathers’ contributions, uniquely or in
conjunction with mothers, to their children’s religious identity development are not
well understood (Baker-Sperry, 2001).
While parents clearly have an important role in transmitting their religious
identities to their children, and many are able to successfully do so, the mechanism
by which this is accomplished remains under-investigated. That is, when parents
read Bible stories to their children, or teach them prayers and rituals, how are the
meanings of parents’ religious identities transmitted along with the story, prayer, or
ritual? This study considers the intergenerational transmission of religious identities
by examining mothers’ and fathers’ roles in the religious socialization of their
children.
Religious identities are conceptualized in terms of their subjective value to the
individual, propensity to be situationally called-up, and the frequency of behavior in
them. Although their terminology differs, each of these concepts can be found in
identity theory and in the sociology of religion and are used here in line with this
9 Religious Identity in a Proximate Social Structure: Mothers, Fathers, and. . . 181
extant research. The current investigation builds on the integrative work linking
identity theory and social psychological theories of the socialization of children
(Stryker & Serpe, 1983). This research novelly integrates socialization into identity
theory, illustrating intergenerational transmission of identity, using the religious
identity as an example. Religious socialization of children is conceptualized as set
of direct paths in a structural model between measures of each identity concept for
parents’ religious identities to those of their children.
Identity Theory
First, we must consider what we mean by an “identity.” In everyday English, as well

as in much of the sociological writing outside of sociological social psychology, the
word “identity” is used in multiple ways, many of which lack systematic,
operationalizable definitions (e.g., Appiah, 2018; Fukuyama, 2018). Thus, identity
theory was developed to clearly define and understand the functioning of identities:
how they are created and maintained, and how individuals use them and are
constrained by them in their everyday lives (Stryker, 1968; 1980).
Identities are an organizing feature of the self. They provide meaning for indi-
viduals by helping them to understand who they are, locating where they belong, and
letting them know what is expected of them by others. Three distinct forms of
identities construct and maintain the individual’s place in the world. First, identities
can be linked to group memberships such as race and ethnicity, gender and sexuality,
and religious and political affiliations. These group identities provide meaning for
the individual that are focused on connections to others based on shared social
characteristics (Hogg & Abrams, 1988; Hogg et al., 1995; Stets & Burke, 2000).
Second, identities also can be linked to role relationships such as familial, profes-
sional, and occupational connections. These role identities define individuals and
provide meaning for them in terms of the roles they perform in response to the
counter-performances of others (e.g., parents and their children; professors and their
students, and so on [Burke, 1980; Stryker 1980]). Third, identities can be linked to
deeply held personal characteristics. These identities, called person identities, are the
meaningful adjectives that individuals use to describe themselves, such as a moral
(Stets & Carter, 2011, 2012), introverted, or generous person. A single part of the
self, such as one’s religious self, may be categorized in one or more of these three
ways. For example, the individual’s religiosity may be reflected in one’s identity as a
Roman Catholic (a social identity), as a parishioner at St. Mary’s (a role identity),
and as a devout believer (a person identity).
To understand the functioning of identities, identity theory defines a set of key
constructs and their interrelationships. In addition to the ramifications of identity
enactment on the maintenance or change in the meaning of an identity (Burke, 1991;
Cast & Burke, 2002; Stets and Cast Stets & Cast, 2007), a primary concern of
identity theory is how individuals choose which identity they will enact within a
situation (Stryker, 1968; 1980). The first of these concepts is identity prominence.
182 P. S. Brenner
Although individuals are comprised of multiple identities, they are not necessarily
valued equally by the individuals who hold them. Some identities are valued, even
cherished, by individuals while other identities may be only mildly appreciated or
even disliked (Asencio, 2013; Asencio & Burke, 2011). The importance or subjec-
tive value of the identity to the individual, called identity prominence, is a key
construct in understanding how identities function. The higher the value placed on
an identity, the higher its placement in the prominence hierarchy, an ordering of
identities given their value to the individual (Stryker & Serpe, 1994; McCall &
Simmons, 1978). Religious identities, for those who have them, are typically highly
prominent as religious individuals tend to value their religious faiths and beliefs and
value being religious. The identity’s placement in the prominence hierarchy impor-
tantly influences the likelihood of its enactment; the more prominent the identity, the
more likely the individual will opt to enact it (Brenner, 2011a; Brenner et al., 2014;
Stryker & Serpe, 1982).
This second concept, the likelihood of defining a situation as one in which the
identity is relevant, is called identity salience. Within a given situation, individuals
can choose which of their identities they will call forward and enact (Stryker 1980).
Salience too is organized in a hierarchy, with those identities placed higher in the
salience hierarchy having a higher propensity of being deemed relevant for enact-
ment. Moreover, the more the salient identity, the greater the individual’s motivation
to define the situation as one in which an identity is relevant and may be enacted
(Stryker & Serpe, 1982, 1994; Brenner & DeLamater, 2014, 2016). In sum, identity
prominence influences identity salience and identity salience influences behavior in
the identity.
The extant research supports these theoretical suppositions for identities in
general (Brenner et al., 2014) and for the religious identity specifically (Brenner,
2011a). However, some amount of the relationship between prominence and
salience as measured on surveys is artifactual. The religious identity is viewed by
many Americans as strongly normative. Like other normative identities, the religious
identity has been prone to substantial bias when measured using survey self-reports
(Hadaway et al., 1993, 1998; Presser & Stinson, 1998). This phenomenon is
understood to be a function of prominence: as individuals place more importance
or value on their identity as a religious person they tend to exaggerate their
likelihood of performing that identity (Brenner, 2011b, 2016; Brenner & DeLamater,
2016).
Whether the relationship between religious identity prominence and religious
identity salience is real or artifactual, the higher the measured value of prominence of
the religious identity, the higher the measured value of salience. Secondly, by
definition, the higher the salience of the religious identity, the more likely it is to
be defined relevant for enactment in a given situation (Ploch & Hastings, 1998).
Finally, extant research demonstrates that identity prominence has a secondary but
direct effect on behavior in addition to its indirect effect through salience (Brenner,
2011a, 2016; Brenner et al., 2014; Stryker & Serpe, 1982).
Thus, the first three hypotheses reflect these well established relationships for
both parents and their children. The primary interest of this investigation is
Fig. 9.1 Heuristic model
estimating the contributions of mothers’ and fathers’ religious identity prominence,

salience, and behavior in the religious identity to those of their children. Because
structural models (to be discussed shortly) require explicit or implicit assumptions
about interrelationships between all variables, parents’ religious identity processes
must also be hypothesized. Excluding these paths could result in a specification error
that could bias estimated paths between parents’ and children’s religious identity
constructs. Thus, models assume relationships between parents’ religious identity
constructs for model fitting follow the primary tenets of identity theory.
Hypothesis 1. Each family member’s (mother’s, father’s, and child’s) religious
identity prominence positively influences their own religious
identity salience. (see Fig. 9.1).
identity salience positively influences their own behavior in the
religious identity.
identity prominence positively influences their own behavior in the
religious identity.
Whereas the first three hypotheses focus on the function of the religious identity
within the self, the next hypotheses concern the intergenerational transmission of the
religious identity. That is, how do parents’ religious identities influence their chil-
dren’s identities and how, if at all, do children’s religious identities affect their
parents’ religious identities?
184 P. S. Brenner
Parents raise their children to resemble themselves by attending to the same

norms, identifying with the same groups, and sharing their beliefs, values, and
habits, including religious norms, affiliations, beliefs, values, behaviors, and so
on. In so doing, parents’ religious identities influence those of their children. But
how can this influence be operationalized within the concepts of identity theory?
That is, to what extent and by what mechanisms do parents’ religious identity
prominence, religious identity salience, and behavior in their religious identities
influence those of their children?
A direct link is posited between parents’ and children’s religious identity prom-
inence. Children tend to value their religious identity to the same extent that their
parents do (Regnerus et al., 2004). But how does this intergenerational transference
of religious identity prominence happen? Parents inculcate their children into their
faith by introducing them to their religious community or ushering them into
membership in that community through baptism, confirmation, bar or bat mitzvah,
or the like. These actions integrate children into the religious community and instruct
them on the importance of these connections to others their community. Teaching
children to value these social connections with their co-religionists increases chil-
dren’s affective commitment, the value the individual places on connections to
others known through that identity; a cause of identity prominence (Stryker &
Serpe, 1994).1 By integrating children into a religious community and encouraging
them to value the connections with others in that community, parents may influence
their children’s religious identity prominence. Thus, the value that parents place on
their religious identities positively influences the value their children place on their
religious identities.
Hypothesis 4. Each parent’s (mother’s and father’s) religious identity
prominence positively influences their children’s religious identity
prominence.
In addition to instilling the value that parents hold in their religious identities in
their children, these moments of religious socialization also demonstrate the enact-
ment of religious identities for children. While mothers typically take the leading
role (Pearce & Thornton, 2007), both mothers and fathers provide models for their
children’s religious identity enactment. Parents show their children how and when to
perform their religious identities by telling them religious stories, demonstrating
rituals and styles of worship at religious services, and teaching them hymns and
prayers recited or sang in their congregation. But children’s religious identity
salience is also influenced through interactional commitment, the extensiveness of
the individual’s connections to others through an identity; a cause of identity
salience (Stryker & Serpe, 1994). By introducing children into a religious commu-
nity and ensuring their connections with others in that community, parents may
positively influence their children’s religious identity salience. Thus, we posit a link
between parents’ and children’s religious identity salience.
1
Unfortunately, measures of religious identity commitment are not available in these data.
Hypothesis 5. Each parent’s (mother’s and father’s) religious identity salience

positively influences their children’s religious identity salience.
Prominence is conceptually subjective—one values an identity to the extent that
they feel that they do. Thus, by definition, the individual is privy to the prominence
of their religious identity (and other identities), making prominence readily measur-
able: the value of one’s religious identity is what individuals believe it to
be. Salience, on the other hand, is conceptually objective although individuals are
not necessarily privy to their own propensity to enact a religious identity (Stryker &
Serpe, 1994). Moreover, the measurement of salience is typically complicated by the
use of subjective questions on surveys that ask the respondent to reflect on their
behavior in hypothetical situations, a violation of measurement best practices (Con-
verse & Presser, 1986). Using subjective questions to measure salience, an objective
concept, allows prominence, as its precedent, to bias its measurement (Brenner,
2011a). Therefore, an objectively measured outcome, actual behavior in the religious
identity, is included in this model as it avoids systematic measurement errors that can
be introduced in self-report measures of normative behavior (Brenner & DeLamater,
2016).
As families who attend religious services typically attend together, and children
tend not to attend without their parent(s), the direction of influence is from parents to
children (Flor & Knapp, 2001; Francis, 1993; Regnerus et al., 2004; Storm & Voas,
2012). Some part of this relationship is simply that parents decide what their children
will do and where they will go. Some part of children’s religious behavior is,
therefore, coerced. Some other part of children’s religious behavior, however, is
the outcome of socialization and the identity process. Children learn how to behave,
religiously and otherwise, first by imitating their parents, but later by replicating their
behavior either intentionally or habitually (Ploch & Hastings 1998). Thus, parents’
and children’s religious behavior are positively associated (Jennings et al., 2009).
Hypothesis 6. Each parent’s (mother’s and father’s) behavior in their religious
identities positively influences their children’s behavior in the
religious identity.
While mothers’ and fathers’ religious identities are likely associated, we should
not assume that one parent’s religious identity prominence unidirectionally influ-
ences that of the other parent given the use of cross-sectional data and a lack of
research or theory to suggest a causal relationship from mothers to fathers or vice
versa. Unidirectional paths from one parent’s salience or behavior to that of the other
parent are similarly unwarranted. Therefore, each of these constructs is allowed to
covary with the same construct in the other parent but no direction is posited for
these associations.
Thus far, the posited connections between parents’ religious identity prominence
and their children’s religious identity salience is indirect: either through the parent’s
religious identity salience or via children’s identity prominence. This implies that the
value parents place on their religious identities will influence the likelihood of their
186 P. S. Brenner
child enacting his or her religious identity by first influencing the value the child
places on his or her religious identity. But a direct connection may be needed. A
direct path from parents’ prominence to children’s salience implies that children’s
propensity to call up their religious identities may primarily exist to please parents
who strongly value their own religious identities.2 This path would allow a child to
engage in a somewhat perfunctory identity, calling up an identity without necessarily
valuing it, yielding a religious identity of relatively high salience compared to its
relatively lower prominence; a typical pattern for some obligatory or coerced
identities.
prominence is a positive predictor of their children’s religious
identity salience.
A return path that allows children to influence their parent is also plausible (Glass
et al., 1986; Stolzenberg et al., 1995; Stryker & Serpe, 1983). Research has
established that children’s religious identity development and involvement in their
religious communities may help to reinforce parents’ religious identity, as well as
that of other adults in their religious community (Gallagher, 2007). Although the
biblical passage in Isiah (11:6; “a little child shall lead them”) is commonly
misinterpreted, there is a long-noted association between parenthood and religiosity
for young adults (Stolzenberg et al., 1995; Wilcox, 2002). Children’s religious
socialization and the development of their religious identities may motivate
increased propensity of religious identity enactment for their parents. In this way,
the development of religious identity prominence of children could influence the
religious identity salience of their parents.
Hypothesis 8. Children’s religious identity prominence positively influences their
mother’s and father’s religious identity salience.
Finally, the causal relationship established in prior research for individuals
(Brenner et al., 2014) is novelly extended to dyads, estimating the influence of
each parent’s religious identity prominence on the other parent’s religious identity
salience. Although it has not been previously hypothesized or investigated, this
relationship is plausible given research finding couple’s mutual ability to influence
the values, religious and otherwise, of each other (Jennings & Stoker, 2001). A path
between mothers and fathers replicates a similar path between parents and their
children, allowing one parent to increase their propensity of religious identity
enactment in light of the value placed on it by the other parent. This association
may occur for various reasons, such as increasing one’s likelihood to perform their
2
This mechanism is arguably tapping into the child’s identity as a “good child.” It is possible that
some of the effect of parents’ religious identity prominence on child’s religious identity salience
could be explained by measures of child identity prominence, were they available in these data.
Indeed, identities exist in a dynamic ecology by definition as they interact as individuals decide
which to situationally enact and change over the individual’s life course (Smith-Lovin, 2003).
religious identity to please their partner independent of the prominence of their

religious identity.
prominence is a positive predictor of the other parent’s religious
identity salience.
Data and Methods
These hypotheses are tested using data from the 500 Family Study, a multi-mode
study of middle-class American families from eight communities from different
regions of the country (Schneider & Waite, 2008). This clustered design surveyed
multiple members of each family, including resident parents and their resident
adolescent children, ages 11 to 18. Given the potential for differences in their
roles, I include both mothers and fathers and estimate each of their contributions
to the religious socialization of their children. The analytic dataset includes
564 mothers, 453 fathers, and 465 co-resident children.
Analyses use measures of three key concepts reflecting the religious identity:
religious identity prominence, religious identity salience, and behavior in the reli-
gious identity. Religious identity prominence is measured using a single-question
ordinal variable: “Do you think of yourself as a religious person? No, not at all (0);
yes, somewhat (1); yes, very (2).” Mothers report a mean prominence level of 1.0,
corresponding to “somewhat” prominent religious identities. Fathers’ and children’s
mean prominence are somewhat lower than mothers (.88 and .81, respectively). (See
Table 9.1). While this is a somewhat weak measure of religious identity prominence,
it arguably reflects the construct reasonably well. Similar measures in the 2018
General Social Survey measuring the individual’s self-view as a “religious person”
and the subjective strength of religious affiliation are strongly correlated (ρ ¼ .65 and
.80) and reflective of a single underlying factor (α ¼ .86).
Religious identity salience is operationalized as a retrospective measure of
religious behavior in the recent past: “In the past year, about how often have you
attended religious services?” Responses were recoded to represent their likelihood of
attending in a given week: not at all (0); several times a year or less (652 ¼ .115);
about once a month (14 ¼ .25); two or three times a month (2.54 ¼ .625); about
once a week (1.0); and more than once a week (1.0). Mean religious identity salience
is relatively similar for mothers (.44), fathers (.40), and children (.48) (see
Table 9.1). While respondents typically report past religious behavior with bias, it
is arguably still a better operationalization of salience than other commonly used
measures (Brenner, 2011a). Such questions ask for anticipated behavior in hypo-
thetical situations which, paraphrasing Converse and Presser (1986), are likely to
garner hypothetical answers.
Behavior in the religious identity is operationalized as that reported using the
experience sampling method (ESM). In this application of ESM, respondents were
188
Table 9.1 Means/percents, standard deviations, and correlations

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1 Prominence, mother 1.00
2 Prominence, father 0.43 1.00
3 Prominence, child 0.40 0.32 1.00
4 Salience, mother 0.55 0.48 0.48 1.00
5 Salience, father 0.55 0.61 0.43 0.78 1.00
6 Salience, child 0.53 0.39 0.51 0.74 0.66 1.00
Behaviors
7 At church, mother 0.33 0.32 0.22 0.51 0.45 0.46 1.00
8 Religious event, mother 0.38 0.33 0.20 0.38 0.41 0.37 0.60 1.00
9 Relig. practice, mother 0.21 0.25 0.20 0.34 0.31 0.18 0.32 0.26 1.00
10 At church, father 0.26 0.35 0.23 0.40 0.48 0.42 0.53 0.42 0.16 1.00
11 Religious event, father 0.22 0.32 0.20 0.31 0.38 0.30 0.41 0.36 0.14 0.68 1.00
12 Relig. practice, father 0.14 0.28 0.12 0.20 0.27 0.18 0.12 0.19 0.26 0.26 0.23 1.00
13 At church, child 0.23 0.18 0.28 0.35 0.29 0.42 0.41 0.32 0.22 0.33 0.31 0.23 1.00
14 Religious event, child 0.19 0.17 0.22 0.24 0.19 0.26 0.31 0.33 0.16 0.27 0.29 0.17 0.68 1.00
15 Relig. practice, child 0.14 0.15 0.16 0.16 0.14 0.15 0.18 0.17 0.31 0.15 0.16 0.20 0.29 0.18 1.00
Mean/Percent 1.04 .88 .81 .44 .40 .48 29% 17% 11% 25% 18% 8% 24% 14% 4%
(standard deviation) (.63) (.72) (.66) (.41) (.40) (.43)
Note: Correlations between prominence and salience are Pearson correlation coefficients. Correlations between behaviors and prominence or salience are point-
biserial coefficients. Correlations between behaviors are phi coefficients
P. S. Brenner
asked to report their current activity at randomly-selected times when they were
notified by a beeper. Respondents were asked to record their location, main activity
(“What was the main thing you were doing? Please be specific.”), and secondary
activity (“What else were you doing at the same time? Please be specific.”). These
verbatim responses were coded into an extensive set of categories. Religious behav-
ior was operationalized as respondents reporting behavior categorized as participat-
ing in religious events and activities (coded as religious events, religious holiday
activities, other religious activity), devotional behavior (reading religious materials,
praying, and meditating), and other behavior reported as occurring in a religious
building. Across all types, 33 percent of mothers, 30 percent of fathers, and 25 per-
cent of their children, respectively, reported religious behavior (see Table 9.1).
The application of ESM is useful here because it avoids asking the respondent
direct questions about specific behaviors that may encourage an unwarranted nor-
mative response. Direct questions about religious behavior can be directive if they
motivate respondents to report inaccurately on their behavior if its prominence is
high and out of sync with its (relatively lower) salience (Brenner & DeLamater,
2016). Thus, rather than asking “did you attend religious services on that Sunday?”,
the ESM procedure asks respondents simply to note what they were doing when the
beeper sounds or vibrates. This procedure avoids measurement bias attributable to
the prominence of the identities upon which the respondent is reporting. Table 9.1
presents correlations for parents’ and their children’s identity constructs and behav-
ior in the religious identity.
Three demographic covariates are also included in these models given research
that demonstrates differences in the religious identity and religious socialization by
age (Erickson, 1992), sex (Hoge et al., 1982; Kieren & Munro, 1987), and race
(Demo & Hughes, 1990). Models include the adolescent child’s age in years
(mean ¼ 15, sd ¼ 1.9), sex (52 percent female), and race (80 percent white,
20 percent non-white [including 8 percent Black]).
Analysis plan. Structural equation models are estimated using full information
maximum likelihood estimation to retain cases with missing observations. This
estimation approach yielded an analytic sample of 642 observations within 516 clus-
ters (families). The variance-covariance matrix and standard errors are adjusted
given the clustered sampling design. Models test the hypothesized relationships
between religious identity prominence and salience and behavior in the religious
identity for parents and their children. Paths are estimated between parents’ and
children’s identity constructs while allowing parents’ constructs to covary. This
approach assumes that parents’ identities are a cause of their children’s, with one
route allowing children’s identities to influence those of their parents and another
route allowing parents to influence each other. Models also include three demo-
graphic variables and allows each to influence prominence, salience, and behavior
for parents and children. Demographics are allowed to influence parents’ religious
identity constructs given the potential for children’s characteristics to influence
parental treatment, such as differential treatment of sons and daughters and tailoring
of religious activity to be appropriate for the child’s age.
190 P. S. Brenner
Results
Reporting of results starts with the key associations hypothesized by the identity
model; the associations between prominence, salience, and behavior within each
family member. Religious identity prominence has positive associations with
salience for individual family members. Fathers’ religious identity prominence is
positively associated with their own salience (b ¼ .25, p .001), mothers’ religious
identity prominence is positively associated with their own salience (b ¼ .22,
p .001), and children’s religious identity prominence is positively associated
with their own salience (b ¼ .11, p .001; see Table 9.2).
Religious identity salience is positively associated with religious behavior.
Mothers’ and fathers’ religious identity salience is positively associated with their
own religious behavior (b ¼ .35, p .001; and b ¼ .39, p .001, respectively).
Children’s salience is also positively associated with their own religious behavior
(b ¼ .22, p .05).
Religious identity prominence is positively associated with religious behavior.
Mothers’ and fathers’ religious identity prominence is positively associated their
own religious behavior (b ¼ .08, p .05; and b ¼ .08, p .05, respectively).
Children’s prominence is positively associated with their own religious behavior
(b ¼ .07, p ¼ .07) although the effect is only marginally significant.
Next, the paths connecting parents’ religious identities to those of their children
are reviewed. Mothers’ and fathers’ religious identity prominence is positively
associated with the religious identity prominence of their children (b ¼ .28,
p .001; b ¼ .19, p .001, respectively). Mothers’ and fathers’ religious identity
salience is positively associated with the religious identity salience of their children
(b ¼ .51, p .001; b ¼ .15, p ¼ .076, respectively) although the latter is only
marginally statistically significant. Mothers’ and fathers religious behavior is posi-
tively associated with their children’s religious behavior (b ¼ .36, p .001; b ¼ .20,
p ¼ .09) although the latter is only marginally statistically significant.
Only one of the hypothesized cross paths from parent’s prominence to children’s
salience is found to be significant. Mothers’ religious identity prominence is posi-
tively associated with their children’s salience (b ¼ .08, p .01). Father’s promi-
nence is not significantly associated with their children’s salience. The return paths
from children’s prominence to their parents’ salience are found to be significant.
Children’s religious identity prominence is positively associated their fathers’
salience (b ¼ .12, p .001) and their mothers’ salience (b ¼ .15, p .001).
The final set of paths connect the religious identities of the parents. Fathers’
religious identity prominence is positively associated mothers’ salience (b ¼ .14,
p .001). Mothers’ religious identity prominence is positively associated with
fathers’ salience (b ¼ .15, p .001). The model also allowed the three religious
identity constructs to covary between parents. Mothers’ and fathers’ prominence are
positively associated (b ¼ .19, p .001), as are their salience (b ¼ .06, p .001),
and their religious behavior (b ¼ .05, p .001).
Table 9.2 Structural equation model and measurement model coefficients and standard errors
Prominence, Prominence, Prominence, Salience, Salience, Salience, Behavior, Behavior, Behavior,
father mother child father mother child father mother child
coef. p coef. p coef. p coef. p coef. p coef. p coef. p coef. p coef. p
Structural model
Prominence, father .19 *** .19 *** .25 *** .14 *** .00 .08 * .08 * .01
Prominence, mother .19 *** .28 *** .15 *** .22 *** .08 ** .01 .08 * .02
Prominence, child .12 *** .15 *** .11 *** .02 .05 .07
Salience, father .06 *** .15 .39 *** .07 .15
Salience, mother .06 *** .51 *** .07 .35 *** .01
Salience, child .14 .09 .22 *
Behavior, father .05 *** .20
Behavior, mother .05 *** .36 **
Control variables
Race (white) .32 ** .23 * .04 .06 .03 .09 * .08 .12 * .01
Age, years .02 .02 .02 .02 * .01 .02 *** .01 .01 .01
Gender (female) .01 .04 .09 .02 .04 .03 .00 .03 .02
Measurement model
At religious place 1.00 1.00 1.00
Relig. event, holiday .73 *** .70 *** .60 ***
Pray, meditate, read .20 ** .31 *** .14 ***
9 Religious Identity in a Proximate Social Structure: Mothers, Fathers, and. . .
Notes: Covariances in italics

*p .05
**p .01
***p .001
191
192 P. S. Brenner
Structural equation models are defined as much by the hypothesized, estimated

paths as by those assumed to be zero. By estimating these other possible paths, the
robustness of the hypothesized model can be tested. The first set of possible paths
allowed mothers’, fathers’ and children’s religious identity salience to influence each
others’ behavior in the religious identity. These six paths failed to yield any
statistically significant or marginally significant effects. A second set of possible
paths were estimated to test the influence of parents’ religious identity prominence
on each other’s and their children’s behavior in the religious identity. Only one of
these six paths, from fathers’ prominence to mothers’ behavior, rises to conventional
levels of statistical significance. An increase in fathers’ religious identity prominence
is positively associated with a mothers’ religious behavior (b ¼ .08, p .05).
Demographic controls. Mothers of white children report lower prominence
(b ¼ .23, p .05) and more frequent religious behavior (b ¼ .12, p .05) than
mothers of non-white children and fathers of white children report lower religious
identity prominence (b ¼ .32, p .01). White children report higher salience than
non-white children (b ¼ .09, p .05). Each additional year of a child’s age is
associated with a reduction in child’s salience (b ¼ .02, p .001) and father’s
salience (b ¼ .02, p .05).
Discussion
The first three hypotheses are drawn from the main tenets of identity theory. The first
hypothesis posits a positive relationship between religious identity prominence and
salience. For mothers, fathers, and children, findings demonstrate that the value the
respondent places on their religious identity positively influences their propensity of
enacting that identity, thus supporting the first hypothesis. The second hypothesis,
that religious identity salience is a positive predictor of behavior in the religious
identity is also supported, although the coefficient for this path is somewhat weaker
for children. The third hypothesis, that religious identity prominence is a positive
predictor of religious behavior, is only partially supported. Prominence is a positive
predictor of behavior for mothers but the association is just outside conventional
levels of statistical significance for children and fathers.
Taken together, the first three hypotheses support to the main tenets of identity
theory: the value one places on an identity influences its propensity of enactment
which in turn influences behavior in the identity. Unsurprisingly, however, this
process is relatively weaker for adolescent children, whose religious identities are
still in development. The value children place on their religious identities does
influence their stated propensity to enact that identity, but the link between that
stated propensity and actual behavior is attenuated relative to their parents. This may
be at least partially due to the regimentation of children’s behavior. Children’s
religious behavior is influenced not only by their religious identity salience but
also, and most importantly, by positive or negative constraints set by parents.
Thus, for children, the link between salience and behavior may be interrupted by
parental mandates. They may be forced into (or potentially prevented from
performing) religious behavior, such as attending services, reading scripture, reciting
prayers, or performing rituals, regardless of their own religious identity salience.
These constraints are illustrated in Hypothesis 6, to be discussed shortly. Other
research has investigated the role of constraints in the relationship between religious
identity salience and religious behavior (Brenner, 2016), but more research, espe-
cially on children’s religious identities, is needed.
The next three hypotheses focused on integrating religious socialization into the
identity model. These hypotheses connected religious identities intergenerationally
through a set of direct effects between the same identity concepts from parents to
their children. Hypothesis 4 directly connected the religious identity prominence of
mothers and fathers to that of their children. This hypothesis was supported. The
value each parent places on their religious identity directly and positively influences
the value their children place on their religious identities.
Hypothesis 5 posited that each parent’s religious identity salience is a direct and
positive influence on that of their children. This hypothesis was only partially
supported. Mothers’ salience was positively associated with children’s salience,
but the association of fathers’ salience and children’s salience was only marginally
significant. Hypothesis 6 posited that each parent’s religious behavior is directly and
positively related to that of their children. This hypothesis was also only partially
supported. Mothers’ behavior was positively associated with children’s behavior,
but the association of fathers’ behavior and children’s behavior was only marginally
significant.
The final hypothesized pathway from parents to children, hypothesis 7, posits a
direct influence from parents’ religious identity prominence to the religious identity
salience of their children. This hypothesis was only partially supported. The value
that mothers place on their religious identities positively influences their children’s
propensity of performing their religious identities. Fathers’ prominence, however,
failed to influence the religious identity salience of their children. Thus, children may
be more likely to define situations as relevant for their religious identities to appease
their mothers’, but not their fathers’, religious identities.
Taken together, these findings further reveal how parents, and particularly
mothers, contribute to their children’s religious socialization and illustrate some
important differences in the roles that mothers and fathers play in the religious
socialization of their children and in their children’s religious identity formation.
The relationship between mothers religious identities and those of their children
operates essentially as hypothesized. However, the relationship between fathers’
religious identities and those of their children diverges markedly from the proposed
model. Thus, findings suggest the primary route by which children are socialized
into their religious communities and traditions is through their mothers’ religious
identities. In sum, children value their religious identity because their parents do;
children see situations as relevant for their religious identities because their par-
ents—primarily their mothers—do and because their mothers’ value their religious
identities; and children enact their religious identities because their parents—pri-
marily their mothers—do.
194 P. S. Brenner
Each of these findings suggest ways by which parents’ religious identities

influence those of their children rather than the reverse. Yet, one route by which
children’s religious identities influence those of their parents was hypothesized.
Children’s religious identity prominence was hypothesized to have a direct effect
on parents’ religious identity salience: the more value their children place on their
religious identities, the more likely their parents are to enact their own religious
identities. This hypothesis, hypothesis 8, was supported for both mothers and
fathers. This single pathway captures the “a child shall lead them” understanding:
parents’ religious identities are supported by having and rearing children. The size of
these paths from children’s prominence to parent’s salience are approximately equal
to those from one parent’s prominence to the other parents’ salience, and are among
the weaker paths in the model, especially in comparison to the mother’s direct
influences from her religious identity prominence, salience, and behavior to those
of her children. Thus, while children do exert some influence on their parents in this
way, it is small in size relative to other estimated associations.
The final hypothesis, 9, estimates a similar path, allowing each parent’s promi-
nence to influence the salience of the other parent. These paths are positive and
significant, and support the hypothesis. All other direct connections between parents
were estimated as covariances. Each of these associations—between parents’ reli-
gious identity prominence, salience, and behavior in the religious identity—is
positive and significant suggesting some level of alignment between parents’ reli-
gious identities. Although no assumption is made about the direction of these
relationships, future research may attempt to establish causal relationships. While
assortative mating tends to shuffle individuals into relatively religiously homoga-
mous relationships (McClendon, 2016), some directional influences occur between
partners during relationships (Jennings & Stoker, 2001). Given the strong role
mothers play in influencing their children’s religious identities as established in
these findings, it seems plausible that mothers may also positively influence fathers
to a greater extent than fathers influence mothers.
Finally, twelve additional paths were estimated to test the robustness of model
specifications, each representing a possible but unhypothesized relationship. Only
one of these paths, from father’s prominence to mother’s behavior, rose to conven-
tional levels of statistical significance. Given the small size of the coefficient, and the
likelihood that this finding is simply due to chance, it will not be discussed further
here, although future research may wish to investigate the potential of this associ-
ation in greater detail. Notably, however, none of the paths connecting parents’ and
children’s salience to others’ behavior were significant. This finding supports the
model’s assumptions that behavior in the religious identity is directly influenced by
others’ behavior rather than others’ religious identity salience.
Taken together, these findings support hypotheses based in identity theory for the
role that parents play in the religious socialization of their children. Clearly, parents’
religious identities influence their children’s religious identities. Children value the
identities their parents value, and express a propensity to perform the identities that
their parents do. However, as suggested by the extant research, the influence of
mothers is stronger and more consistent than that of fathers (Acock & Bengtson,
1978; Hayes & Pittelkow, 1993; Pearce & Thornton, 2007). In short, mothers appear
to be the driving force behind their children’s religious identity with fathers playing a
relatively marginal supporting role.
These findings also demonstrate the limitations of parental influence. Neither
parents’ religious identity prominence nor salience influence their children’s reli-
gious behavior. The only significant direct influence on children’s behavior is their
mothers’ behavior. Thus, children’s religious behavior appears to be obligatory—
children pray, read scripture, perform rituals, and go to religious services because
their mothers do. However, as children age, they become more able to make
decisions on which identities are meaningful for them and how they will spend
their time. Thus, as children transition toward young adulthood and grow more
independent, the relationship between religious identity prominence, salience, and
behavior are apt to change. This is illustrated in the model. Age of the child, included
as a covariate, is significant in its negative associations with a number of identity
constructs across family members. As children age, the salience but not the prom-
inence of their religious identities is reduced. This finding is in line with previous
research on adolescents that suggests a negative relationship between autonomy and
the propensity for religious behavior but not with the value adolescents place on the
religious identity (Armet, 2009; Lopez et al., 2011). Thus, parents can still influence
their adolescent children’s religious identity prominence, even if they have less
control over their religious identity salience and religious behavior.
Some part of the negative effect of age may be linked to identity commitments;
the quality and quantity of links that an individual has to an identity. The first type,
affective commitments, influence the prominence of an identity: the more intensive
the connections to others one knows through the identity, the higher the prominence
of the identity. The second type, interactional commitments, influence the salience of
an identity: the higher the number of connections to others via the identity, the higher
the salience of the identity (Stryker & Serpe, 1994). As children age and enter social
contexts outside of the family, they gain more identities—student, friend, musician,
artist, athlete, and so on. As their number of identities grows and those connections
to others become more numerous and strong, the relative prominence and salience of
the identities they previously held, such as child, sibling, member of a religious
group, and so on may be reduced in light of competition from these other identities
(Desmond et al., 2010). As a result, the relationship between parents’ and their
children’s religious identities tends to weaken as children age and others, such as
friends, teachers, and peers, gain social influence as the relative influence of parents
and their values wanes.
Estimating a series of models separately for children of different ages (so-called
“groups” models), may help to elucidate this process allowing different slopes for
children of different ages. Unfortunately, these data are limited by a relatively small
sample size and such a model may not be robustly estimated. Moreover, measures of
religious identity commitments were not included in the model tested here given
limitations of these data. Future research should investigate the role of changing
identity commitments as children transition to young adulthood, in addition to aging
196 P. S. Brenner
and psychosocial development on the development of the religious identity and

religious socialization of children.
Conclusion
One of my mother’s favorite television programs when I was a child was the
Waltons; a show that chronicled the life of a large family in the mountains of
Virginia during the Great Depression and Second World War. The family’s matri-
arch and patriarch were both religious, the matriarch devoutly so, although the
patriarch typically stayed home while his wife and their seven children attended
weekly services at their Baptist church. Although I haven’t seen this television show
in decades, it seems, given my somewhat hazy recollections of it, a useful illustration
of the findings of this research: a mother with high religious identity prominence,
high religious identity salience, and frequent religious behavior that motivates the
prominence, salience, and behavior of her adolescent children, while their father
remains on the sidelines of the family’s religious life. Not to belabor the analogy, but
it holds right down to the religious socialization and religious identity development
of the family’s oldest son who himself saw his religious identity salience wane over
time, accompanied by a reduced tendency for religious behavior.
This paper reinterprets religious socialization as a function of identity processes
of parents and children. It views the strong influence of parental religiosity on
children, but also distinguishes between the different roles of mothers and fathers.
Mothers carry much of the weight of religious socialization and religious identity
development in their children, with some assistance from fathers, although it is likely
that fathers too are influenced by the religious identities of their partners. As parents,
especially mothers, value their religious identities and have high propensities of
performing them, children too place higher values of these identities and have high
propensities of performing them. These constructs, religious identity prominence
and salience, help us understand religious behavior.
Moreover, this research also acknowledges that identities change. Children’s
religious identity formation is not unidirectional toward increased religiosity. Rather,
an important caveat emerged in this research that may create stumbling block for the
successful religious socialization of offspring: children get older. These findings
suggest that as children age, the influence of parents’ religiosity may wane. It may be
that other identity commitments that take a lead role in the movement of children
away from their religious identities. The increasing connections of children with
others outside the family provides opportunities for other identities to increase in
prominence and salience and, potentially, overtake or even dislodge the religious
identity from its high place in the prominence and salience hierarchies. This rela-
tionship between children’s increasing quantity and quality of other identities and
reductions in religious identity prominence and salience should be further explored
in future research.
Future research should also address the additional limitations of this work,
notably its inability to fully model the relationship between mothers and fathers
and model change over time. A large panel data set, clustered by family unit, would
be ideal for such an endeavor.
References
Acock, A. C., & Bengtson, V. L. (1978). On the relative influence of mothers and fathers: A
covariance analysis of political and religious socialization. Journal of Marriage and Family, 40
(3), 519–530.
Appiah, K. A. (2018). The lies that bind: Rethinking identity. Liveright.
Armet, S. (2009). Religious socialization and identity formation of adolescents in high tension
religions. Review of Religious Research, 50(3), 277–297.
Asencio, E. K. (2013). Self-esteem, reflected appraisals, and self-views: Examining criminal and
worker identities. Social Psychology Quarterly, 76(4), 291–313.
Asencio, E. K., & Burke, P. J. (2011). Does incarceration change the criminal identity? A synthesis
of the labeling and identity theory perspectives on identity change. Sociological Perspectives,
54(2), 163–182.
Bainbridge, W. S. (1982). Shaker demographics 1840-1900: An example of the use of U.S. census
enumeration schedules. Journal for the Scientific Study of Religion, 21(4), 352–365.
Baker-Sperry, L. (2001). Passing on the faith: The father’s role in religious transmission. Socio-
logical Focus, 34(2), 185–198.
Bengtson, V. L., Copen, C. E., Putney, N. M., & Silverstein, M. (2009). A longitudinal study of the
intergenerational transmission of religion. International Sociology, 24(3), 325–345.
Bengtson, V. L., Putney, N. M., & Harris, S. C. (2013). Families and faith: How religion is passed
down across generations. Oxford University Press.
Boyatzis, C. J., & Janicki, D. L. (2003). Parent-child communication about religion: Survey and
diary data on unilateral transmission and bi-directional reciprocity styles. Review of Religious
Research, 44(3), 252–270.
Brenner, P. S. (2011a). Investigating the biasing effect of identity importance in self-reports of
socially desirable behavior. Sociological Focus, 44(1), 55–75.
Brenner, P. S. (2011b). Identity importance and the overreporting of religious service attendance:
Multiple imputation of religious attendance using American time use study and the general
social survey. Journal for the Scientific Study of Religion, 50(1), 103–115.
Brenner, P. S. (2016). Time as a situational constraint on role-identity performance. In J. E. Stets &
University Press.
Brenner, P. S., & DeLamater, J. (2014). Social desirability bias in self-reports of physical activity: Is
an exercise identity the culprit? Social Indicators Research, 117(2), 489–504.
Brenner, P. S., & DeLamater, J. (2016). Lies, damned lies, and survey self-reports? Identity as a
cause of measurement bias. Social Psychology Quarterly, 79(4), 333–354.
Brewer, P. J. (1984). The demographic features of the Shaker decline, 1787-1900. Journal of
Interdisciplinary History, xv(i), 31–52.
Burke, P. J. (1991). Identity processes and social stress. American Sociological Review, 56(6),
836–849.
Cast, A. E., & Burke, P. J. (2002). A theory of self-esteem. Social Forces, 80(3), 1041–1068.
198 P. S. Brenner
Converse, J. M., & Presser, S. (1986). Survey questions: Handcrafting the standardized question-
naire. Sage.
DellaPergola, S. (2005). Was it the demography? A reassessment of US Jewish population
estimates, 1945-2001. Contemporary Jewry, 25(1), 85–131.
Demo, D. H., & Hughes, M. (1990). Socialization and racial identity among Black Americans.
Desmond, S. A., Morgan, K. H., & Kikuchi, G. (2010). Religious development: How (and why)
does religiosity change from adolescence to young adulthood? Sociological Perspectives, 53(2),
247–270.
Dollahite, D. C., & Thatcher, J. Y. (2008). Talking about religion: How highly religious youth and
parents discuss their faith. Journal of Adolescent Research, 23(5), 611–641.
Erickson, J. A. (1992). Adolescent religious development and commitment: A structural equation
model of the role of family, peer group, and educational influences. Journal for the Scientific
Study of Religion, 31(2), 131–152.
Fletcher, A. C., Elder, G. H., Jr., & Mekos, D. (2000). Parental influences on adolescent involve-
ment in community activities. Journal of Research on Adolescence, 10(1), 29–48.
Flor, D. L., & Knapp, N. F. (2001). Transmission and transaction: Predicting adolscents’ internal-
ization of parent religious values.
Francis, L. J. (1993). Parental influence and adolescent religiosity: A study of church attendance and
attitude toward christianity among adolescents 11 to 12 and 15 to 16 years old. International
Journal for the Psychology of Religion, 3(4), 241–253.
Fukuyama, F. (2018). Identity: The demand for dignity and the politics of resentment. Farrar,
Straus, and Giroux.
Gallagher, S. K. (2007). Children as religious resources: The role of children in the social
re-formation of class, culture, and religious identity. Journal for the Scientific Study of Religion,
46(2), 169–183.
Glass, J., Bengtson, V. L., & Dunham, C. C. (1986). Attitude similarity in three-generation families:
Socialization, status inheritance, or reciprocal influence? American Sociological Review, 51(5),
685–698.
Grusec, J. E. (2011). Socialization processes in the family: Social and emotional development.
Annual Review of Psychology, 62, 243–269.
Gunnoe, M. L., & Moore, K. A. (2002). Predictors of religiosity among youth aged 17–22: A
longitudinal study of the national survey of children. Journal for the Scientific Study of Religion,
41(4), 613–622.
Hadaway, C. K., Marler, P. L., & Chaves, M. (1993). What the polls don’t show: A closer look at
US church attendance. American Sociological Review, 58(6), 741–752.
Hadaway, C. K., Marler, P. L., & Chaves, M. (1998). Overreporting church attendance in America:
Evidence that demands the same verdict. American Sociological Review, 63(1), 122–130.
Hayes, B. C., & Pittelkow, Y. (1993). Religious belief, transmission, and the family: An Australian
study. Journal of Marriage and Family, 55(3), 755–766.
Hoge, D. R., Petrillo, G. H., & Smith, E. I. (1982). Transmission of religious and social values from
parents to teenage children. Journal of Marriage and Family, 44(3), 569–580.
Hogg, M. A., & Abrams, D. (1988). Social identifications: A psychology of intergroup relationsand
group processes. Routledge.
Hogg, M. A., Terry, D. J., & White, K. M. (1995). A tale of two theories: A critical comparison of
identity theory with social identity theory. Social Psychology Quarterly, 58, 255–269.
Hout, M., Greeley, A., & Wilde, M. J. (2001). The demographic imperative in religious change in
the United States. American Journal of Sociology, 107(2), 468–500.
Jennings, M. K., Stoker, L., & Bowers, J. (2009). Politics across generations: Family transmission
reexamined. Journal of Politics, 71(3), 782–799.
Jennings, M. K., & Stoker, L. (2001). Political similarity and influence between husbands and
wives. UC Berkeley Working Paper 2001-14. https://escholarship.org/uc/item/9s54f2mc
Kieren, D. K., & Munro, B. (1987). Following the leaders: Parents’ influence on adolescent
religious activity. Journal for the Scientific Study of Religion, 26(2), 249–255.
Lerer, N., & Mayer, E. (1989). In the footsteps of ruth: A sociological analysis of converts to
Judaism in America. Papers in Jewish Demography, 172–184.
Liefbroer, A. C., & Elzinga, C. H. (2012). Intergenerational transmission of behavioral patterns:
How similar are parents’ and children’s demographic trajectories? Advances in Life Course
Research, 17(1), 1–10.
Lopez, A. B., Huynh, V. W., & Fuligni, A. J. (2011). A longitudinal study of religious identity and
participation during adolescence. Child Development, 82(4), 1297–1309.
Marks, L. (2004). Sacred practices in highly religious families: Christian, Jewish, Mormon, and
Muslim perspectives. Family Process, 43(2), 217–231.
Martin, T. F., White, J. M., & Perlman, D. (2003). Religious socialization: A test of the channeling
hypothesis of parental influence on adolescent faith maturity. Journal of Adolescent Research,
18(2), 169–187.
McCall, G. J., & Simmons, J. L. (1978). Identities and interactions: An examination of human
associations in everyday life. Free Press.
McClendon, D. (2016). Religion, marriage markets, and assortative matings in the United States.
Journal of Marriage and Family, 78(5), 1399–1421.
McQuillan, K. (2004). When does religion influence fertility? Population and Development Review,
30(1), 25–56.
Min, J., Silverstein, M., & Lendon, J. P. (2012). Intergenerational transmission of values over the
family life course. Advances in Life Course Research, 17(3), 112–120.
Myers, S. M. (1996). An interactive model of religiosity inheritance: The importance of family
context. American Sociological Review, 61(5), 858–866.
Pearce, L. D., & Thornton, A. (2007). Religious identity and family ideologies in the transition to
adulthood. Journal of Marriage and Family, 69(5), 1227–1243.
Perry, S. L., & Schleifer, C. (2019). Are the faithful becoming less fruitful? The decline of
conservative protestant fertility and the growing importance of religious practice and belief in
childbearing in the US. Social Science Research, 78, 137–155.
Ploch, D. R., & Hastings, D. W. (1998). Effects of parental church attendance, current family status,
and religious salience on church attendance. Review of Religious Research, 39(4), 309–320.
Presser, S., & Stinson, L. (1998). Data collection mode and social desirability bias in self-reported
religious attendance. American Sociological Review, 63(1), 137–145.
Regnerus, M. D., Smith, C., & Smith, B. (2004). Social context in the development of adolescent
religiosity. Applied Developmental Science, 8(1), 27–38.
Schneider, B., & Waite, L. J. (2008). The 500 Family Study [1998-2000: United States]. In
ICPSR04549-v1. Inter-university Consortium for Political and Social Research.
Sherkat, D. E. (1998). Counterculture or continuity? Competing influences on baby boomers’
religious orientations and participation. Social Forces, 76(3), 1087–1114.
Sherkat, D. E. (2003). Religious socialization: Sources of influence and influences of agency. In
M. Dillon (Ed.), Handbook of the sociology of religion (pp. 151–163). Cambridge University
Press.
Silverstein, M., & Bengtson, V. (2018). Linked religious lives across generational time in family
lineages: Grandparents as agents of transmission. In D. F. Alwin, D. H. Felmlee, & D. A.
Kreager (Eds.), Social networks and the life course: Integrating the development of human lives
and social relational networks (pp. 415–429). Springer.
Smith-Lovin, L. (2003). Self, identity, and interaction in an ecology of identities. In P. J. Burke,
(pp. 167–178). Klewer/Plenum.
Snell Herzog, P., & Mitchell, S. (2016). Intergenerational transmission of religious giving: Instilling
giving habits across the life course. Religions, 7, 93. https://doi.org/10.3390/rel7070093
Quarterly, 63(3), 224–237.
200 P. S. Brenner
Quarterly, 74, 192–215.
Sociological Review, 77, 120–140.
Stets, J. E., & Cast, A. E. (2007). Resources and identity verification from an identity theory
perspective. Sociological Perspectives, 50(4), 517–543.
Stolzenberg, R. M., Blair-Loy, M., & Waite, L. J. (1995). Religious participation in early adulthood:
Age and family life cycle effects on church membership. American Sociological Review, 60(1),
84–103.
Storm, I., & Voas, D. (2012). The intergenerational transmission of religious service attendance.
Nordic Journal of Religion and Society, 25(2), 131–150.
Stryker, S. (1968). Identity salience and role performance: The relevance of symbolic interaction
theory for family research. Journal of Marriage and the Family, 30, 558–564.
Stryker, S. (1980/2003). Symbolic interactionism: A social structural version. Blackburn.
Stryker, S., & Serpe, R. T. (1982). Commitment, identity salience, and role behavior: Theory and
research example. In W. Ickes & E. S. Knowles (Eds.), Personality, roles, and social behavior
(pp. 197–218). Spring-Verlag.
Stryker, S., & Serpe, R. T. (1983). Toward a theory of family influence in the socialization of
children. Research in Sociology of Education and Socialization, 4, 47–71.
overlapping, or complimentary concepts? Social Psychology Quarterly, 57, 16–35.
Vedder, P., Berry, J., Sabatier, C., & Sam, D. (2009). The intergenerational transmission of values
in national and immigrant families: The role of zeitgeist. Journal of Youth and Adolescence, 38,
642–653.
Wilcox, W. B. (2002). For the sake of the children: Family-related discourse and practice in the
mainline. In R. Wuthnow & J. H. Evans (Eds.), The quiet hand of god: Faith-based activism and
the public role of mainline protestantism (pp. 287–316). University of California Press.
Chapter 10
Identity, Attribution, and Self-Esteem:
When Employment Status Matters
Michael M. Harrod and Richard T. Serpe
Introduction
Identity theory is grounded in structural symbolic interaction (Stryker, 1968, 1980;

Stryker & Burke, 2000). It takes as one of its starting points the impact of social
structure facilitating or constraining the opportunity to engage in social action.
Identity theory posits that internalized identity meanings are based on interaction
that includes meanings located in society, others, and the self. Most identity theory
research has focused on active research programs developing research questions
around how identities are impacted through structural and perceptual processes
(Burke & Stets, 2009; Serpe & Stryker, 2011; Stets & Serpe, 2013). These structural
and perceptional processes emphasize the significance of internalizing identity
meanings, and these meanings are an essential aspect of defining the self. Identity
theory research focuses on how social factors are embedded in large, intermediate,
and proximal social structures. Placement in the social structure shapes the meanings
of one’s identities and the opportunities to perform one’s identities (Owens et al.,
2010; Serpe & Stryker, 2011).
We build on these connections between the self and the social structure by
investigating how actors’ self-attribution of responsibility for their current employ-
ment status affects the worker identity’s prominence and, ultimately, how positively
or negatively the self is viewed. The worker identity was chosen because it is
commonly found within the general population. Moreover, work holds a central
place in people’s lives, allowing them to provide for themselves and their families; it
M. M. Harrod (*)
Department of Sociology, Central Washington University, Ellensburg, WA, USA
R. T. Serpe
Department of Sociology, Kent State University, Kent, OH, USA
202 M. M. Harrod and R. T. Serpe
locates the person within a role in society; it also provides a structure for daily
activity and expectations for the future. We focus on two groups, those who are
working and those who are unemployed but actively seeking employment. Self-
attributions about the cause of one’s employment status helps us to understand how
one explains why they are or are not employed. Self-attributions are classified as
internal (ability and effort) or external (task difficulty and luck) (Kelley, 1967;
Weiner, 1986). Self-attributions are not static and are based on the situation and
context of the event. A person may attribute the cause of the same event or outcome
differently based on the context of the social action. Importantly, one’s placement in
the social structure either facilitates or constrains the range of information and
experience that helps the person make the self-attribution to cause.
Research on social structure and self-attributions has demonstrated that gender,
race, education, occupation, and income have differential effects on self-attribution
(Howard, 2000; Howard & Pike, 1986; Shepelak & Alwin, 1986; Stryker &
Gottlieb, 1981). Placement within the social structure both enables and restricts
social networking possibilities (Stryker et al., 2005). Within these social structures,
each of our lives comes to shape how we understand the world. From these social
positions, we internalize meanings and shared expectations of how the world
operates, including how we fit into society. These internalized meanings exist within
identities. When a person makes an attribution of successful performance (e.g.,
passing an exam, winning a competition, securing employment, etc.) to internal
causes, or when they make an attribution of unsuccessful performance (e.g., failing
an exam, losing a competition, failing to find employment, etc.) to external causes,
these attributions influence the self and the internalized meanings of their identity as
a student, a competitor, worker identity, etc.
Identity theory research finds that the more individuals receive verification of an
identity, the more they will experience positive outcomes such as increased efficacy
and positive emotions (Stets & Burke, 2014). Self-attributions for responsibility of
successful or unsuccessful outcomes are likely to influence identity processes such
as prominence, salience (Markowski & Serpe, 2018; Owens & Serpe, 2003; Stets &
Serpe, 2013), and verification by providing self-relevant feedback on the identity
(Burke & Harrod, 2005; Stets & Burke, 2005). The prominence of an identity,
whether high or low, impacts the self and one’s level of self-esteem (Brenner
et al., 2018; Markowski & Serpe, 2018).
We test whether self-attributions of the cause of employment status, as a source of
feedback about the self, influence worker identity prominence.1 The prominence of
the worker identity is based on how central or important the worker identity is to
one’s self-concept (Brenner et al., 2014, 2018; Rosenberg, 1979). Self-attributions
are the perceived explanation or the cause of one’s accomplishments as well as one’s
failures (Howard, 2000; Howard & Pike, 1986; Shepelak & Alwin, 1986; Stryker &
1
Self-attributions can impact identity salience as well. This research focuses on self-esteem. If the
dependent variable were a behavior e.g., time-in-role, number of activities associated with seeking
employment, identity salience would be included.
10 Identity, Attribution, and Self-Esteem: When Employment Status Matters 203
Gottlieb, 1981). A person’s evaluation of the cause for successful/unsuccessful

performance or outcomes, in the case of being employed or unemployed, is an
appraisal and impacts the prominence of an identity.
Research on the relationship between self-attribution and self-esteem shows that
positive attributions are related to higher self-esteem, and negative attributions are
related to lower self-esteem (Rosenberg et al., 1995; Wallace & Baumeister, 2002)
and that self-attributions are part of the development of a self-concept (Rosenberg,
1979). The relationship of self-attributions to successful or unsuccessful employ-
ment, occupational status, work on self-esteem also demonstrates that success
associated with work leads to higher levels of self-esteem (de Araujo & Lagos,
2013; Gecas & Seff, 1989; Wang et al., 2010).
This research investigates the processes of self-attributions, identity prominence,
and self-esteem. We posit that self-attributions regarding the causes of employment
impact both identity prominence and self-esteem. Self-attributions of the cause of
being employed or unemployed are theorized to impact the location of the worker
identity in the prominence hierarchy and the person's level of self-esteem. We also
posit that worker identity prominence will impact one’s self-esteem.
Attribution and Identity Theories
Identity theory (Stryker, 1980) posits that identities are arranged into hierarchies
representing which identities are more important or more likely to be enacted. One of
these is the prominence hierarchy. The prominence hierarchy envisions identities as
a rank-ordering of each identity’s importance or centrality to the person’s self-
concept (McCall & Simmons, 1978; Rosenberg, 1979). The more important or
central an identity is the higher that identity’s location in the prominence hierarchy.
Among the factors influencing an identity’s location in the prominence hierarchy are
the social relationships connected to identity, feedback from others regarding the
identity in form of reflected appraisals, and the person’s self-view of the identity.
Within this research, we posit that self-attributions to the cause of outcomes asso-
ciated with the identity (e.g., exam performance, competitive performance, obtaining
employment, etc.) are sources of information informing a person’s self-view.
In the context of one’s placement in the social structure, persons make self-
attributions that are self-evaluations of outcomes associated with an identity (How-
ard, 2000; Howard & Pike, 1986; Shepelak & Alwin, 1986). Attributions have been
theoretically conceptualized as part of the self-concept (Rosenberg, 1979, 1988;
Stets & Burke, 2005; Stryker & Gottlieb, 1981). Self-attributions are essential
because they allow actors to interpret and understand self-relevant events (Howard
& Pike, 1986). Self-attributions are a uniquely personal source of feedback regarding
how the outcome of identity enactment is evaluated.
Heider (1958) posits that explanations of behavior are either external attributions
or internal attributions. External attributions are those that are blamed on situational
forces, while internal attributions are blamed on individual characteristics and traits.
Central to Heider’s formulation of elements that are used in making attributions is

the understanding of how the social world and a person’s placement within it
influences individuals’ interpretation of the outcomes of social action. Heider con-
ceived of people as striving for understanding and control over their environment.
From this perspective, one is motivated to understand how the world operates by
determining an outcome’s root cause. Heider conceived of outcomes as being either
attributable to the actor or an external source. If the person perceives they are
responsible for an outcome, they make an internal attribution (e.g., ability or effort).
Alternatively, when a person perceives that something outside their control is the
cause (e.g., task difficulty or luck), they make an external attribution.
Attributions reflect the person’s perception of the cause of an outcome, consid-
ering the context (self and others) of the interaction and the structural environment
(location and characteristics) and whether the context and the location for the person
is more open or closed behavior (Kelley, 1967). Rather than drawing conclusions
about others, self-attribution focuses on how people draw “conclusions about their
own motives or underlying characteristics when interpreting the causality of events”
(Rosenberg, 1979, p. 71, italics in the original). When our actions and the outcomes
are predictable (Crittenden, 1983; Crittenden & Wiley, 1980), actors come to
associate their actions with an outcome. For example, someone who successfully
makes multiple home repairs will begin to see themselves as “handy.” However, the
same person’s repeated failure to cook a good meal results in the self-attribution of
being a lousy cook. People can associate their efforts to achieve the desired outcome
with their underlying characteristics and skills. Here, self-attributions represent the
actor’s understanding of the cause of events and whether they were responsible for
the outcomes experienced.
Two Components of Self-Esteem
Self-esteem represents the positive or negative attitude one has about the self
(Rosenberg, 1979). Although social psychologists have been researching self-
esteem for decades, there remains some disagreement about how to best conceptu-
alize it. Among the conceptualizations offered, one relevant for the present argument
is the actor’s motivation around self-esteem: self-enhancement or self-consistency.
We do not claim to be directly contributing to this literature. Instead, we engage this
literature to contextualize our approach to the measurement of self-esteem.
A leading measure of self-esteem is the Rosenberg Self-Esteem (RSE) scale,
initially envisioned as a unidimensional global self-esteem scale. The ubiquity of the
RSE in research endeavors led some to question its unidimensionality. Huang and
Dong’s (2012) meta-analysis looked at 23 studies that consisted of 80 independent
samples. Their meta-analysis analyzed over 32,000 respondents leading them to
recommend that researchers treat the RSE as a unidimensional construct unless there
is a compelling reason to use the positive and negative subscales. Since we are
looking at two identity statuses, one normative and one counter-normative that have
been shown to have differential effects (Adams & Serpe, 2020; Markowski & Serpe,
2018), for both theoretical and empirical grounds, we separate the RSE into self-
confidence and self-deprecation subscales for this analysis (Huang & Dong, 2012;
Owens, 1993, 1994).
Theoretically, we posit that using RSE’s positive and negative subscales is
justified to capture the contextual differences in self-evaluation between persons
employed full-time and unemployed persons who are actively seeking employment.
Importantly, we characterize those who are employed as holding a normative worker
identity and those who are unemployed (excluding those who are students, retired,
etc.) holding the counter-normative unemployed worker identity. Specifically, we
posit that those who are unemployed and wish to be employed will view the
unemployed status negatively. We posit that the unemployed will agree with the
self-denigrating items, e.g., “All in all, I am inclined to feel that I am a failure.” The
self-denigrating items are reverse-coded; therefore, the unemployed will likely have
a lower self-esteem score for the self-denigrating sub-scale. An examination of the
positive and negative subscales of the RSE aligns with self-evaluations associated
with identity theory when people hold a normative and counter-normative identity
and is consistent with the internal and external dimensions of attribution theory as
influencing people’s sense of self. As discussed above, both theories posit that
people seek consistency, controllability, and predictability to their social worlds.
An analysis of both dimensions of the RSE enables us to test for the two competing
self-motivations of self-relevant feedback found in identity theory: self-enhancement
and self-consistency (Burke & Harrod, 2005; Stets, 2003; Stets & Burke, 2014).
Similarly, by examining both dimensions, we follow Owens’ (1993, 1994) theoret-
ical justification and subsequent empirical illustration of the RSE’s two subscales.
In discussing the positive dimension, Owens (1993) points out that much of the
work on self-esteem assume actors are motivated to see themselves in the best
positive light—a self-enhancement motive. Research shows that the self-enhance-
ment motive leads people to interpret events and their self-image—identity—as
positively as possible (Markus & Wurf, 1987). In service to the self-enhancement
motive, people may selectively process and remember events positively (Markus &
Wurf, 1987).
To explain the negative aspects of self-esteem, Owens draws on both attribution
and Swann’s self-verification theories. The link to attribution theory is rooted in that
actors with negative self-views also seek consistent feedback from others as to who
they are (Kelley, 1971). Likewise, drawing on Swann’s self-verification theory, the
focus is on actors receiving feedback consistent with their existing self-views.
Swann’s self-verification approach has a clear affinity to identity theory and empha-
sizes receiving feedback consistent with existing self-views.
Emerging research shows that normative and counter-normative groups experi-
ence the world differently and supports examining the two subdimensions of self-
esteem. Normative identities are consistent with societal expectations. Counter-
normative identities, especially those not held by choice, are inconsistent with
societal expectations (Adams & Serpe, 2020; Long, 2016; Markowski & Serpe,
2018; Stets & Serpe, 2013). Occupying normative and counter-normative identities
Female
Self-
Attributional Confidence
Trait
Personal Prominence
Income
Educational Self-
Attainment Deprecation
Age
Fig. 10.1 Heuristic model
locate the person differentially within the social structure and societal expectations.
Some examples of normative identities are worker, parent, and religious person.
Counter-normative identities include the unemployed, voluntarily childless, and
non-religious. Those holding normative and counter-normative identities are likely
to experience different levels of support for their identities and thus experience
different levels of prominence of their identity than those holding normative iden-
tities (Adams & Serpe, 2020; Long, 2016; Markowski & Serpe, 2018).
Based on our conceptualization of the relationships between self-attributions,
identity prominence, and self-esteem, we estimate the following model. The model
specifies that self-attributions impact identity prominence, and both dimension of
self-esteem, self-confidence and self-deprecation and identity prominence impacts
both dimensions of self-esteem. The model also estimates the background variables
of gender, income, education, and age. The background variables are estimated to
account for placement in the social structure (Fig. 10.1).
The model assumes an ordering of the self-attributions, identity prominence, and
self-esteem. Following the theory of self-attributions, we hypothesize that internal
and external attributions will have differential effects on identity prominence. We
formalize these expectations with the following hypotheses.
H1a: An internal attribution to ability will positively impact identity prominence
for the employed.
H1b: An internal attribution to ability will negatively impact identity prominence
for the unemployed.
H1c: An internal attribution to effort will positively impact identity prominence
for the employed.
H1d: An internal attribution to effort will negatively impact identity prominence
for the unemployed.
H2a: An external attribution to task difficulty will negatively impact identity
prominence for the employed.
H2b: An external attribution to task difficulty will negatively impact identity

prominence for the unemployed.
H2c: An external attribution to luck will negatively impact identity prominence for
the employed.
H2d: An external attribution to luck will negatively impact identity prominence
for the unemployed.
We focus on understanding situational understandings from the actor’s vantage
point, as called for by Stets and Burke (2014). Doing this allows for determining
how an actor’s attribution of responsibility for events affects the self. We focus on
two groups, those who are working and those who are unemployed but actively
seeking employment. Working and not working identities provide the opportunity to
investigate how the attribution made for the respondent’s work status influences their
self, in this case, self-esteem (Adams & Serpe, 2020). Evaluations of self-esteem
within the context of different identities should reflect how a person's structural
location affects the comparisons they make about themselves relative to others
(Rosenberg, 1979). Research has demonstrated that identity prominence affects
self-esteem (Adams & Serpe, 2020; Brenner et al., 2018; Markowski & Serpe,
2018). Using normative and counter-normative identities allows us to test if the
impact of identity prominence on self-confidence and self-deprecation. Specifically,
we expect that the relationships between the employed and unemployed will differ
on self-confidence and self-deprecation. We hypothesize that the differences by
employment status and self-esteem dimensions will reflect the locations of the
worker identity prominence. We formalize these expectations as follows:
H3a: Identity prominence will positively impact self-confidence among the
employed.
H3b: Identity prominence will negatively impact self-confidence among the
unemployed.
H4a: Identity prominence will negatively impact self-deprecation among the
employed.
H4b: Identity prominence will positively impact self-deprecation among the
unemployed.
When people attribute success or failure to internal causes, they may experience
specific emotions (e.g., hope/hopelessness, pride, or guilt) or higher or lower
evaluations of self-esteem. Self-attributions should impact how one feels, how
their actions are responsible for events, and how one feels about themselves.
Because of this, we anticipate differences based on the internal/external dimensions.
Because internal attributions implicate self-behavior as responsible for one’s work
status, we anticipate internal attributions will increase reports of both dimensions of
self-esteem. Alternatively, external attributions of responsibility decrease reports of
self-esteem. We do not hypothesize different patterns for the employed and unem-
ployed. We have no theoretical reason for believing the underlying psychological
process differs based on holding a normative or counter-normative identity. We

formalize these expectations below.
H5a: An attribution to ability will positively impact self-confidence.
H5b: An attribution to ability will positively impact self-deprecation.
H5c: An attribution to effort will positively impact self-confidence.
H5d: An attribution to effort will positively impact self-deprecation.
H6a: An attribution to task difficulty will negatively impact self-confidence.
H6b: An attribution to task difficulty will negatively impact self-deprecation.
H6c: An attribution to luck will negatively impact self-confidence.
H6d: An attribution to luck will negatively impact self-deprecation
Methods
Data were collected using a web-based survey during the summer of 2014, which
generates a national, web-based panel of U.S. adults. Participants were recruited to
the panel using traditional random-digit-dialing techniques and methods to target
households, including cellphone-only households. These joint strategies generate a
nationally representative sample that mirrors standard RDD telephone samples
(Braunsberger et al., 2007; Yeager et al. 2011). Participants who agree to participate
in panels are then contacted via email when a survey is available. A total of 1955
interviews were completed.
Variables and Measures
Dependent Variables
Self-Confidence contains the five positively worded items of the Rosenberg Self-
Esteem Scale (RSE). The questions are “I feel that I am a person of worth, at least on
an equal basis with others”; “I feel that I have a number of good qualities”; “I am able
to do things as well as most other people”’; “I take a positive attitude towards
myself”; and “On the whole, I am satisfied with myself.” Responses were anchored
from 1 “Strongly Agree” to 4 “Strongly Disagree.” Items were then coded such that a
higher score reflects higher self-esteem among the self-confidence items. These were
then summed and averaged into the self-confidence scale. Alpha reliability scores
were high for both employed (α ¼ .89) and unemployed (α ¼ .88) respondents.
Self-Deprecation contains the five negatively worded items of the RSE. The
questions are “All in all, I am inclined to feel that I am a failure”; “I feel that I do
not have much to be proud of”; “I wish I could have more respect for myself”; “I
certainly feel useless at times”; and “At times I think that I am no good at all.”
Responses were anchored from 1 “Strongly Agree” to 4 “Strongly Disagree.” Items
were coded such that a higher score reflects higher self-esteem associated with the
self-deprecation items. These were then summed and averaged into the self-depre-
cation scale. The scale has high inter-item reliability for both employed (α ¼ .83) and
unemployed (α ¼ .86) respondents. The factor analysis for the two subscales is
reported in Appendix A.
Identity Prominence
Identity prominence is assessed by four items asking respondents to indicate how

much they agreed with how important their employment status was to how they see
themselves. The four items included the following questions: “Being [employed/
unemployed] is an important reflection of who I am”; “I have come to think of
myself as [employed/unemployed]”; “I have a strong sense of belonging as a
[employed/unemployed person]; and “Being [employed/unemployed] is an impor-
tant part of my self-image.” Responses were anchored from 1 “Strongly Agree” to
4 “Strongly Disagree.” Items were coded such that a higher value represents greater
identity prominence. These were summed and averaged into the identity prominence
measure. The four items form a single factor with an eigenvalue of 2.79. The factor
loadings of the identity prominence items can be found in Appendix B. The identity
prominence measure has high inter-item reliability for both employed (α ¼ .82) and
unemployed (α ¼ .78) respondents.
Attributional Traits
Four constructs measure attributions: ability, effort, task difficulty, and luck
(cf. Weiner, 1986). Ability is measured by summing the responses to the following
two questions: “Please indicate how well you believe each word reflects you as a
[full-time worker or unemployed worker]. Respondents were then shown the two
adjectives “competent” and “skilled.” The correlations between these two variables
are (r ¼ .56 and r ¼ .66) for the unemployed and employed, respectively. Responses
were anchored from 0 “Not at All Like Me” to 10 “Just Like Me.” Items were coded
such that a higher value represents greater ability.
Effort is measure by summing the responses to the following two questions:
“Please indicate how well you believe each word reflects you as a [full-time worker
or unemployed worker]. Respondents were then shown the two adjectives “hard-
working” and “driven.” The correlations between these two variables are (r ¼ .67
and r ¼ .79) for the unemployed and employed, respectively. Responses were
anchored from 0 “Not at All Like Me” to 10 “Just Like Me.” Items were coded such
that a higher value represents greater effort.
Task Difficulty is measure by summing the responses to the following two
questions: “Please indicate how well you believe each word reflects you as a [full-
time worker or unemployed worker]. Respondents were then shown the two adjec-
tives “complicated” and “difficult.” The correlations between these two variables are
(r ¼ .67 and r ¼ .76) for the unemployed and employed, respectively. Responses
were anchored from 0 “Not at All Like Me” to 10 “Just Like Me.” Items were coded
such that a higher value represents greater task difficulty.
Luck is measure by summing the responses to the following two questions:
“Please indicate how well you believe each word reflects you as a [full-time worker
or unemployed worker]. Respondents were then shown the two adjectives “lucky”
and “blessed.” The correlations between these two variables are (r ¼ .51 and r ¼ .53)
for the unemployed and employed, respectively. Responses were anchored from
0 “Not at All Like Me” to 10 “Just Like Me.” Items were coded such that a higher
value represents greater luck.
Background Variables
Work status is captured by respondents’ responses to the following question: “Which

of the following best describes your work status last week?” We restrict our analyses
to only those respondents indicating they were working full-time or those unem-
ployed and looking for work. We do not include respondents who are students,
keeping house, retired, part-time, and self-employed who are not seeking employ-
ment. Respondents reporting they were working full time in the prior week to the
study are coded as 1, and the unemployed respondents are coded as 0.
Sex is the respondent’s self-reported sex, female (1) or male (0). Education
contains the highest level of education the respondent obtained. Response categories
include the following: 1 “Less than high school”; 2 “High school graduate”; 3 “Some
college or technical school”; 4 “College graduate”; and 5 “Graduate or professional
degree.” Age reflects the respondent’s self-reported age in years.
Personal income refers to the income category reported by each respondent.
Income categories are: 1 “less than $9999”; 2 “between $10,000 and $14,999”;
3 “between $15,000 and $19,999”; 4 “between $20,000 and $24,999”; 5 “between
$25,000 and $29,999”; 6 “between $30,000 and $34,999”; 7 “between $35,000 and
$39,999”; 8 “between $40,000 and $44,999”; 9 “between $45,000 and $49,999”;
10 “between $50,00 and $59,999”; 11 “between $60,000 and $74,999”; and
12 “above $75,000” (Table 10.1).
Table 10.1 Means, standard deviations, and significant differences between the unemployed and
employed
Unemployed Employed
Variables N Mean SD N Mean SD
Female 160 0.56* 0.50 643 0.46* 0.50
Personal income 145 2.75* 2.70 622 8.39* 3.14
Educational attainment 160 2.98* 0.92 637 3.71* 0.93
Age 160 39.95* 10.96 644 37.77* 9.90
Prominence 156 2.26* 0.80 643 3.28* 0.59
Ability 159 14.33* 4.32 643 16.27* 3.11
Effort 160 14.14* 4.83 644 16.08* 3.19
Task difficulty 159 5.11 4.76 644 5.66 4.58
Luck 160 9.49* 5.85 644 14.24* 4.02
Self-confidence self-esteem 158 3.13* 0.71 641 2.85* 0.87
Self-deprecation self-esteem 157 2.60* 0.87 641 2.85* 0.87
*Different at p .05
Results
Correlations
The correlations are presented in Table 10.2. We begin by highlighting the bivariate
relationships of interest. Among the employed, we see that three of the four
attributional items are positively correlated with identity prominence (ability
r ¼ .23; effort r ¼ .27; and luck r ¼ .37). Conversely, among the unemployed,
there are no significant correlations between the attributional items and identity
prominence.
Turning next to the correlations between identity prominence and the self-esteem
dimensions. We see that among the employed, identity prominence is significantly
correlated with the self-confidence dimension (r ¼ .35), but not self-deprecation.
Among the unemployed, the pattern is the opposite. We see a significant correlation
between identity prominence and self-deprecation (r ¼ .34), but no significant
relationship with self-confidence.
Lastly, we see some interesting patterns among the attributional items and the two
dimensions of self-esteem. Nearly all the attributional items are significantly corre-
lated with self-esteem. Self-confidence is correlated with ability (r ¼ .49 and
r ¼ .50), effort (r ¼ .41 and r ¼ .49), task difficulty (r ¼ .14 and r ¼ .24),
and luck (r ¼ .34 and r ¼ .51) for the employed and unemployed, respectively. Self-
deprecation is correlated with ability (r ¼ .26 and r ¼ .25), effort (r ¼ .19 and
r ¼ .22), task difficulty (r ¼ .57 and r ¼ .50), and luck (r ¼ .01ns and r ¼ .28)
for the employed and unemployed, respectively.
212
Table 10.2 Zero-order correlations among variables, employed group above the diagonal and unemployed group below the diagonal
Variables 1 2 3 4 5 6 7 8 9 10 11
1. Female 1.00 .19* .02 .15* .02 .04 .15* .04 .11* .04 .02
2. Personal income .06 1.00 .45* .14* .05 .11* .03 .09* .12* .13* .06
3. Educational attainment .24* .14 1.00 .08* .04 .08* .03 .11* .05 .09* .02
4. Age .14 .09 .08 1.00 .04 .16* .03 .20* .03 .06 .26*
5. Prominence .05 .11 .12 .12 1.00 .23* .27* .01 .37* .35* .01
6. Ability .07 .00 .05 .22* .10 1.00 .58* .10* .36* .49* .26*
7. Effort .00 .09 .06 .14 .02 .67* 1.00 .08* .45* .41* .19*
8. Difficulty .07 .05 .00 .20* .12 .17* .12 1.00 .11* .14* .57*
9. Luck .11 .06 .05 .05 .02 .35* .42* .14 1.00 .34* .01
10. Self-confidence .02 .05 .09 .08 .01 .50* .49* .24* .51* 1.00 .30*
11. Self-deprecation .07 .02 .01 .30* .34* .25* .22* .50* .28* .56* 1.00
*p .05
M. M. Harrod and R. T. Serpe
Analysis
We use Structural Equation Modeling (SEM) to analyze the data, which provides us
with a parsimonious approach for analyzing differences among parameters between
groups. SEM was selected to assess differences between the employed and unem-
ployed, and the ginvariant subroutine in STATA supported the use of a groups
analysis. Working from an unconstrained baseline model, we used ginvariant to
guide in the determination of which coefficients can be constrained to equality.
Analyses were conducted using multiple imputations of the missing values account-
ing for missing values within the data. Overall, the fit of the structural equation
model provides a good fit to the data. Chi-Square 30.95, df ¼ 43, p of chi-square .92;
RMSEA ¼ .000. These statistics indicate the model reasonably represents the
relationships presented in the data.
Attributions and Identity Prominence
We begin by examining our first hypotheses, making predictions of the relationship

of self-attributions on identity prominence. Table 10.3 reports the SEM results for
the model. Beginning with the first four hypotheses relating to the internal attribu-
tions of ability and effort on identity prominence. Hypothesis 1a predicts that
internal attributions will increase identity prominence for the employed. The internal
attribution of ability positively impacts identity prominence (β ¼ .11, p < .05).
Hypothesis 1b predicts that internal attributions to ability will reduce identity
prominence for the unemployed. Hypothesis 1b is supported (β ¼ .11, p < .05).
Results also show no statistical difference between the coefficients of ability on
identity prominence for the employed and unemployed. Thus, that coefficient was
constrained to be equal. The next two hypotheses provide predictions about the
effect of effort on both groups. Hypothesis 1c posits that an internal attribution to
effort will positively affect identity prominence. Unexpectedly, this hypothesis is not
supported. Hypothesis 1d predicts that attributions to effort will increase identity
prominence among the unemployed. Hypothesis 1d is not supported.
The next four hypotheses provide our predictions of how external attributions
affect identity prominence. Hypothesis 2a predicts that the external attribution to
task difficulty will negatively impact identity prominence for the employed. This
hypothesis is not supported. Hypothesis 2b predicts that task difficulty will nega-
tively impact identity prominence. Hypothesis 2b is not supported. Hypothesis 2c
predicts that making external attributions to luck will have a negative impact on
identity prominence for the employed. Hypothesis 2c is not supported. Finally,
Hypothesis 2d posits that luck's external attribution will have a negative impact on
identity prominence among the unemployed. The hypothesis is not supported.
Table 10.3 Standardized coefficients of task difficulty, luck, ability, effort, and prominence on
confidence and deprecation role-specific self-esteem
Task Confidence Deprecation
difficulty Luck Ability Effort Prominence R-S SE R-S SE
Unemployed group, N ¼ 160
Task – – – – 2.00 2.09* .36*
difficulty
Luck – – – – .08 .24* .18*
Ability – – – – .11* .34* .19*
Effort – – – – .06 .15* 2.06
Prominence – – – – – .04 .29*
Sex, female 2.07 .09* .04 .10* 2.02 .00 .03
¼1
Personal .07* .03 .04 .13* .04 .01 .08*
income
Educ. .04 2.02 .05 2.02 2.08* .09 2.01
attainment
Age 2.22* 2.02 .13* .04 2.07* .00 .13*
Employed group, N ¼ 644
Task – – – – 2.00 2.13* .51*
difficulty
Luck – – – – .31* .05 .02
Ability – – – – .11* .34* .14*
Effort – – – – .05 .14* .09*
Prominence – – – – – .20* .03
Sex, female 2.07 .13* .06 .15* 2.03 .01 .03
¼1
Personal .08* .16* .06 .00 .07 .02 .09*
income
Educ. .04 2.03 .07 2.04 2.11* .06 2.01
attainment
Age 2.22* 2.03 .18* .06 2.09* .00 .13*
Note: Bolded Variables are constrained across groups; Χ ¼ 30.95, df ¼ 43, p of chi-square .92;
2
RMSEA ¼ .000
*p .05
Identity Prominence and Self-Esteem
Hypotheses 3a and 3b make predictions about the relationship between identity

prominence and self-confidence. Hypothesis 3a expects a positive impact of identity
prominence on self-confidence for the employed, while Hypothesis 3b predicts a
negative impact on the unemployed. Consistent with H3a, identity prominence has a
positive relationship with self-confidence among the employed (β ¼ .20, p < .05).
H3b is not supported with a non-significant effect from prominence on self-confi-
dence for the unemployed.
Hypotheses 4a and 4b offer predictions about the expected relationship between

identity prominence and self-deprecation. Hypothesis 4a expects a negative impact
of identity prominence on self-deprecation for the employed, and Hypothesis 4b
anticipates a positive relationship for the unemployed. For the employed, the effect
of identity prominence on self-depreciation is non-significant, and we find no
support for Hypothesis 4a. For the unemployed, we find no support for Hypothesis
4b. Unexpectedly, the effect of identity prominence negatively impacts self-depre-
cation (β ¼ .29, p < .05).
The next hypotheses provide predictions about the relationship between internal
self-attributions and the two dimensions of self-esteem. Hypotheses 5a-d present
predictions about the two internal dimensions on both dimensions of self-esteem.
Hypothesis 5a posits that ability will have a positive impact on self-confidence.
Hypothesis 5a is supported. For both groups, we see that self-attribution to ability
positively impacts self-confidence (β ¼ .34, p < .05) for both the employed and
unemployed. Moreover, results show no statistical difference between the coeffi-
cients of ability on identity prominence for the employed and unemployed, so the
coefficient was constrained to be equal. Hypothesis 5b posits that attributions to
ability will be positively related to self-deprecation. Hypothesis 5b is supported. For
both groups we see that making attributions to ability has a positive impact on self-
deprecation for the employed (β ¼ .14, p < .05) and unemployed (β ¼ .19, p < .05).
Additionally, the results showed no statistical difference between the groups leads to
constraining the coefficient to be equal between the groups.
Hypotheses 5c and 5d provide the expected relationship between effort and the
two dimensions of self-esteem. Hypothesis 5c is supported. Consistent with expec-
tations, effort has a positive impact on self-confidence. Table 10.3 shows this is true
for both the employed (β ¼ .14, p < .05) and unemployed (β ¼ .15, p < .05).
Additionally, the results showed no statistical difference between the groups leads to
constraining the coefficient to be equal between the groups. On the other hand,
Hypothesis 5d receives mixed support. Support is found for the employed (β ¼ .09,
p < .05), but among the unemployed, it is non-significant.
Hypotheses 6a–d posit the relationship between the two external dimensions and
both dimensions of self-esteem. Hypothesis 6a posits that task difficulty will nega-
tively impact self-confidence. We find support for Hypothesis 6a, attributions to task
difficulty leads to reduced self-confidence for the employed (β ¼ .13, p < .05) and
the unemployed (β ¼ .09, p < .05). The results also show no statistical difference
between the groups leads to constraining the coefficient to be equal between the
groups. Support is also found for Hypothesis 6b, which states that attributions to task
difficulty will negatively impact self-deprecation. Within Table 10.3 we find support
among the employed (β ¼ .51, p < .05) and the unemployed (β ¼ .36, p < .05).
The final two hypotheses provide hypotheses of the negative relationship of luck
on the two dimensions of self-esteem. Hypothesis 6c posits that attributions to luck
will negatively impact self-confidence. Within Table 10.3, we find no support for
Hypothesis 6c. Instead, the results show the unexpected effect that luck has a
positive impact on self-confidence for the unemployed (β ¼ .24, p < .05).
Hypothesis 6d posits a negative relationship between an attribution to luck and
self-deprecation. This hypothesis is not supported. In fact, the results show an

unexpected positive effect (β ¼ .18, p < .05).
To summarize, the results are mixed relative to the relationships proposed.
Beginning with the internal attributions, ability has the most consistent finding. In
all cases, the anticipated effect of ability was supported. Effort’s impact was mixed.
Effort failed to impact identity prominence significantly, but it was supported in
three of four instances related to the two dimensions of self-esteem. The exception
was the impact of effort on self-deprecation among the unemployed group. The
effects of the external attributions were also mixed relative to the hypotheses.
Hypotheses including task difficulty had mixed results. Unexpectedly, task difficulty
failed to have the expected impact on identity prominence. However, task difficulty
had the anticipated effect on both dimensions of self-esteem. Turning to the expected
impact of luck, it failed to be significant in all cases. Finally, only for the hypothesis
linking identity prominence to self-confidence among the employed did we find a
significant effect.
Discussion
This paper’s central question asks how people holding normative and counter-
normative identities evaluate themselves when they attribute their employment
status in terms of internal and external causes. In this research, we have used the
two dimensions of self-esteem to assess if there are differences by employment
status of employed and unemployed. Across the four attributional conditions, the
effects of prominence on self-confidence and self-deprecation demonstrate that
employment status is strongly related to self-esteem. As hypothesized, we find
significant and positive effects of prominence on self-confidence but not on self-
deprecation. We find that prominence has no significant effects on self-confidence
and a significant and negative effect on self-deprecation for the unemployed.
This research provides new findings for identity theory and self-esteem. Self-
attributions provide feedback for one’s identity. Self-attributional feedback may be
different for people holding normative and counter-normative identities. In this
study, those who hold the unemployed counter-normative identity are seeking to
move to the employed normative identity. The unemployed in this analysis wish to
be employed. For them, the meanings associated with being unemployed are likely
to differ from their self-view. Their behavior to actively seek employment is
motivated to change from a counter-normative identity to the normative identity of
being employed. Comparing normative and counter-normative identities provides
the theoretical rationale to look at the two dimensions of self-esteem. Empirically,
Owens’ (1993, 1994) sociological work and the meta-analysis of Huang and Dong
(2012) emphasize that the RSE contains both positive and negative aspects. In our
sample, the tests show that the employed and unemployed groups differ significantly
from one another. If we maintained a unidimensional self-esteem construct, it would
fail to capture how attributions and identity prominence differentially impact self-
esteem for those employed and those who have been unsuccessful in finding
employment. The RSE subscales of self-confidence and self-deprecation provide
the differentiation necessary to compare the two identity statuses. While Rosenberg’s
Self-Esteem Scale is most often used as a unidimensional measure, there is evidence
that the scale has both self-confidence and self-deprecation sub-scales (Huang &
Dong, 2012; Owens, 1993, 1994). Our research supports the theoretical reasons for
using the two sub-scales to address the hypothesized differences between holding a
normative and counter-normative identity.
Self-attributions have been theorized as part of the meaning-making process
within identity theory (Rosenberg, 1988; Stryker & Gottlieb, 1981). Self-attributions
provide information that is relevant to understanding outcomes associated with one’s
identity. Thus, attributions become part of the internalized meanings that persons use
contextually in role-specific self-view.
Our research investigates the relationship between identity, self-esteem, and self-
attributions. Our hypotheses can be grouped into three broad categories based on the
connections of self-attributions, identity, and self-esteem in our model.
The first set of hypotheses examines self-attributions about the cause of the
respondents’ work status and the effect that attributions to ability, effort, task
difficulty, and luck have on the identity prominence for being employed or unem-
ployed. Support for these predictions, while mixed, demonstrates differences based
on work identity status. Overall, the results suggest our predictions for differential
patterns for the normative employed respondents and the counter-normative unem-
ployed respondents are warranted for future research. The internal attribution of
ability positively impacts identity prominence for both the employed and unem-
ployed. However, the internal attribution of effort does not significantly affect
identity prominence for either group. Importantly, this pattern can be explained by
the stability of the two internal attributions. Ability is stable in the context of any
specific identity. If one evaluates the self as having high ability, we expect attribu-
tions to ability to not differ between being employed or not employed in the case of
working. Effort, on the other hand, is unstable in the context of any specific identity.
As an unstable attribution, the causal ascription to effort will vary based on the
person’s perception of their level of effort. In this case, the perceived level of effort
does not significantly affect identity prominence for either group.
The external attributions of the stable task difficulty and unstable luck on identity
prominence are not supported. Task difficulty does not have a significant impact on
identity prominence for either group. As a stable causal attribution, perceptions of
task difficulty do not affect identity prominence. The unstable characteristic of luck
does not significantly affect identity prominence for the unemployed, suggesting that
luck is not self-relevant for the prominence for a counter-identity that one is actively
seeking to change.
Overall, in this study, the impact of self-attributions on identity prominence is
quite modest. Only ability is positively related to prominence, unexpectedly so for
the unemployed. Moreover, the coefficients for ability could be constrained to
equality for the employed and unemployed groups, suggesting that perceptions of
ability may not be affected by the context of normality versus counter-normality.
Ability is a stable, internal casual attribution; therefore, it is not surprising that the
respondents in this sample did not see any variation in their ability to be employed,
even they are unemployed.
The second set of hypotheses focus on identity prominence’s effect on the two
sub-scales of the Rosenberg measure of self-esteem—self-confidence, and self-
deprecation. Again, we find mixed support for the hypotheses by identity status.
The results suggest a general pattern that self-feelings follow societal norms, with the
normative employed identity holding more positive feelings. For the two groups,
identity prominence is linked with self-feelings consistent with their normative or
counter-normative position. Identity prominence increases self-confidence among
the employed. There are no significant effects for identity prominence on the self-
deprecation sub-scale. For the unemployed, the pattern is reversed. Identity promi-
nence of the unemployed the self-confidence sub-scale does not increase or decrease
self-esteem. Importantly, identity prominence for the unemployed on the self-
depreciation sub-scale decreases self-esteem.
The final hypotheses explore how self-attributions influence self-confidence and
self-deprecation. Here, too, we see differential support aligned with internal and
external self-attributions, but the preponderance of the results are consistent with the
hypotheses. Beginning with the internal attributions of ability and effort, we see that
both the self-confidence and self-deprecation sub-scales increase self-esteem. These
findings support earlier theorizing that self-attributions are an important source of
information about the self (Rosenberg, 1979, 1988; Stryker & Gottlieb, 1981).
Believing that the self is responsible because of one’s ability and effort is linked to
increased positive and negative self-esteem dimensions. In both cases, one’s actions
are responsible for the good and the bad things that happen. These findings are
consistent with Owens’ (1993, 1994) theorizing about self-motives and self-esteem.
The external attributions of task difficulty and luck are where we see a less clear
pattern of results. Respondents attributing their work status to task difficulty uni-
formly report a decrease in self-esteem for both the self-confidence and self-depre-
cation sub-scales. When persons perceive a task as difficult, it is not consistent with
positive self-feelings. These are precisely the results that prior literature would
expect.
Unexpectedly, the results for luck do not follow prior attribution research. For the
unemployed, luck's unstable characteristics increase self-esteem for both the self-
confidence and self-depreciation sub-scales. What explains the surprising results?
Traditionally, luck is defined as both external and unstable. The stability dimension
indicates the amount of control over the dimension. Thus, luck is outside of the
actor’s sphere of influence as well as being uncontrolled by the self. If examined
from the actor’s viewpoint, outcomes resulting from luck are not useful predictors of
future events.
In sum, our results focusing on the worker identity are not strong. However, this
one test using the worker identity suggests self-attributions may have a place within
identity theory. Self-attributions are an excellent way to implement Stets and Burke’s
(2014) call to gather data from the actor’s perspective. Self-attributions reflect the
actor’s interpretations of outcomes—the how and why events unfolded as they did.
Research in identity theory is likely to benefit from the inclusion of self-attributions

when focusing on the motivation for identity change, differences between normative
and counter-normative identities, and life events that change the number and con-
figuration of identities to provide information on how the person understands the
outcomes of social interaction.
Additionally, this research provides evidence that a person’s experiences and self-
evaluations are differentially impacted by holding a normative or counter-normative
identity. The unidimensionality of the RSE will continue to be the general use of the
scale. Our research demonstrates that when dealing with data consisting of counter-
normative, not by choice, and are seeking change, the use of the self-confidence and
self-denigrating sub-scales may produce a nuanced set of findings.
The current research is the first to examine how attributions affect identity
prominence and self-esteem. As with any single study, the reader should exercise
a degree of skepticism until additional research can support and refine these results.
In addition to examining other identities, our confidence would be strengthened if we
had longitudinal data to analyze these relationships. With longitudinal data, we
could have confidence in the causality of the relationships explored here. While
the data fit the model very well, cross-sectional data is no substitute for
longitudinal data.
While the results are mostly consistent with theoretical predictions, the unex-
pected effect of luck on self-esteem for the unemployed suggests additional theoret-
ical work is needed. Although this finding is unexpected, it does lend additional
support to the burgeoning literature that people who hold normative and counter-
normative identities depart in significant ways. Precisely how or why it happened
here is something we will need to leave to future researchers.
Future research may wish to expand on the present study in numerous ways.
Among these opportunities would be contributing to the nascent research exploring
how the structural and perceptual control programs intermingle to elucidate an
actor’s behavior (Markowski & Serpe, 2018; Stets et al., 2017). Second, it might
be fruitful to fully explore how structural factors influence respondents’ perceptions
of self. In the future, researchers should replicate and extend our analyses focusing
on different identities and outcomes.
Appendix A. Principle Components Factor Loading

of Role-Specific Self-Confidence and Self-Deprecation
Item Self-confidence Self-deprecation

I feel that I am a person of worth, at least on an equal .81 –
basis with others.
I feel that I have a number of good qualities. .82 –
I am able to do things as well as most other people. .77 –
I take a positive attitude toward myself. .83 –
(continued)
Item Self-confidence Self-deprecation

On the whole, I am satisfied with myself. .76 –
All in all, I am inclined to feel that I am a failure. – .83
I wish I could have more respect for myself. – .76
I certainly feel useless at times. – .88
I feel I do not have much to be proud of. – .81
At times, I think that I am not good at all. – .89
Eigenvalue ¼ 3.18 Eigenvalue ¼ 3.49
Appendix B. Principle Components Factor Loadings

of Identity Prominence
Item Identity prominence

Being [employed/unemployed] is an important reflection of who I am 0.91
I have a strong sense of belonging to other [employed/unemployed]. 0.79
I have come to think of myself as a [employed/unemployed]. 0.77
Being a [employed/unemployed] is an important part of my self-image. 0.87
Eigenvalue ¼ 2.79
References
Adams, R. E., & Serpe, R. T. (2020). Identity meaning discrepenacies and psychological distress: A
partial test of incorporating identity theory and self-definitions into the stress process model. In
R. T. Serpe, R. Stryker, & B. Powell (Eds.), Identity and symbolic interaction: Deepening
foundations, building bridges (pp. 293–316). Springer.
Braunsberger, K., Wybenga, H., & Gates, R. (2007). A comparison of reliability between telephone
and web-based surveys. Journal of Business Research, 60(7), 758–764.
identity theory: An empirical examination. Social Psychology Quarterly., 77(3), 231–252.
Brenner, P. S., Serpe, R. T., & Stryker, S. (2018). Role-specific self-efficacy as the precedent and
Burke, P. J., & Harrod, M. M. (2005). Too good to be believed? Social Psychology Quarterly, 68
(4), 359–374.
Crittenden, K. S. (1983). Sociological aspects of attribution. Annual Review of Sociology, 9,
425–446.
Crittenden, K. S., & Wiley, M. G. (1980). Causal attribution and behavioral response to failure.
de Araujo, P., & Lagos, S. (2013). Self-esteem, education, and wages revisited. Journal of
Economic Psychology., 34(1), 120–132.
Gecas, V., & Seff, M. A. (1989). Social class, occupational conditions, and self-esteem. Social
Psychology Quarterly., 32(3), 353–364.
Heider, F. (1958). The psychology of interpersonal relations. John Wiley and Sons.
Howard, J. A. (2000). Social psychology of identities. Annual Review of Sociology., 26, 367–393.
Howard, J. A., & Pike, K. C. (1986). Ideological investment in cognitive processing: The influence
of social statuses on attribution. Social Psychology Quarterly, 49(2), 154–167.
Huang, C., & Dong, N. (2012). Factor structures of the Rosenberg self-esteem scale: A meta-
analysis of pattern matrices. European Journal of Psychological Assessment, 28(2), 132–138.
Kelley, H. H. (1967). Attribution theory in social psychology. In Nebraska symposium on motiva-
tion (Vol. 15, pp. 192–238). University of Nebraska Press.
Kelley, H. H. (1971). Attribution in social interaction. New York: General Learning Press.
Long, B. L. (2016). Stigmatized identities: Choice, accessibility, and authenticity. In J. E. Stets &
University Press.
of prominence and salience in the verification and self-esteem relationship. In Advances in
group processes (pp. 75–102). Emerald Group Publishing Limited.
Markus, H., & Wurf, E. (1987). The dynamic self-concept: A social psychological perspective.
Annual Review of Psychology, 38, 299–337.
McCall, G. J., & Simmons, J. L. (1978). Identities and interactions: An examination of human
associations in everyday life. Macmillan.
Owens, T. J. (1993). Accentuate the positive—and the negative: Rethinking the use of self-esteem,
self-deprecation, and self-confidence. Social Psychology Quarterly, 56(4), 288–299.
Owens, T. J. (1994). Two dimensions of self-esteem: Reciprocal effects of positive self-worth and
self-deprecation on adolescent problems. American Sociological Review, 59(3), 391–407.
Owens, T. J., Robinson, D. T., & Smith-Lovin, L. (2010). Three faces of identity. Annual Review of
Sociology, 36, 477–499.
Owens, T. J., & Serpe, R. T. (2003). The role of self-evaluation in identity salience and commit-
ment. In P. J. Burke, T. J. Owens, R. T. Serpe, & P. A. Thoits (Eds.), Advances in identity theory
and research (pp. 85–102). Kluwer Academic/Plenum Publishers.
Rosenberg, M. (1988). Self-objectification: Relevance for the species and society. Sociological
Forum, 3(4), 548–565.
(1), 141–156.
Serpe, R. T., & Stryker, S. (2011). The symbolic interactionist perspective and identity theory. In
S. Schwartz, K. Luyckx, & V. Vignoles (Eds.), Handbook of identity theory and research
(pp. 225–248). Springer.
Shepelak, N. J., & Alwin, D. F. (1986). Beliefs about inequality and perceptions of distributive
justice. American Sociological Review, 51(1), 30–46.
Stets, J. E. (2003). Justice, emotion, and identity theory. In P. J. Burke, T. J. Owens, R. T. Serpe, &
P. A. Thoits (Eds.), Advances in identity theory and research (pp. 105–122). Kluwer Academic/
Plenum Publishers.
Stets, J. E., Brenner, P. S., Burke, P. J., & Serpe, R. T. (2017). The science identity and entering a
science occupation. Social Science Research, 64, 1–14.
Stets, J. E., & Burke, P. J. (2005). New directions in identity control theory. Advances in Group
Stets, J. E., & Burke, P. J. (2014). Emotions and identity nonverification. Social Psychology
Quarterly, 77(4), 387–410.
social psychology (2nd ed., pp. 31–60). New York, NY.
Stryker, S. (1968). Identity salience and role performance: The relevance of symbolic interaction
theory for family research. Journal of Marriage and Family, 30(4), 558–564.
Stryker, S., & Burke, P. J. (2000). The past, present, and future of an identity theory. Social
Stryker, S., & Gottlieb, A. (1981). Attribution theory and symbolic interactionism: A comparison.
In J. H. Harvey, W. J. Ickes, & R. F. Kidd (Eds.), New directions in attribution research (Vol.
3, pp. 425–458). Lawrence Erlbaum.
Stryker, S., Serpe, R. T., & Hunt, M. O. (2005). Making good on a promise: The impact of larger
social structures on commitments. In S. Thye & E. Lawler (Eds.), Advances in group processes
(Vol. 22, pp. 93–124). JAI Press.
Wallace, H. M., & Baumeister, R. F. (2002). The effects of success versus failure feedback on
further self-control. Self and Identity., 1(1), 35–41.
Wang, L.-Y., Kick, E., Fraser, J., & Burns, T. J. (2010). Status attainment in America: The roles of
locus of control and self-esteem in educational and occupational outcomes. Sociological
Spectrum., 19(3), 281–298.
Weiner, B. (1986). An attributional theory of motivation and emotion. New York: Springer-Verlag.
Yeager, D. S., Krosnick, J. A., Chang, L. C., Javitz, H. S., Levendusky, M. S., Simpser, A., &
Wang, R. (2011). Comparing the accuracy of RDD telephone surveys and internet surveys
conducted with probability and non-probability samples. Public Opinion Quarterly, 75(4),
709–747.
Chapter 11
General and Specific Sources of Reflected
Appraisals
Mary Gallagher
Introduction
Social psychologists have long conceived of a self that develops through social
interaction (e.g., Kinch, 1963; Mead, 1934; Sullivan, 1953). Kinch (1963, p. 481)
defined the self as “qualities that the individual attributes to himself,” including
personal characteristics and roles. One way that individuals come to attribute certain
qualities to themselves is through their perceptions and interpretations of the ways in
which others behave toward and respond to them during social interaction (Kinch,
1963).
Actual appraisals are others’ views of or responses to an individual (i.e., what
others truly think of a person). Reflected appraisals, a term first used by Sullivan
(1953), are individuals’ perceptions of others’ behavior toward and responses to
them (i.e., what people think or believe others think of them). Sources of reflected
appraisals range from the specific to the general and can come from individuals with
whom a person interacts, groups to which a person belongs, and society (Mead,
1934). Reflected appraisals play an important role in identity theory, as they repre-
sent one of many possible forms of input from the social environment that individ-
uals receive about themselves and their identities.
According to identity theory, an individual has a set of self-meanings for each
group, role, and person identity they claim (Burke & Stets, 2009). During social
interaction, people seek to achieve identity verification, or consistency between
perceptual input meanings (e.g., reflected or actual appraisals, social comparisons)
and the self-meanings attached to the identity or identities that are activated in the
situation. Consistency is obtained by “controlling” perceptual input through mean-
ingful behavioral output. When there is a mismatch between perceptual input and the
M. Gallagher (*)
Department of Sociology, Kent State University at Stark, North Canton, OH, USA
224 M. Gallagher
meanings held in the identity standard (identity discrepancy or non-verification), a

person is expected to engage in behavior aimed at altering subsequent reflected
appraisals so that they, ideally, will more closely align with their identity standard.
At the same time, they may also respond to an identity discrepancy by slowly
shifting their self-image in the direction of the discrepant feedback (Stets et al.,
2020).
Reflected appraisals from others are operationalized in one of three different ways
in identity research: (1) as specific others who are assumed to be generally important
or close to the individual, irrespective of the identity or identities under consideration
(e.g., Stets et al., 2017; Stets & Harrod, 2004), (2) as specific others who are
theoretically pertinent to the particular identity or identities being studied (e.g.,
Asencio, 2011; Asencio & Burke, 2011; Gallagher, 2017), or (3) as non-specific
generalized “others” or “others in general” (e.g., Kalkhoff et al., 2016; Marcussen &
Gallagher, 2017; Marcussen & Large, 2003; Stets & Trettevik, 2016; Trettevik,
2016). The first two approaches are similar to one another in that they both concern
reflected appraisals from specific others. What distinguishes them is whether the
specific others are chosen because they are significant others who are presumed to be
important sources of feedback about all or most identities (the first approach), or
whether the specific others are chosen as sources of reflected appraisals because of
their theoretical relevance to the identity or identities being studied (the second
approach). Using the worker identity as an example, spouses, family, and friends are
generally important significant others who are not tied to the worker role, whereas
supervisors and co-workers are specific role complements who are theoretically
relevant to the worker role.
The third approach tends to be used in identity research under the, often implicit,
assumption that it captures perceived feedback from the generalized other or some
indefinite combination of relevant specific others. Other social psychological theo-
ries suggest potential alternative ways in which respondents might interpret the
meaning of questions about general others. In expectations states theory terms,
thinking of “others” or “others in general” might call up respondents’ sense of social
norms and expectations associated with an identity in ways that are akin to third-
order expectations, which refer to beliefs about what “most people” think about or
expect from a status group (Ridgeway & Correll, 2006). Similarly, in affect control
theory terms, questions about “others” or “others in general” might be interpreted as
being about fundamental sentiments (culturally shared meanings) attached to iden-
tities (Robinson et al., 2006).
Theoretically, reflected appraisals from both general and specific sources con-
tribute to the development and maintenance of self-views, and they also have the
potential to produce identity non-verification when they are inconsistent with self-
views held in the identity standard. Reflected appraisal and identity processes are not
expected to operate differently for different sources of feedback, but reflected
appraisals from specific sources may be more influential than reflected appraisals
from more distal, general sources. Specific others who are significant and familiar
should play a particularly important role in identity processes because the individ-
ual’s relationships with them tend to involve established patterns of interaction and
11 General and Specific Sources of Reflected Appraisals 225
greater investment (Burke, 1991). Groups in which individuals participate directly

(Shibutani, 1955) and those involving primary relationships (Cooley, 1902) also
represent specific others who are centrally important to the self. To assess the
cumulative and/or relative influence of reflected appraisals from specific and general
sources, and to compare findings across different studies, reflected appraisals from
each of these sources would need to be accounted for in a single study. However,
most extant research includes reflected appraisals from one type of source or the
other, but not both.
To address this gap in previous research, I simultaneously examine whether
reflected appraisals from general and specific others are independently associated
with self-views, paying particular attention to the relative strength of the association
between reflected appraisals from each respective source and self-views. I investi-
gate four identities (spouse/partner, friend, worker, and religious/spiritual) to deter-
mine the robustness of associations between different sources of reflected appraisals
and self-views. These identities are normative social roles that most adults hold and
in which they typically invest a great deal of time and energy. They also tend to be
centrally important for self-definition. A secondary aim of this study is to highlight
the theoretical importance of general others in identity processes and suggest
directions for future research that is more intentional in operationalizing reflected
appraisals from this source.
Background
Reflected Appraisals and Self-Views
Symbolic interactionists have long recognized others as central to the development

and maintenance of the self (e.g., Cooley, 1902; Kinch, 1963; Kuhn, 1964a; Mead,
1934; Sullivan, 1953). One way in which others influence the self is through
reflected appraisals, or individuals’ perceptions of others’ attitudes, behavior, and
responses toward them (Sullivan, 1953). During social interaction, individuals
observe the ways in which others react to them, and based on those observations,
make inferences about the type of person they are (Cooley, 1902). In Mead’s (1934)
terms, this process requires role-taking, or individuals’ ability to imagine how others
will respond to their gestures including their words or actions that carry shared
meanings. They use that information to draw conclusions about themselves and
guide their subsequent behavior. Role-taking sometimes involves adopting the
standpoint of the other, allowing the perceived attitudes of the other to become
internalized as one’s own (Turner, 1956). To the extent that the internalized attitudes
of others pertain to the self, individuals’ self-views will be, at least in part, influenced
by their perceptions of how others view them (Kinch, 1963).
An extensive body of research has found somewhat mixed support for the
reflected appraisal process, depending on the study sample and the attributes,
identities, or source(s) of reflected appraisals under consideration. Much of the
226 M. Gallagher
“mixed” support has come from research involving idiosyncratic samples that are
not easily compared with one another. For instance, studies find that college football
players’ self-views depend more on their objective performance, that is, how well
they actually played, than on reflected appraisals from coaches or teammates
(Felson, 1981), but among adolescents, reflected appraisals from parents, teachers,
and classmates have a strong impact on self-views (Bouchey & Harter, 2005; Felson,
1985).
Several more recent studies lend support to the reflected appraisal process across
a variety of adult samples. For example, a study of married couples finds that
reflected appraisals from one’s spouse influence commitment to caregiving and
breadwinning roles (Maurer et al., 2001). Research also shows that reflected
appraisals (how “others” would categorize their race based on their appearance)
influence racial identification among biracial Asian Americans (Khanna, 2004).
There also is evidence that, in counter-normative identities such as mental illness,
stigmatized reflected appraisals from mothers are associated with stigmatized self-
views (Markowitz et al., 2011). These studies demonstrate that reflected appraisals
are associated with self-views, but they utilize cross-sectional data that cannot
establish causal direction.1
The relationship between reflected appraisals and self-views is likely reciprocal
(Kinch, 1963) and influenced by social psychological biases such as projection, or
the tendency to assume that others share our opinions (Felson, 1985). However, the
few longitudinal studies that have examined the reflected appraisal process tend to
find that the direction of causation is primarily from reflected appraisals to self-
views. For instance, Ichiyama (1993) found a reciprocal relationship between
reflected appraisals and self-views on dimensions of dominance, affiliation, and
task orientation among participants in human relations training groups. Some degree
of projection was present, but the effect of reflected appraisals on self-views was
stronger than the reverse for two of the three dimensions examined. Additionally,
research has demonstrated that reflected appraisals from parents affect children’s
self-views related to athletics and popularity (Felson, 1989), and that reflected
appraisals of significant others (i.e., spouse or other important family members
and/or friends), and fellow inmates influence inmates’ criminal and drug user self-
views (Asencio & Burke, 2011).
Taken together, extant research generally supports the theoretical assumption that
reflected appraisals from many different sources influence self-views. But, because
most studies have focused on reflected appraisals from a single type of source
(general or specific), this body of research collectively leaves open questions about
the relative influence of reflected appraisals from different sources on self-views. As
discussed in more detail below, theory and research on reflected appraisals generally
suggests that reflected appraisals from specific others may exert a particularly strong
1
Markowitz et al. (2011) utilized two waves of data to examine relationships between stigma,
reflected appraisals, and recovery outcomes, but self and reflected appraisals were only measured in
the second of those waves, so longitudinal relationships between them could not be estimated.
influence on self-views. Additionally, most prior research has relied on purposive

samples that cannot be used to make inferences to a general population. Examining
the association between general and specific reflected appraisals and self-views
across a broader range of typical adult social roles will help to determine the extent
to which a particular source of reflected appraisal may be more consequential for
self-views in some identities more than others.
Specific Sources of Reflected Appraisals
According to Mead (1934), sources of reflected appraisals fall on a continuum

between the specific and the general. On the more specific end of the spectrum are
role partners, or individuals with whom a person interacts by virtue of holding a
particular social position or identity. Role partners facilitate each other’s role
performances and rely on one another for resources such as social support (Stryker
& Statham, 1985). Kuhn (1964b) noted that significant others (who he also terms
“orientational others”) have a strong influence on the self because relationships with
them involve emotional and psychological commitment, provide the self with a
general vocabulary and categories of self, other, and meaningful roles, and they
contribute to the maintenance of and changes in the self. Similarly, James (1890)
argued that romantic partners may have an especially large effect on the self because
of the intense emotions involved in such intimate relationships.
Although specific role partners are important influences on self-views, one does
not necessarily have to take the role of any specific individual to account for and
adopt the values and attitudes of a group as their own (Turner, 1956). Reference
groups provide members with norms, goals, social comparisons, and an audience
whose approval is sought (Kemper, 1966). Just as personal relationships have
different characteristics, groups vary along many different dimensions (Kuhn,
1964b). Groups in which individuals participate directly and those that involve
primary relationships are thought to be most influential for individuals’ sense of
self (Shibutani, 1955). For example, Cooley (1902) argued that primary groups such
as the family and others who are present during childhood have the largest impact on
the self. Thus, in addition to role partners, reference groups are a type of specific
other that contribute to the development and maintenance of self-views.
Early symbolic interactionist theories highlight the important role that specific
others play in the development and maintenance of self-views, and identity research
supports that basic tenet, especially for romantic partners. For example, men’s
spouses/partners have a large impact on their self-views as a parent (Collett et al.,
2019). Additionally, members of newly married couples whose spouse’s appraisals
of them differ from their own change their self-views over time to more closely
match their spouse’s views of them (Cast & Cantwell, 2007).
Recall that one common operationalization of reflected appraisals is specific
others who are assumed to be generally important or close to the individual,
irrespective of the identity or identities under consideration (e.g., Stets et al., 2017;
228 M. Gallagher
Stets & Harrod, 2004). These others often include spouses/partners as well as other
role complements such as family members, friends, and co-workers, but reflected
appraisals from multiple specific sources are typically combined into a single
composite score. In these studies, composite scores are justified by the theoretical
assumption or empirical demonstration that reflected appraisals from two or more
others are highly correlated and/or that as a group, they capture one’s general sense
of how they are viewed by others with whom they regularly interact while
performing roles associated with their identities. However, combining reflected
appraisals from different sources precludes the detection of any potential differences
in their effects on outcomes, and there are theoretical and empirical reasons to expect
that various specific others might have different effects on self-views and other
outcomes of interest to identity researchers.
Proximate social structures such as patterns of relationships among members of
interpersonal networks shape individuals’ self-views and behaviors (Merolla et al.,
2012; Stryker et al., 2005). Each member of a person’s social network likely affects
their self-views to some degree, but the extent of that influence for any given
individual may differ depending on the exact nature of the relationship with that
person. Identity theorists address these potential differences in predicting that incon-
sistent feedback about one’s self that is perceived to come from a significant, close
other will be more distressing than feedback that is perceived to come from an
acquaintance or stranger (Burke, 1991, 1996). Given that personal relationships vary
on several dimensions such as length of acquaintance, depth of relationship, type of
relationship (e.g., partner versus friend), and relative status, reflected appraisals from
different specific sources who vary on those and other factors may have different
consequences for self-views.
A second common operationalization of reflected appraisals is specific others
who are theoretically pertinent to the identity or identities being studied. Research
that takes this approach finds evidence of the differential effects of specific others on
self-views, based on their characteristics and/or the nature of their relationship with
the object of their feedback (i.e., the person whose self-views are being assessed).
For instance, Asencio and colleagues found that reflected appraisals from familiar,
frequent, and legitimate others differentially affected criminal and worker self-views
in a sample of incarcerated adults (Asencio, 2011; Asencio, 2013; Asencio & Burke,
2011). In a sample of newly married couples, Cast et al. (1999) found that the lower
status spouse’s self-views and views of their spouse conformed to that of the higher
status spouse. Finally, Stets and Asencio (2008) found that, in a lab setting, partic-
ipants reacted more strongly to feedback that they thought came from a supervisor
(a legitimate authority figure) versus a non-supervisor.
Regardless of the differences that exist between various specific others whose
feedback is theoretically relevant to an identity, compared to general others who are
more abstract and whose significance to the identity is less apparent, specific others
may exert a stronger influence on self-views. In fact, identity theory (Burke, 1991,
1996) and foundational symbolic interactionist theories ascribe central importance to
reflected appraisals from specific others who are role partners (Kuhn, 1964b) and
members of groups in which individuals directly participate (Cooley, 1902;
Shibutani, 1955) in the development and maintenance of self-views. Nevertheless,

reflected appraisals from general sources may also impact self-views.
General Sources of Reflected Appraisals
The concept of the generalized other is broad, encompassing roles, social norms, and
expectations that members of a given culture have learned through socialization. Just
as individuals can take the role of specific others with whom they interact and of
specific groups to which they belong, they can also take the role of their society and
draw conclusions about who they are, and how they are being perceived and
evaluated in relation to overarching cultural standards and expectations (Mead,
1934).
Much social psychological theory and research has been devoted to answering the
question of whether, and to what extent, individuals are able to accurately distin-
guish between the views about them that are held by specific other people (see
Wallace & Tice, 2012 for a relatively recent review). Felson (1989) argued that
individuals’ perceptions of how they are viewed by specific significant others cannot
be distinguished, and thus, it is only reflected appraisals from general others that
influence self-views. Wallace and Tice (2012) suggested that this argument is
consistent with Mead’s (1934) conception of the generalized other, because the
judgements driving the reflected appraisal process are assumed to involve the
assessment of the collective views of others rather than the views of individual
specific others.
A third common operationalization of reflected appraisals is non-specific gener-
alized “others” or “others in general” (e.g., Kalkhoff et al., 2016; Marcussen &
Large, 2003; Marcussen & Gallagher, 2017; Stets & Trettevik, 2016; Trettevik,
2016). This approach tends to be used in identity research under the, often implicit,
assumption that it captures perceived feedback from the generalized other or some
indefinite combination of specific others whose views are relevant to an identity. It
also tends to be used in studies that test aspects of the perceptual control process. For
example, Kalkhoff et al. (2016) found that discrepancies between reflected
appraisals from general others and self-views in the spouse, parent, friend, and
person identities were associated with negative emotions and distress.
The focus on general others in the aforementioned studies is consistent with
Felson’s (1989) argument about the ways in which reflected appraisals are perceived
as the views of the collective rather than of individual others. It also is consistent
with Wallace and Tice’s (2012) interpretation of the generalized other. If people
perceive the views of others as a collective whole rather than as unique feedback
from specific individuals, then reflected appraisals from general others may be more
strongly associated with self-views compared to reflected appraisals from any single,
specific source.
Almost no prior research has compared the relative influence of reflected
appraisals from general and specific others in identity processes. The one exception
230 M. Gallagher
is Marcussen and Asencio’s (2016) study of individuals who reported having a

mental health problem. Respondents rated the extent to which they thought others in
general and specific others (coworkers, partner, and family members) saw them
primarily in terms of their mental illness. They found that reflected appraisals from
coworkers, partners, and family members were significantly associated with mental
illness self-views, but reflected appraisals from others in general were not. This
finding supports symbolic interactionist theorizing about the stronger impact that
reflected appraisals from specific others should have on self-views relative to
reflected appraisals from general others, though it has yet to be empirically demon-
strated in more conventional adult social roles.
The Present Study
Prior research demonstrates that the impact of reflected appraisals from others on
self-views may depend upon the source and/or the identities or attributes being
evaluated, but it is limited in two ways. First, most studies have focused on reflected
appraisals from a single type of source (general or specific). This hinders our ability
to compare results across studies and systematically identify patterns in the relative
influence of reflected appraisal from different sources in the reflected appraisal,
perceptual control, and other processes of interest to social psychologists. Second,
much of the prior research on the reflected appraisals processes has included either
samples of adolescents or purposive samples of adults who hold somewhat transitory
(e.g., college football players) or atypical (e.g., criminal) identities that cannot be
used to make inferences to a general population of adults in established or common
social roles.
The main contribution of this research is to simultaneously examine reflected
appraisals from specific (role partners or groups) and general others across four
different identities in a sample of adults in the general population. This will advance
our understanding of the scope of the reflected appraisal process by directly
assessing whether reflected appraisals from specific and general sources are inde-
pendently associated with self-views and exploring the relative strength of those
associations. It will also highlight the theoretical importance of general others in
identity processes and suggest directions for future research that is more intentional
in operationalizing reflected appraisals from this source.
Methods
Sample
Data for this study were collected in 2012 as part of a random-digit-dialing telephone
survey of identity meanings, self-evaluation, and psychological distress among a
national multi-stage probability sample of 502 individuals who were 18 or older and
living in the 48 contiguous United States. In this study, I examine four identities:
spouse, worker, friend, and religious.2 Two of the four identities can be conceptu-
alized as obligatory identities that most individuals will take on at some point in their
lives (spouse and worker), and the other two are more voluntary identities (friend
and religious) that individuals can more easily enter and exit (Thoits, 2003). Each
respondent was asked about two identities (one from the first category and one from
the second category) that were randomly chosen from the total number of identities
for which they were screened as eligible.3
Five-hundred and two respondents participated in the survey, and as Table 11.1
illustrates, the final analysis sample is comprised of 501 of those individuals (one
was excluded due to missing data on age). Eighty-four percent of the respondents are
white, nine percent are black, two percent are Latino, and five percent identify as a
member of a racial/ethnic group other than those three. Seventy-one percent of
respondents are female, 56% are employed, 30% are single, 3% percent are
cohabiting, and 69% are married. The average age of the sample is 56 years old.
The average number of years of education is 15, which represents a few years of
education beyond high school.
Measures
Self-views are the dependent variables, and they are measured for each of the four
identities by asking respondents, “On a scale of 0–10, where 0 is not at all good, and
10 is very good, how would you rate yourself as a [name of identity]?” (Stets &
Harrod, 2004; Stets et al., 2017). Higher values indicate more positive evaluations of
self in each respective identity. Descriptive statistics for self-views appear in
Table 11.1. Mean self-views are quite positive across all four identities, ranging
from 8.08 in the friend identity to 9.00 in the worker identity.
Reflected Appraisals are the independent variables. Adults in established social
roles have developed relatively stable patterns of interaction during which they
continuously receive information about others’ views and evaluations of them.
When considering others’ evaluations, people tend to focus on whether they are
2
Six identities were included in the original study. Each respondent was questioned about two of the
six identities which were randomly selected from those for which they were screened as eligible.
The family identity was excluded from this study because of its potential overlap with the spouse
identity. The volunteer/social group identity was excluded because it was less well-defined than the
other identities, and thus, could represent different social roles for different respondents (e.g., a
book club, volunteer work at a hospital, volunteer tasks related to a child’s schooling or athletics).
3
To ensure a relatively equal number of cases for each identity, the more common identities (e.g.,
friend) were capped around 170 cases. Interviews lasted for about 15 min, and the response rate was
58%. The response rate was calculated using the American Association for Public Opinion
Research (AAPOR) response rate 4 (American Association for Public Opinion Research, 2009).
232 M. Gallagher
Table 11.1 Means, standard Mean SD Range

deviations, and ranges for
Demographic variables (N ¼ 501)
study variables
White 0.84 0–1
Black 0.09 0–1
Latino 0.02 0–1
Other 0.05 0–1
Female 0.71 0–1
Age (in years) 55.68 14.85 18–93
Education (in years) 15.37 3.04 2–22
Employed 0.56 0–1
Single 0.30 0–1
Cohabiting 0.03 0–1
Married 0.69 0–1
Identity variables
Spouse (N ¼ 184)
Self-view 8.11 1.72 0–10
General reflected appraisal 8.13 1.77 0–10
Specific reflected appraisal 8.29 1.92 0–10
Friend (N ¼ 184)
Self-view 8.08 1.84 0–10
Worker (N ¼ 136)
Self-view 9.00 1.06 0–10
Religious (N ¼ 172)
Self-view 8.23 1.63 0–10
being perceived in “some globally positive, negative, or neutral way” (Shrauger &
Schoeneman, 1979, p. 559). The reflected appraisal measures in this study assess
respondents’ general perceptions of how positively others evaluate them in particular
social roles.
For each of the four identities, respondents are asked to report reflected appraisals
of general and specific others. The general reflected appraisal question asks, “On a
scale of 0–10, where 0 is not at all good, and 10 is very good, how would others rate
you as a [name of identity]?” (Stets & Harrod, 2004; Stets et al., 2017). The specific
reflected appraisal questions are worded similarly, except “others” is replaced with a
role partner or group that is relevant to each respective identity. In the spouse/partner
identity, for example, the specific other question asks, “On a scale of 0–10, where
0 is not at all good, and 10 is very good, how would your spouse/partner rate you as a
spouse/partner?” The specific others for the remaining identities are closest friend for
the friend identity, co-workers for the worker identity, and others who share the same
religious/spiritual affiliation for the religious/spiritual identity.
For all sources of reflected appraisal, higher values indicate a more positive
perceived evaluation of the respondent in that identity. As shown in Table 11.1,
reflected appraisals from general others are quite positive across all four identities,
ranging from 7.56 in the religious identity to 8.81 in the worker identity. Reflected
appraisals from specific others are also very positive across all four identities,
ranging from 7.67 in the religious identity to 8.89 in the friend identity.
Demographic Measures The linear regression analyses control for measures of
social statuses that may be associated with self-views. Employment status is coded
as a binary variable where 1 ¼ employed full or part time, with unemployed as the
comparison group. Education is coded as a continuous variable representing years of
schooling completed. Race is coded into binary variables for black ¼ 1, Latino ¼ 1,
and other ¼ 1, with white as the comparison group. Gender is coded as a binary
variable where 1 ¼ female, with male as the comparison group. Marital status is
coded into binary variables for single ¼ 1 and cohabiting ¼ 1, with married as the
comparison group. Age is coded as a continuous variable representing years of age.
Analyses
First, I examine bivariate correlations between key theoretical variables. Second, to

test the relative strength of the associations between general and specific reflected
appraisals and self-views in each respective identity, I estimate ordinary least squares
(OLS) linear regression models with robust standard errors.4 Because each respon-
dent answered questions about two identities, self-views in each identity are likely
not independent from one another. Robust standard errors are estimated to correct for
this potential autocorrelation. I also conduct postestimation Wald tests to compare
the effect sizes of the coefficients for general and specific reflected appraisals in each
identity.
4
The dependent variables are ordinal because their categories can be ranked (e.g., a self-view rating
of 8 is higher than a self-view rating of 6), but the precise distance between them is unknown. Linear
regression models assume that the distance between categories are equal, but ordinal regression
models do not require that assumption to be met. Therefore, I also estimated the models using
ordered probit regression (in analyses not shown, but available upon request) and found that the
results were substantively very similar to the results from the linear models. For ease of interpre-
tation, I present the results of the linear models here.
234 M. Gallagher
Results
The correlations between self-views, general reflected appraisals, and specific

reflected appraisals for each identity are shown in Table 11.2. As expected, self-
views and reflected appraisals from general and specific sources in each respective
identity are significantly associated ( p < 0.001). The correlation coefficients are all
on the lower end of what can be considered a large effect size (average r ¼ 0.68),
which suggests that, although self-views and reflected appraisals are strongly asso-
ciated with one another, they are not so highly correlated as to raise questions about
whether they are measuring different things. Associations (not shown) between self-
views and each source of reflected appraisals for the two identities on which
respondents reported were statistically significant, but modest in size (r ¼ 0.29 for
self-views, r ¼ 0.21 for specific others, and r ¼ 0.27 for general others). This
suggests that respondents were differentiating between their identities. In other
words, there does not seem to be a response set effect operating, since respondents
generally did not report the same self-views or reflected appraisals for the two
different identities about which they were questioned.
Table 11.3 displays associations between reflected appraisals and self-views in
the spouse (Model 1), friend (Model 2), worker (Model 3) and religious (Model 4)
identities. In preliminary analyses (not shown), I estimated models in which reflected
appraisals from different sources were entered separately and models in which they
were entered together. Because both approaches produced substantively identical
results, only the models in which they are included together are presented here.
Variance inflation factors (VIFs) for all models are below 2.00, indicating that
multicollinearity is not an issue in the regression analyses (O’Brien, 2007).
As shown in Table 11.3, general and specific reflected appraisals are significantly
and positively associated with self-views in three of the four identities (spouse,
worker, and religious), supporting the basic tenet of symbolic interactionism that
reflected appraisals influence self-views. In general, the more positively respondents
perceive that general and specific others rate them in their identities, the more
positively they view themselves in those same identities. The only exception to
Table 11.2 Correlations between self-views and reflected appraisals

Spouse identity Friend identity Worker identity Religious identity
(N ¼ 184) (N ¼ 184) (N ¼ 136) (N ¼ 172)
(1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3)
(1) Self- 1.00 1.00 1.00 1.00
view
(2) General 0.64 1.00 0.82 1.00 0.66 1.00 0.72 1.00
Reflected
appraisal
(3) Specific 0.62 0.54 1.00 0.64 0.68 1.00 0.71 0.75 1.00 0.70 0.75 1.00
Reflected
appraisal
All correlations are statistically significant ( p 0.001)
Table 11.3 Unstandardized coefficients and robust standard errors for linear regressions of self-
views on general and specific reflected appraisals
Spouse identity Friend identity Worker identity Religious identity
(N ¼ 184) (N ¼ 184) (N ¼ 136) (N ¼ 172)
Model 1 Model 2 Model 3 Model 4
b (SE) b (SE) b (SE) b (SE)
General reflected 0.26** 0.89*** 0.32* 0.27**
appraisal
(0.09) (0.06) (0.13) (0.09)
Specific reflected 0.45*** 0.06 0.39** 0.35***
appraisal
(0.11) (0.09) (0.13) (0.08)
Female 0.24 0.04 0.20 0.35
(0.20) (0.14) (0.16) (0.19)
Age 0.01 0.02** 0.00 0.01*
(0.01) (0.01) (0.01) (0.00)
Single 0.07 0.48** 0.02 0.08
(0.29) (0.16) (0.14) (0.18)
Cohabiting 0.22 0.03 1.42* 0.18
(0.84) (0.11) (0.66) (0.29)
Employed 0.01 0.32* – 0.09
(0.21) (0.15) (0.17)
Education 0.03 0.02 0.01 0.01
(0.04) (0.02) (0.01) (0.03)
Black 0.02 0.18 0.30 0.13
(0.36) (0.24) (0.31) (0.23)
Latino 1.08*** 0.02 0.75 0.47
(0.19) (0.56) (0.78) (0.29)
Other race 0.55 0.34 0.20 0.56
(0.49) (0.29) (0.41) (0.39)
Intercept 1.66 0.50 2.56 2.62
R-Squared 0.46 0.80 0.50 0.62
*p 0.05; **p 0.01; ***p 0.001
this pattern is the friend identity, where general reflected appraisals are significantly
associated with self-views (b ¼ 0.89, p < 0.001), but specific reflected appraisals
are not.
Wald tests comparing the coefficients for specific and general reflected appraisals
in the spouse, worker, and religious identities are all nonsignificant ( p > 0.05),
indicating that the null hypothesis that the parameters are equal cannot be rejected. In
other words, both sources of reflected appraisal contribute to self-views in those
three identities, but not to different degrees. In the friend identity, reflected appraisals
from general others are significantly associated with self-views but reflected
appraisals from a specific other (closest friend) are not, and a Wald test confirms
236 M. Gallagher
that the coefficients for general and specific reflected appraisals in the friend identity
are significantly different ( p < 0.001).
Taken together, this study’s findings demonstrate that reflected appraisals from
general others are associated with self-views in all identities examined here, and that
reflected appraisals from specific others are associated with self-views in all but the
friend identity. For the spouse, worker, and religious identities, neither source of
reflected appraisal is more strongly associated with self-views than the other. This
provides further evidence that reflected appraisals from both general and specific
others independently affect self-views in those three identities. In the friend identity,
a different pattern is observed, where general reflected appraisals are more strongly
associated with self-views than specific reflected appraisals.
Discussion
The pattern of findings in this study lend support for the basic symbolic interactionist
premise that reflected appraisals are associated with self-views in several common
adult social roles, some of which have not been examined in previous research on the
reflected appraisal process. The more positively respondents felt that specific others
(role partners or members of their groups) and general others (“others”) evaluated
them in their spouse, worker, and religious identities, the more positively they rated
themselves in those identities. This suggests that the reflected appraisal process is
quite robust, in that it operates in the same way across different identities. It also
indicates that both general and specific sources of reflected appraisal independently
contribute to the development and maintenance of self-views. However, in the friend
identity, reflected appraisals from general others were significantly associated with
self-views, but reflected appraisals from specific others (the respondent’s closest
friend) were not.
It is possible that this one departure from the general pattern of findings could be
due to chance, but it might also be the case that there is something unique about the
friend identity when it comes to the influence of different sources of reflected
appraisals. For most of the identities in this study, the associations between general
and specific reflected appraisals and self-views were statistically equal. Because
reflected appraisals are an important source of feedback in identity theory, these
findings also have implications for the perceptual control process. Discrepancies
involving different sources of reflected appraisals may imply different self-relevant
meanings for different bases of identity, and thus, have different consequences for
outcomes such as self-esteem.
Identity theorists have long recognized that identities are connected to outcomes
through shared meanings (Burke & Reitzes, 1981). Role identities are based on
positions held in social structure and emphasize role performance, group identities
are based on membership in groups and emphasize belonging, and person identities
are based on personal characteristics and emphasize the ways in which one is a
unique individual (Burke & Stets, 2009). Stets and Burke (2014) have demonstrated
that discrepancies in role, group, and person identities are most strongly associated
with different bases of self-esteem (self-efficacy, self-worth, and authenticity,
respectively).
In addition to their relevance to distinct outcomes, the differential meanings of
role, group, and person identities implicate different sources of reflected appraisal
that might be particularly influential in identity verification processes for each
respective base of identity. Specifically, the focus on role performance in role
identities suggests that discrepancies with specific role partners who provide feed-
back about the extent to which one adequately fulfills role-related responsibilities
may be most consequential for mastery. The focus on belonging in group identities
suggests that discrepancies with specific others who are reference groups that
provide feedback about the extent to which one adequately contributes to and
represents the group’s goals and ideals may be most important for self-esteem.
The focus on unique, individual characteristics in person identities suggests that
discrepancies with specific significant others who provide feedback about the extent
to which one possesses certain personal attributes may be most strongly associated
with authenticity.
Limitations and Directions for Future Research
This study has limitations that could be addressed in future research. First, although
the self-view and reflected appraisal measures utilized here are consistent with some
previous identity research (Stets & Harrod, 2004; Stets et al., 2017) and are relevant
to a wide range of identities, they do not account for the multiple dimensions of
meaning that comprise identity standards (Burke & Tully, 1977). Instead, they
capture just one dimension of meaning (evaluation). The more typical way of
measuring identity standard meanings in identity theory involves asking respondents
to rate themselves on a series of semantic differentials that are anchored by pairs of
adjectives representing the positive and negative poles of an attribute related to an
identity. Reflected appraisals are then measured by asking respondents how others
would rate them on those same dimensions of meaning (Stets & Trettevik, 2014).
The relative influence of general and specific sources of feedback in the reflected
appraisal process should be reexamined in research using this measurement tech-
nique to verify the pattern of findings presented here.
Second, who respondents were thinking of when they responded to questions
about how “others” would rate them in their identities is unknown. When a person is
asked to think about how “others” or “others in general” view them in an identity, it
could call to mind a number of different things, including their general sense of how
they are evaluated with respect to social and cultural norms and expectations tied to
that identity, how they are evaluated by specific groups or individuals, a composite
perception of evaluations from multiple role complements or multiple significant
others who are not directly tied to the role, or some combination of the above.
238 M. Gallagher
While the questions about reflected appraisals in this study did reference specific
identities, respondents were not promoted to think of anything in particular when it
came to the term “others,” nor were they questioned about what exactly they had in
mind when they thought about others’ views of them in their identities. Notably,
similar measures are quite common in identity research (e.g., Marcussen & Large,
2003; Marcussen & Gallagher, 2017; Stets & Trettevik, 2016; Trettevik, 2016),
although their limitations are rarely acknowledged (see Kalkhoff et al., 2016 for an
exception). Moving forward, identity research might benefit from increased theoret-
ical and empirical attention to the different facets of general reflected appraisals.
Identity theorists assume that general others represent some unspecified combi-
nation of specific others who are role complements for a given identity and/or
generally important sources of feedback about all or most identities. Other social
psychological theories involving general others identify additional facets of the
concept. For example, in expectations states theory, third-order expectations refer
to beliefs about what “most people” think about or expect from status groups such as
those based on race and class distinctions (Ridgeway & Correll, 2006). There may
also be third-order beliefs representing most people’s expectations for those who
hold a given identity. Similarly, in affect control theory, fundamental sentiments
refer to culturally shared meanings attached to actors, behaviors, objects, and social
settings (Robinson et al., 2006). For any identity that is activated and recognized by
others during social interaction, there may also be culturally shared meanings
concerning the characteristics, behaviors, and attitudes of a person who holds that
identity.
Methodologically, a couple of strategies could be employed to measure these
different facets of general reflected appraisals. The simplest approach would be to
preface any question(s) about them with a prompt in which “others” or “others in
general” (or any other term one may use for the concept) is defined for the
respondent. For example, if one is interested in measuring general reflected
appraisals in terms of culturally shared meanings related to an identity, then respon-
dents might be instructed that “others” refers to society’s standards and expectations
for people who hold the identity in question. This would help increase confidence
that respondents are interpreting relatively vague terms like “others” or “others in
general” in ways that are consistent with their theoretical meaning.
Alternatively, one might utilize a measurement approach that seeks to understand
the meaning of general sources of reflected appraisals from the perspective of
respondents to determine the extent to which they are interpreting questions about
general others in ways that are consistent with researchers’ theoretical assumptions.
For example, interpretive, open-ended questions could be included along with the
more typical close-ended questions to help elucidate how respondents view and
make distinctions between reflected appraisals from role-partners, groups, and
society. Similarly, questions regarding who or what exactly a respondent was
thinking about when they answered queries about others or others in general could
be asked. This information could expand our understanding of the extent to which
individuals perceive unique or overlapping reflected appraisals from different
sources and clarify whether reflected appraisals from general others are distinct
from or merely a proxy for reflected appraisals from one or more specific role
partners or groups.
Conclusions
Despite the noted limitations, this study advances our understanding of the ways in
which different sources of reflected appraisals operate in three ways. First, it pro-
vides strong and consistent support for the basic tenet of the reflected appraisals
process that perceptions of how others behave toward and respond to a person are
associated with that person’s own self-views (Kinch, 1963) in several conventional
adult social roles, including some that have not been examined in previous research
on the reflected appraisal process.
Second, findings presented here also suggest that certain sources of reflected
appraisal may not be equally relevant for self-views in all identities. By extension,
identity discrepancies involving reflected appraisals from different sources of feed-
back may affect different outcomes, or the same outcomes to different degrees.
Established theoretical distinctions between identities may provide a useful frame-
work for examining these associations. To that end, I have suggested that the
different meanings associated with role, group, and person identities implicate
different sources of feedback, and that discrepancies involving those sources may
be differentially associated with self-efficacy, self-worth, and authenticity,
respectively.
Third, this study outlines several directions for future research to refine the
measurement of reflected appraisals from general others. Measures of reflected
appraisals from general others are well-represented in survey research on identities,
but it is often not clear which facet(s) of the generalized other they might be
capturing or whether respondents are interpreting them in ways that are consistent
with their theoretical meaning. Studies that incorporate measures of reflected
appraisals that capture the generalized other and its full range of possible meanings
could advance our understanding of the independent and cumulative effects of
different sources of feedback in reflected appraisal and perceptual control processes.
Acknowledgement Data collection for this research was funded by a National Science Foundation
Dissertation Improvement Grant (SES-1129879) awarded to Kristen Marcussen (PI) and Mary
Gallagher (co-PI).
References
Asencio, E. K. (2011). Familiarity, legitimation, and frequency: The influence of others on the
criminal self-view. Sociological Inquiry, 81(1), 34–52.
Asencio, E. K. (2013). Self-esteem, reflected appraisals, and self-views: Examining criminal and
worker identities. Social Psychology Quarterly, 76(4), 291–313.
240 M. Gallagher
Asencio, E. K., & Burke, P. J. (2011). Does incarceration change the criminal identity? A synthesis
of labeling and identity theory perspectives on identity change. Sociological Perspectives, 54
(2), 163–182.
Bouchey, H. A., & Harter, S. (2005). Reflected appraisals, academic self-perceptions, and math/
science performance during early adolescence. Journal of Educational Psychology, 97(4),
673–686.
836–849.
Burke, P. J. (1996). Social identities and psychological stress. In H. B. Kaplan (Ed.), Psychosocial
stress: Perspectives on structure, theory, life course, and methods (pp. 141–174). Academic
Press.
Burke, P. J., & Reitzes, D. C. (1981). The link between identity and role performance. Social
Burke, P. J., & Tully, J. C. (1977). The measurement of role identity. Social Forces, 55(4),
881–897.
Cast, A. D., & Cantwell, A. M. (2007). Identity change in newly married couples: Effects of
positive and negative feedback. Social Psychology Quarterly, 70(2), 172–185.
Collett, J. L., Vercel, K., & Pierce, K. D. R. (2019). The role of significant others in shaping fathers’
identities and behavior. In J. E. Stets & R. T. Serpe (Eds.), Identities in everyday life
(pp. 359–380). Oxford University Press.
Cooley, C. H. (1902). Human nature and the social order. Scribner.
Felson, R. B. (1981). Self- and reflected appraisal among football players: A test of the median
hypothesis. Social Psychology Quarterly, 44(2), 116–126.
Felson, R. B. (1985). Reflected appraisals and the development of self. Social Psychology Quar-
terly, 48(1), 71–78.
Felson, R. B. (1989). Parents and the reflected appraisal process: A longitudinal analysis. Journal of
Gallagher, M. (2017). Obligatory and voluntary identity discrepancies, self-evaluation, and psy-
chological distress. Society and Mental Health, 7(2), 51–68.
Ichiyama, M. A. (1993). The reflected appraisal process in small-group interaction. Social Psychol-
ogy Quarterly, 56, 87–99.
James, W. (1890). The principles of psychology. Holt.
Kalkhoff, W., Marcussen, K., & Serpe, R. T. (2016). To thine own self be true? Clarifying the
effects of identity discrepancies on psychological distress and emotions. Social Science
Research, 58, 14–33.
Kemper, T. D. (1966). Self-conceptions and the expectations of significant others. The Sociological
Quarterly, 7(3), 323–343.
Khanna, N. (2004). The role of reflected appraisals in racial identity: The case of multiracial asians.
Kinch, J. W. (1963). A formalized theory of the self-concept. American Journal of Sociology, 68(4),
481–486.
Kuhn, M. H. (1964a). Major trends in symbolic interaction theory in the past twenty-five years. The
Sociological Quarterly, 5(1), 62–84.
Kuhn, M. H. (1964b). The reference group reconsidered. The Sociological Quarterly, 5(1), 5–21.
Marcussen, K., & Asencio, E. K. (2016). Stigma resistance and the mental illness self-view. In J. E.
Stets & R. T. Serpe (Eds.), Advances in identity theory and research (pp. 473–508).
Marcussen, K., & Gallagher, M. (2017). The role of aspirations and obligations in explaining the
relationship between identity discrepancies and psychological distress. Sociological Perspec-
tives, 60(6), 1019–1038.
Marcussen, K., & Large, M. D. (2003). Using identity discrepancy theory to predict psychological
distress. In P. J. Burke, T. J. Owens, R. T. Serpe, & P. A. Thoits (Eds.), Advances in identity
theory and research (pp. 151–166). Kluwer Academic/Plenum.
Markowitz, F. E., Angell, B., & Greenberg, J. S. (2011). Stigma, reflected appraisals, and recovery
outcomes in mental illness. Social Psychology Quarterly, 74(2), 144–165.
Maurer, T. W., Pleck, J. H., & Rane, T. R. (2001). Parental identity and reflected-appraisals:
Measurement and gender dynamics. Journal of Marriage and Family, 63(2), 309–321.
Mead, G. H. (1934). Mind, self and society. The University of Chicago Press.
149–172.
O’Brien, R. M. (2007). A caution regarding rules of thumb for variance inflation factors. Quality
and Quantity, 41(5), 673–690.
Ridgeway, C. L., & Correll, S. J. (2006). Consensus and the creation of status beliefs. Social
Forces, 85(1), 431–453.
Robinson, D. T., Smith-Lovin, L., & Wisecup, A. K. (2006). Affect control theory. In J. E. Stets &
J. H. Turner (Eds.), Handbook of the sociology of emotions (pp. 179–202). Springer.
Shibutani, T. (1955). Reference groups as perspectives. American Journal of Sociology, 60(6),
562–569.
Shrauger, J. S., & Schoeneman, T. J. (1979). Symbolic interactionist view of self-concept: Through
the looking glass darkly. Psychological Bulletin, 86(3), 549–573.
Stets, J. E., & Asencio, E. K. (2008). Consistency and enhancement processes in understanding
emotions. Social Forces, 86(3), 1055–1078.
Stets, J. E., Brenner, P. S., Burke, P. J., & Serpe, R. T. (2017). The science student identity. Social
Science Research, 64(4), 1–14.
Stets, J. E., & Burke, P. J. (2014). Self-esteem and identities. Sociological Perspectives, 57(4),
409–433.
symbolic interaction (pp. 65–88). Springer.
Stets, J. E., & Trettevik, R. (2016). Happiness and identities. Social Science Research, 58, 1–13.
Stets, J. E., & Trettevik, R. (2014). Emotions in identity theory. In J. E. Stets & J. H. Turner (Eds.),
Handbook of the sociology of emotions (Vol. II, pp. 33–49). Springer.
Stryker, S., & Statham, A. (1985). Symbolic interaction and role theory. In G. Lindzey &
E. Aronson (Eds.), Handbook of social psychology (pp. 311–378). Random House.
Sullivan, H. S. (1953). In H. S. Perry & M. L. Gladwell (Eds.), The interpersonal theory of
psychiatry. Norton.
The American Association for Public Opinion Research. (2009). Standard definitions: Final
dispositions of case codes and outcome rates for surveys (6th ed.). AAPOR.
Thoits, P. A. (2003). Personal agency in the accumulation of multiple role-identities. In P. J. Burke,
(pp. 179–194). Kluwer Academic/Plenum.
Trettevik, R. (2016). Identities, goals, and emotions. Social Psychology Quarterly, 79(3), 263–283.
Turner, R. H. (1956). Role-taking, role standpoint, and reference-group behavior. American
Journal of Sociology, 61(4), 316–328.
Wallace, H. M., & Tice, D. M. (2012). Reflected appraisal through a 21st-century looking glass. In
M. R. Leary & J. P. Tangney (Eds.), Handbook of self and identity (pp. 124–140). Guilford
Press.
Part III
Networks, Homophily, and the Physical
Environment
Chapter 12
Does the Self Reflect Society? Examining
the Relationship Between Personal Network
Assortativity and Role-Based Groups
Mark H. Walker
Identity theorists (Stryker, 1980; Stryker & Serpe, 1982) argue that individuals live
their lives in relatively small and specialized role-based groups, where social behav-
ior is based on the roles that are ascribed to individuals and the identities that are
assumed to be attached to those roles. Although theoretical work in identity theory
and structural symbolic interactionism accept the notion that there is likely substan-
tial overlap between these role-based groups, empirical work has generally exam-
ined identities and role-based groups in isolation (Burke & Reitzes, 1981; Riley &
Burke, 1995; Stryker et al., 2005; Stryker & Serpe, 1982). Although social networks
provide a crucial link between the self and society for identity theory (Stryker, 1980;
Stryker et al., 2005), relatively little is known about the interplay between social
network processes and identity processes. Research in identity theory has not
adequately addressed the question of how the patterning of ties between the people
in an individual’s personal social network matter for identity processes.1 Recent
research applying social network methods to identity processes highlight the impor-
tance of the structure of social relations within and between role-based groups,
which is a collection of individuals with whom interaction is based on a given
role-identity. Walker and Lynn (2013) find that social ties between members of an
individual’s role-based groups and social contacts outside of a given role-based
group increase identity salience because these ties effectively extend identity mean-
ings outside a given role-domain. Other research suggests that the density of social
ties within a given group constrains identity change (McFarland & Pals, 2005). Still
other research highlights the impact of tie multiplexity or social contacts that are
based on multiple roles on emotional well-being (Markowski, 2019).
1
See McFarland and Pals (2005), Walker (2015), Markowski (2019), and Walker and Lynn (2013)
for notable exceptions.
M. H. Walker (*)
Louisiana State University, Baton Rouge, LA, USA
246 M. H. Walker
Though this research points to the importance of the structure of social relations
within and between role-based groups for identity processes, little is known about
the general patterns in the structure of social relations in these groups. The extent to
which social networks are fragmented around role-based groups, the general demo-
graphic patterns in the structure of social relations within and between role-based
groups, and the comparative levels of social closure of role-based groups are unclear.
A better understanding of the extent and nature of social overlap between role-based
groups could inform the assumptions of structural identity theory as well as the
design of future studies linking social networks to identity processes.
Using personal social network data drawn from a convenience sample of students
at a large Midwestern university, I address three issues regarding the structure of
social relations within and between role-based groups. First, I examine the extent to
which social ties in individuals’ personal networks cluster around role domains such
as church, school, and work. I assess the extent to which an individual’s social
contacts who are members of a given role-based group are more likely to be socially
tied to one another than contacts outside of that role-based group, using the network
measure of assortativity. To provide a baseline, I compare role-based assortativity to
assortativity based on gender, race, and tie strength. Second, I compare average
levels of assortativity of five role-based groups: family members, coworkers, church
contacts, schoolmates, and friends. Third, I examine the extent to which the demo-
graphic characteristics and level of role-involvement of respondents are associated
with assortativity.
The aim of this study is to provide a descriptive account of the patterns in social
relations within and between role-based groups. Although the sample and population
of study utilized here are not ideal for estimating population parameters, I am
unaware of any nationally representative study with network data that gathers
extensive network data that measures the structure of social relations between
respondents’ social contacts and includes information on social contacts’ role-
memberships. Thus, this research should not be taken as representative of population
parameters, but as a first attempt at gaining insight into the overall landscape of the
structure of social relations within and between role-based groups. These findings
will be of use for scholars who are interested in examining the “upstream” processes
within structural identity theory concerned with the interplay between large-scale
social structures, social networks and commitment, and identity salience.
Theory
Social Networks and the Self
According to identity theory (Stryker, 1980), social networks serve as the link
between the self and social structure. Although early formulations of identity theory
included discussion of various network measures, researchers have argued that
social networks impact the self by enhancing one’s social and emotional
12 Does the Self Reflect Society? Examining the Relationship Between Personal. . . 247
commitment to a given role-based group; social ties that are based on a given role-
identity act as social and emotional investments in the group (Walker & Lynn,
2013). Essentially, social ties based on an identity are commitment. Increased social
and emotional investment in the group, in turn, is thought to increase the salience of
the identity associated with the group, with salience defined as the likelihood that the
identity will be activated in and across social encounters. Finally, the more salient an
identity, the more likely the identity is to shape behavior. Thus, in the traditional
formulation, social networks are important in identity theory because social ties
constitute commitment to role-based groups, which shapes the self and behavior.
Because of the social and emotional investment view of commitment, social
networks are operationalized within identity theory as the extensiveness and inten-
siveness of ties to a role-based group (Stryker, 1968).2 The extensiveness of ties is the
number of ties that one has to members of a given role-based group. This reflects the
notion that individuals who have a large number of ties based upon a particular role-
identity will feel socially invested in the identity, as well as a great sense of loss if
they were no longer able to perform that identity, and therefore forgo the ties with
these role-based others. The intensiveness of ties reflects the strength, or emotional
closeness, of the ties to members of a given role-based group. This aspect of
commitment reflects the fact that not all social ties are equal, and individuals are
more easily influenced by strong ties and feel a greater sense of loss at the dissolution
of strong ties than the dissolution of ties with mere acquaintances based on an
identity.
In the language of social network analysis, identity theory’s link between com-
mitment and salience is rooted in a bonding/binding mechanism (Borgatti & Halgin,
2011; Borgatti & Lopez-Kidwell, 2011). The bonding/binding mechanism involves
social ties serving as bonds to other social actors and social groups, while simulta-
neously binding individuals to the norms, expectations, values, and rules (both
formal and informal) of that group. However, recent research suggests that other
network mechanisms aside from the bonding/binding mechanism might matter more
for identity internalization.
Building on Stryker’s (1968, 1980) early theoretical statement regarding the link
between social networks and identity salience, Walker and Lynn (2013) argued that
the structure of social relations within and between role-based groups matter for
identity salience. Indeed, their findings revealed that the embeddedness of role-based
others in one’s overall social network affected identity salience more than either the
number of ties to role-based others or the strength of those ties for work, religion, and
2
Commitment also has been divided in terms of affective and interactional commitment. The
extensiveness and intensiveness of ties to role-based others are key indicators of interactional and
affective commitment, respectively, but there also are other indicators that are not discussed here.
For example, frequency of interaction with role-based others is often used as an indicator of
interactional commitment. I focus here on the extensiveness and intensiveness of ties because
these concepts align most clearly with network measures and concepts.
248 M. H. Walker
student role-identities.3 These findings have two implications: (1) the structure of
social relations between social contacts in individuals’ personal social network is
important for identity processes, and (2) ties between role-based groups as well as
ties extending from role-based others to non-role-based others matter for the self.
Thus, it may be important to understand the extent to which social networks are
characterized by segmented social networks clustered around role-based groups, or
ties that crosscut role domains.
Given the importance of intergroup ties and network structure for identity pro-
cesses, research comparing the extent of clustering across role-based groups associ-
ated with different identities as well as research illustrating demographic patterns in
the structure of role-based groups may help refine predictions within identity theory
regarding the interplay between social networks and the self. For example, though
Walker and Lynn (2013) focus on how intergroup ties matter for identity salience,
research on social capital suggest that dense ties within a given group should increase
trust, reciprocity, cohesion, and shared norms among group members (Coleman,
1988; Granovetter, 1992). In turn, these factors should increase identity salience and
group commitment. Taken together with the fact that previous research suggests that
network density and peer homogeneity reduce the likelihood of identity change
(McFarland & Pals, 2005), it is surprising that the density of ties between role-based
others was not associated with identity salience.
One potential explanation for the inconsistent findings regarding density is that
the network mechanisms that drive the internalization of an identity may depend on a
group’s stage of development. Long standing groups where group membership is
relatively static, such as a family, likely do not display much in terms of intragroup
network dynamics. Thus, within-group density is unlikely to “move the needle” on
identity salience. Instead, for this type of group, commitment may be indicated by
the extent to which members are “brought in” to one’s other social circles or via
intergroup ties. Thus, on the one hand, intragroup density may be more important for
commitment and salience for groups that are at early stages of development and
therefore more amenable to within-group dynamics. On the other hand, intergroup
ties may be more strongly linked to group commitment and identity salience for
long-standing groups that are more resistant to within-group network dynamics.
Although I cannot directly address these two possibilities here, a descriptive account
of comparative levels of assortativity among role-based groups provides background
information that can inform the plausibility of this and other accounts of the
upstream processes involved in linkages between large-scale aspects of social
structure, commitment, and identity salience.
3
Walker and Lynn (2013) define the embeddedness of role-based others as the proportion of non-
role-based others who are tied to a member of a given role-based group based on an identity. For
example, the embeddedness of role-based others for the religious role identity reflects the extent to
which people in one’s personal network who do not attend religious services with them (i.e., non-
role-based others) are tied to at least one person who does attend religious services with them. Thus,
embeddedness in this context reflects the extent to which members of a role-based group have a
breadth of ties to other members of one’s personal network.
Homophily in Social Networks
A large area of study in social network analysis concerns itself with the extent to
which networks are characterized by homophily, which is the tendency of individ-
uals to associate with others who are socially similar to themselves (McPherson
et al., 2001). Generally, studies of homophily focus on similarity in terms of
demographic characteristics (Marsden, 1987, 1988; Yamaguchi, 1990) or political
opinions and social attitudes (Huckfeldt & Sprague, 1995; Knoke, 1990). However,
researchers have begun to examine the idea of “focal closure,” (Kossinets & Watts,
2006, 2009) where ties are formed based on shared “social interaction foci” (Feld,
1981). Social interaction foci are “the various groups, contexts, and activities around
which social life is organized and which in turn facilitate interpersonal interactions”
(Kossinets & Watts, 2009, p. 418). Focal closure describes the extent to which social
ties are formed based on shared membership in a group or shared participation in an
organization. Social roles, such as student, religious congregant, and employee
constitute “social interaction foci” that facilitate interpersonal interactions.
Some empirical research has examined the impact of focal closure on the struc-
ture of large-scale networks that focus on a particular domain of social life such as
email exchanges amongst students at a large university (Kossinets & Watts, 2006).
However, I am unaware of any study that examines the role that focal closure plays
in structuring personal social networks containing social contacts from multiple
social domains. Social network data that elicits social contacts from multiple
domains of social life is particularly important for identity theory, since the self is
viewed as being composed of multiple, interrelated identities that are linked to
various organizations and institutions.
While the research on focal closure in large-scale social networks can inform us
about how this network process matters for a given social system, it provides little
information regarding how focal closure impacts the immediate social environment
surrounding individuals. Research on sociocentric networks that examines the
structure of social ties between all members of a particular social system such as
friendship networks in a high school or advice networks at a workplace can uncover
how focal closure processes unfold in a particular setting such as schools or
workplaces. However, it is unable to examine how the multiple, interrelated social
domains that constitute focal individuals’ social environment can impact one another
and even combine to impact individuals. Additionally, from an identity theory
perspective, network processes that occur in one’s immediate social environment
(e.g., the pattern of ties between one’s friends) are more important for the self than
more distal network processes (e.g., the pattern of ties between friends of friends).
To address this, I utilize egocentric network data, which samples individuals (ego)
and elicits the focal individual’s social contacts (alters) as well as the structure of ties
between social contacts (alter-alter ties). I examine the extent to which the social
contacts in individuals’ personal social networks display homophily based on
co-membership in role-based groups.
250 M. H. Walker
In summary, this research contributes to identity theory by providing a first look

at the overall landscape of social ties within and between role-based groups. By
providing a descriptive account of the overall level of assortativity of various role-
based groups, I assess the extent to which social networks reflect identity theory’s
view of individuals living their lives in relatively small and separate pockets of role-
based groups. I also highlight differences among role-based groups in terms of their
assortativity. Finally, I identify demographic patterns in assortativity that may have
important implications for the link between large-scale social structures and com-
mitment to role-based groups. This research also contributes to the social networks
literature by directly examining the impact that focal closure and institutional
boundaries have on tie formation within (and between) role-based groups.
Methods
To assess the extent to which role-based groups based on identities structure the
patterning of ties within individuals’ personal social networks, data that elicits the
content (including demographic and role-based characteristics of social contacts)
and structure of individuals’ personal social networks is needed. Further, since role-
based groups tend to consist of both emotionally close as well as weak social ties,
extensive personal social networks consisting of both strong and weak ties is
necessary. Since data of this kind are not publicly available, I use an original data
set consisting of extensive personal social networks of a convenience sample of
students from a large Midwestern University (N ¼ 431). I study the patterning of ties
between respondents’ social contacts, as well as information regarding the charac-
teristics of the social contacts that comprise respondents’ personal networks. Com-
pared with the University’s overall demographic composition, this sample is
disproportionately white and female.
Data
Eliciting personal social networks. To elicit the structure and content of personal
social networks, I use a name generator designed to gather both strong ties as well as
social contacts with whom individuals may not be emotionally close, but who they
nonetheless interact with on a regular basis. The personal social networks gathered in
this study were elicited using a novel vignette-style name generator:
Imagine that you’ve just won an award that you are very excited about. Those presenting the
award to you want to throw a party in your honor. They will pay for all of the expenses
related to the party, including food, drinks, travel and housing for guests.
Think carefully about all the people you would want to celebrate with. Who would you
invite?
Respondents were instructed to create a guest list consisting of up to 25 names.

Since previous research indicated that the guest list tended to generate strong ties
(Walker, 2015; Walker & Lynn, 2013), additional role-based probes were used to
generate social contacts with whom the respondents were less close, but who were
nonetheless important fixtures in their daily lives. Respondents were probed about
members of the various role-based groups they participate in including school,
religion, family, friend, and work by asking them to list role-based contacts that
they interact with on a regular basis (regardless of how close they are to them).
Respondents were allowed to list up to five additional names for each role-based
probe, resulting in a maximum possible network size of 50 (25 + [5 5] ¼ 50).
Generating role-based subgroups. To partition egocentric networks into role-
based groups, I use role-based name interpreters. Respondents are shown their list of
contacts (including the guest list and names generated from role-based probes) and
are to indicate which of the people on the list belongs to a given role-based group.
For example, to generate the religion-based group, respondents indicate which of the
people on their list of contacts “attend the same religious services as they do.” To
generate the student-based group, respondents indicate which of the contacts are
“students at Midwestern University;” the friend-based group is generated by respon-
dents indicating who on their list they “enjoy spending time with, socially;” the
family-based group is generated by respondents reporting “which of the people on
your list are family members;” and the work-based group is generated by asking
respondents “which of the people on your list are co-workers.” To create groups
based on race and gender, respondents report which of the people on their list are “a
different race than you” and “female,” respectively.
Eliciting alter-alter ties. To assess the level of homophily based on role-based
group membership and demographic characteristics in individuals’ personal social
networks, I identify the structure of social relations within personal networks or who
is socially tied to whom. Alter-alter ties are measured by presenting respondents with
all possible alter-alter pairs and asking “which of these pairs know each other well
enough to stop and chat if they passed one another on the street?” This level of
“knowing” is similar to that used in the 2006 General Social Survey “number
known” module (Diprete et al., 2011).
Measures
Dependent Variable—Assortativity. To assess the extent to which role-based

groups, demographic characteristics, and network processes structure the pattern of
ties within individuals’ personal social networks, I use the metric of assortativity
(Newman, 2002, 2003). Assortativity assesses the level of homophily or the extent to
which alters who share a given characteristic are more likely to be tied to one another
than those who do not share the characteristic within a given network. Although
there are other methods for assessing the impact of homophily on tie formation
within social networks, assortativity is ideally suited for the purposes of this study
252 M. H. Walker
because it provides a common metric which can be compared across multiple social
networks. Similar to the commonly used correlation coefficient, Pearson’s r,
assortativity yields a number between 1 and 1, where a value of 1 indicates a
perfect correspondence between sharing a nominal characteristic and tie formation
and a value of 1 indicates a perfect dissortativity, such that every tie connects two
nodes of different types. An assortativity score of 0 indicates no relationship between
nominal characteristics and tie formation or a random network.
I calculate assortativity based upon five role-based groups (friend, work, family,
student, religion), two demographic characteristics (race and gender), and tie
strength between ego and alter. Thus, a positive religion assortativity coefficient
indicates that ties between alters who attend religious services with ego (and ties
between alters who do not attend religious services with ego) are more prevalent than
ties between religion-based alters and non-religion-based alters. Conversely, a
negative religion assortativity coefficient would indicate that ties between religion-
based alters and non-religion-based alters are more prevalent than ties within the
religion-based group and/or ties amongst non-religion-based alters.
Another important aim of this study is to examine the extent to which assortativity
differs based on the individual characteristics of respondents. This study examines
variability in assortativity based on individual characteristics: (1) demographic
characteristics, (2) network size, and (3) role-involvement.
Demographic characteristics of respondent (ego). I examine variability of
assortativity based on six demographic factors: gender (1 ¼ female, 0 ¼ male),
race (white ¼ 1, non-white ¼ 0), marital status (1 ¼ married, 0 ¼ unmarried),
parental income (in thousands of dollars), age (in years), whether this is the respon-
dent’s first semester at Midwestern University (1 ¼ yes, 0 ¼ no), and a series
of dummy variables indicating student status at Midwestern University
(freshman ¼ baseline).
Role-based behavior. An important individual-level factor that might impact the
level of role-based assortativity and race- and gender-based assortativity is one’s
overall involvement in various role-based groups. I operationalize role-involvement
in terms of the amount of time one spends in a role-domain. Involvement in the
student role is measured by how many hours per week they spend on school work
outside of class; involvement in the family role is measured by how many hours per
week they spend time socializing with family; and involvement in the work role is
measured by how many hours per week they work for pay. Responses that they
could select for the amount of hours they spent each week in each role domain
included “less than 5 h,” “5–10 h,” “10–15 h,” “15–20 h,” “20–30 h,” “30–40 h,”
and “more than 40 h.” Responses were recoded to reflect the center of the range such
as “less than 5 h”¼2.5, “5–10 h”¼7.5 for interpretability. Finally, involvement in the
religious role is measured in terms of the frequency of religious service attendance.
Responses and their corresponding codes included 1 ¼ “never,” 2 ¼ “less than once
a year,” 3 ¼ “about once or twice a year,” 4 ¼ “several times a year,” 5 ¼ “about
once a month,” 6 ¼ “2–3 times a month,” 7 ¼ “nearly every week,” 8 ¼ “every
week,” and 9 ¼ “several times a week”.
Network size. Network size is a control variable since it may be associated with
individual-level characteristics as well as assortativity. For example, men have been
shown to have more expansive social networks than women (Moore, 1990; Pugliesi
& Shook, 1998), and whites have been shown to have larger networks than
non-whites (Ajrouch et al., 2001). Further, individuals with more expansive net-
works may demonstrate higher levels of assortativity because individuals with large
networks find it more difficult to connect people across role-based groups. It might
also be the case that individuals with higher levels of role-involvement tend to have
larger networks because of their higher levels of social involvement compared with
individuals who report low levels of role-involvement.
Analytic Strategy
As outlined above, this paper addresses the extent to which individuals’ social
networks are structured by the social roles they participate in, and whether individual
characteristics explain variation in the level of assortativity in personal social
networks. To address the first issue, I present mean levels of role-based assortativity
and gender- and race-based assortativity, and test whether these values are statisti-
cally different than zero (see Table 12.2). To address the second issue, I estimate
OLS regression models predicting assortativity from role-involvement variables and
demographic and network control variables.
Results
The descriptive statistics for this sample are presented in Table 12.1. Overall, the
sample is mostly female (73.78%), white (83.53%), and tend to come from upper
middle-class families (mean parental income ¼ $98,980, median parental
income ¼ $85,000). Just over ten percent of the sample (10.86%) is married, and
nearly one-third of the sample (32.85%) is in their first semester at Midwestern
University. As this was a sample of college students, the majority of respondents are
in their early 20’s or younger. Nearly 80% of respondents are younger than 24 and
half of the sample was age 20 or younger. As indicated by the average network size
in Table 12.1 (mean ¼ 28.03), the survey is successful in generating relatively
extensive personal social networks. Over 75% of all respondents list at least 20 social
contacts. Finally, student status is relatively evenly distributed, with roughly 20% of
the sample falling into each student status category. Juniors and graduate students
are the most frequent with 21.69%, and sophomores are the least frequent, with
16.63%.
With respect to role-involvement, respondents spend the most time on average
doing school work outside of class (15.82 h per week), followed by work (13.89 h
per week), and just over ten hours (10.67 h) interacting with their family per week.
254 M. H. Walker
Table 12.1 Descriptive statistics (N ¼ 431)

Variable Mean/% Std. dev. Min Max
%Female 73.78 – – –
%White 83.53 – – –
%Married 10.67 – – –
%First semester 32.85 – – –
Parental income (in thousands of USD) 98.98 57.84 5 225
Age 22.30 5.74 18 60
Network size 28.03 9.61 2 50
Hours studying 15.82 9.60 2.5 45
Hours with family 10.67 13.19 2.5 45
Hours of work 13.89 15.36 0 55
Student status
%Freshman 20.96 – – –
%Sophomore 16.63 – – –
%Junior 21.69 – – –
%Senior 19.04 – – –
%Graduate student 21.69 – – –
Religious attendance
%Never 25.52 – – –
%Less than once a year 9.28 – – –
%About once or twice a year 17.87 – – –
%Several times a year 20.88 – – –
%About once a month 5.8 – – –
%2–3 times a month 5.34 – – –
%Nearly every week 8.12 – – –
%Every week 4.41 – – –
%Several times a week 2.78 – – –
Role-involvement in the religious domain (in terms of frequency of religious

attendance) is clearly the lowest. Over one-quarter of the sample (25.52%) do not
attend religious services at all, over half of the sample (52.67%) report attending
religious services “once or twice a year” or less, and only 15.31% report attending
religious services “nearly every week” or more.
Roles and Social Structure
Table 12.2 displays the mean values of assortativity for each of the five role-based
groups, as well as gender, race, and ego-alter closeness. For each dimension,
assortativity is significantly different from zero ( p < .001), suggesting that role-
based group membership, demographic characteristics, and ego-alter closeness each
significantly structure the patterning of ties of individuals’ personal social networks.
Table 12.2 Assortativity Assortativity N Mean SE Std. dev.

based on role-based groups,
Ego-alter closeness 427 0.12*** 0.01 0.21
demographic characteristics,
and ego-alter closeness Race 311 0.05*** 0.01 0.19
Gender 430 0.02*** 0.01 0.11
Friend 402 0.19*** 0.01 0.27
Work 232 0.67*** 0.02 0.35
Family 426 0.43*** 0.01 0.28
Student 375 0.25*** 0.02 0.30
Religion 282 0.13*** 0.02 0.25
***p < 0.001
With the exception of gender, mean associativity is positive, indicating that (1) ties
within role-based groups are more likely to form than ties between role-based others
and non-role-based others, (2) ties are more likely to form between same-race alters
than interracial ties, and (3) ties are more likely to form between alters that are
similarly close with ego. Gender is the only demographic characteristic that
displayed dissortativity, suggesting that cross-gender ties are (slightly) more likely
than same-gender ties. Overall, this suggests that role-based group membership,
race, and ego-alter closeness significantly impact the structure of social ties within
individuals’ personal networks.
Additional patterns are apparent in the assortativity coefficients. First, role-based
groups generally appear to structure the patterning of social ties more than demo-
graphic characteristics and tie strength. As shown in Fig. 12.1, assortativity for all
role-based groups is significantly higher than race- and gender-based assortativity
( p < .05), and with the exception of religion, role-based assortativity is significantly
higher than assortativity based on tie strength. Second, there is substantial variation
in the strength of assortativity across role-based groups. Work-based ties demon-
strate the highest levels of assortativity (r ¼ .67), followed by family members
(r ¼ .43) and school-based ties (r ¼ .25), while religion-based ties demonstrate the
lowest levels of assortativity (r ¼ .13). All mean differences between role-based
groups are statistically significant ( p < .05).
Individual Characteristics, Role-Behavior, and Assortativity
Models 1–5 in Table 12.3 displays the OLS regression models predicting student,
religion, family, friend, and work-based assortativity. Beginning with the student
role, being a graduate student and network size are positively and significantly
associated with student-based assortativity ( p < .001). Additionally, individuals
who spend more time doing schoolwork tend to display higher levels of student-
based assortativity ( p < .05). Finally, role-involvement in the family domain is
negatively associated with student-based assortativity ( p < .01).
256 M. H. Walker
Mean Assortavity
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
E-A closeness Race Gender Friend Work Family Student Religion
-0.1
Fig. 12.1 Mean assortativity and 95% confidence interval for role-based groups and demographic
characteristics
In Table 12.3, we find that whites display significantly lower levels of religion-
based assortativity than non-white respondents (p < .01). Additionally, like the
student role, religious participation is positively associated with religion-based
assortativity (p < .01).
With respect to the family role, both network size and the hours worked per week
are positively associated with family-based assortativity (p < .001 and p < .05,
respectively). Interestingly, the amount of time spent with one’s family is negatively
associated with family-based assortativity (p < .05), suggesting that people who
spend more time with their family are less likely to display family-based
assortativity.
In model 4, the results reveal that there is a negative impact of age on the
assortativity of the friend-based group (p < .01). Further, the number of hours an
individual works per week is positively associated with friend-based assortativity
(p < .001).
Model 5 shows the results for work-based assortativity. Of the variables included
in the model, only gender and hours worked per week affect work-based
assortativity. Women display a higher level of work-based assortativity than men
( p < .05). Controlling for other variables, women’s networks reveal an assortativity
coefficient that is approximately .15 higher than that of men, on average (see
Table 12.3, model 5). As with the student and religion roles, role-involvement in
the workplace is associated with increased clustering of work-based and non-work-
Table 12.3 OLS regression predicting assortativity
12
Role-based assortativity Demographics/tie strength

Student Religion Family Friend Work Ego-alter closeness Gender Race
Variables (1) (2) (3) (4) (5) (6) (7) (8)
Female .002 .037 .034 .001 .147* .005 .013 .017
(.036) (.038) (.034) (.034) (.069) (.026) (.013) (.028)
White .029 .141** .036 .070 .104 .022 .007 .148***
(.042) (.049) (.040) (.039) (.074) (.031) (.015) (.029)
Married .039 .014 .015 .043 .028 .070 .020 .102
(.071) (.073) (.066) (.063) (.108) (.051) (.026) (.054)
Parental income .000 .000 .000 .000 .000 .000 .000* .000
(.000) (.000) (.000) (.000) (.000) (.000) (.000) (.000)
Age .0031 .004 .004 .011** .009 .004 .002 .001
(.004) (.004) (.004) (.004) (.006) (.003) (.002) (.003)
First semester .078 .075 .006 .103* .131 .056 .022 .049
(.049) (.050) (.046) (.044) (.087) (.036) (.018) (.037)
Student status
Sophomore .004 .037 .037 .002 .078 .044 .007 .057
(.062) (.063) (.061) (.059) (.133) (.047) (.023) (.050)
Junior .045 .111 .015 .070 .079 .036 .010 .067
(.060) (.062) (.058) (.056) (.129) (.045) (.022) (.048)
Senior .046 .080 .006 .077 .047 .043 .028 .078
(.067) (.069) (.065) (.063) (.134) (.050) (.025) (.052)
Graduate student .204** .041 .010 .036 .150 .027 .047* .011
Does the Self Reflect Society? Examining the Relationship Between Personal. . .
(.064) (.066) (.061) (.060) (.130) (.047) (.023) (.048)

Network size .009*** .002 .005*** .002 .000 .001 .003*** .000
(.002) (.002) (.002) (.002) (.003) (.001) (.001) (.001)
(continued)
257
258
Role-based assortativity Demographics/tie strength

Student Religion Family Friend Work Ego-alter closeness Gender Race
Variables (1) (2) (3) (4) (5) (6) (7) (8)
Time spent on schoolwork .003* .002 .002 .002 .003 .002 .000 .003*
(.002) (.002) (.002) (.002) (.003) (.001) (.000) (.001)
Religious service attendance .001 .026** .001 .006 .003 .003 .001 .001
(.007) (.008) (.006) (.006) (.012) (.005) (.003) (.005)
Time spent with family .005*** .003 .003* .000 .002 .001 .000 .002
(.002) (.002) (.001) (.001) (.003) (.001) (.000) (.001)
Hours worked for pay .002 .001 .003* .005*** .006* .002* .000 .001
(.001) (.001) (.001) (.001) (.003) (.001) (.000) (.001)
Constant .102 .086 .356*** .326** .392 .220** .057 .219**
(.114) (.115) (.105) (.101) (.203) (.082) (.040) (.083)
N 353 257 390 368 207 391 394 287
R2 .213 .123 .082 .126 .098 .050 .133 .168
Standard errors in parentheses
*p < 0.05
**p < 0.01
***p < 0.001
M. H. Walker
based ties. A one-hour increase in hours worked per week is associated with a .0057
increase in work-based assortativity ( p < .05).
Gender and Race-based Assortativity
Models 6–8 display the results of models predicting assortativity based on ego-alter
closeness, race, and gender. Parental income is positively associated with gender-
based assortativity (p < .05), suggesting that the respondents who come from high
SES families tend to belong to more gender segregated networks than people who
come from low SES families. Further, graduate students display significantly less
gender segregation in their networks than freshman (p < .05). Finally, network size
is positively associated with gender-based associativity (p < .001), suggesting that
individuals with larger personal social networks tend to have more gender segrega-
tion in their networks (i.e., personal networks with more ties within gender than
between genders).
The last column displays the impact of demographic characteristics on race-based
assortativity. Interestingly, whites display less racial assortativity then non-whites
(p < .001), indicating that whites (with at least 2 non-white contacts) have a racial
assortativity coefficient that is .15 lower than non-whites, on average. Finally, hours
per week spent on schoolwork is positively associated with a racially segmented
social network ( p < .05).
Discussion
Consistent with identity theory, these results suggest that social roles and role-based
groups do significantly structure the patterning of ties within individuals’ personal
social networks. Further, role-based groups appear to have a stronger impact on
assortativity than gender or race for the respondents in this study. Thus, role-based
others appear to represent a group in a sense that is meaningful beyond merely
participating in the same institution or organization. Although role-based groups do
have an impact on the structure of personal social networks, they clearly do not
strictly determine the patterning of social ties. While intragroup ties are more
common than intergroup ties, intergroup ties are nonetheless pervasive. Taken
together with recent research outlining the importance of intergroup ties for identity
salience (Walker & Lynn, 2013), these findings show that intergroup ties appear to
be both impactful and pervasive. Thus, the interrelations between individuals’
various identities should be a key concern for future empirical research in identity
theory.
In addition to revealing substantial social overlap between role-domains, this
research indicates significant variation in assortativity across role-based groups. By
providing a descriptive account of the level of role-based assortativity, this research
260 M. H. Walker
can inform future work linking social network processes to identity processes. For
example, if social ties extending outside of role-based groups are important for
identity salience, it is useful to know that assortativity tends to be higher amongst
work-contacts than amongst friends or church contacts. Additionally, the generally
positive relationship between assortativity and time spent in a role suggests that
causal direction in the link between social networks, the self, and behavior may be a
complication that deserves more attention in future research. Since social networks
are hypothesized to impact identity salience, which in turn affects role-choice
behavior, it is important to know whether role-involvement is causing changes in
social networks or vice-versa.
This study also highlights the impact that demographic characteristics and role-
involvement have on role-based assortativity and may shed light on general network
processes that occur within role-based groups. For instance, while role-involvement
tends to increase role-based assortativity for the student, religion, and work roles, it
appears to have the opposite effect for the family role. While spending increased
time on school work, at religious services, and at work tends to result in more
segmentation between role-based others and non-role-based others, increased time
spent with family members tends to reduce segmentation between family members
and non-family members, effectively integrating family members into the rest of
one’s personal social network.
Though I can only speculate at this stage, this pattern may be explained by
sequencing and life course processes. Familial ties are, in almost all cases,
longstanding role-based relationships, who are part of one’s network of origin. On
the other hand, since the respondents in this sample are college students at a large
state university, student-based, religion-based, and work-based ties are likely ties
that have been formed within the past four years. Thus, it could be that transitivity or
the general tendency for ties to form between two people who are tied to the same
person (Wasserman & Faust, 1994) may be more likely to occur within role-domains
at early stages of role-participation. Increased role-involvement in newly formed
role-based groups may be more likely to increase ties between the role-based others
that ego has formed ties with (e.g., ego may introduce two fellow students who had
previously not known one another) than it is to increase ties between role-based
others and non-role-based other (e.g., ego introducing a friend from school to a
friend from “back home”). Conversely, for longstanding role-based groups like the
family, within-group transitivity may be saturated, and role-involvement might only
serve to integrate members of the role-based group into one’s overall network. For
example, spending more time with one’s family could make it more likely that those
family members will be actively involved in the various domains of one’s life,
thereby integrating family members into ego’s overall social network and decreasing
assortativity.
In general, this research highlights the potential for integrating social network
analysis with identity theory. While theoretical statements of identity theory have
long acknowledged the interdependence of identities and the social overlap between
role-based groups, empirical work has only recently begun to directly address these
issues (e.g., McFarland & Pals, 2005; Stryker et al., 2005; Walker & Lynn, 2013).
Social network analysis is particularly well-suited to address the issue of social

overlap between role-based groups because one key aspect of social overlap between
role-based groups are intergroup ties. The strength of social network analysis is the
focus on directly eliciting social contacts and the structure of social relations
between social contacts from survey respondents so that one can concretely analyze
network structure.
Additionally, social network analysis can enrich identity theory by providing the
methodological tools to assess potential network mechanisms driving identity pro-
cesses. For example, research suggest that both ties between role-based groups and
the multiplexity of ties of role-based others matter for identity processes
(Markowski, 2019; Walker & Lynn, 2013). However, almost by definition, multi-
plex ties tend to produce intergroup ties and vice-versa. That is, a coworker who
attends the same church as ego is likely tied to ego’s contacts from both church and
work. Additionally, ties between ego’s work and religious networks increases the
likelihood that coworkers will begin attending ego’s church. However, multiplex ties
have been theorized to enhance social and emotional commitment to a role-based
group, which is a bonding/binding mechanism (Stryker, 1980), whereas intergroup
ties are viewed as indicative of other network mechanisms such as the flow of
information and meaning, and environmental adaptation. Social network methods
are essential for distinguishing between these possible mechanisms.
Although this research takes an important first step in understanding the extent to
which role-based groups structure the patterning of social ties in personal social
networks, there are some limitations that should be considered when interpreting the
findings. First, this study uses a convenience sample of college students from a
single university. Thus, the network patterns observed are not necessarily represen-
tative of any population. Indeed, the patterns observed would likely differ from those
of middle-aged careerists or elderly retirees, given the difference in life course stage
and the corresponding difference in social environments. Future research using
probability sampling procedures or research that focuses on individuals at different
stages in the life course is needed to fully address this issue. However, this research
may shed some light on the basic network processes that govern tie formation within
and across role-based groups. Further, the threshold in this survey for the existence
of alter-alter ties or “which of these pairs knows each other well enough to stop and
chat if they passed one another on the street” is relatively low. It is unclear if the
patterns observed here would hold if social relations were set above the level of
“knowing” to some stronger threshold.
Overall, this research suggests that, while role-based groups do structure individ-
uals personal networks, there is still considerable overlap in social relations between
role-based groups. Thus, the image of individuals as living their lives in relatively
stable and specialized role-based groups is only partially accurate. Thus, future
research on the link between social networks and the self-structure should attend
not only the number and strength of ties to role-based others, but also consider the
structure of social relations within and between role-based groups. Additionally, this
research suggests that network and identity processes do not occur in isolation,
262 M. H. Walker
highlighting substantial interdependence and social overlap between role-based

groups.
Given the recent upsurge of interest in applying a social networks perspective to
identity theory (Markowski, 2019; McFarland & Pals, 2005; Walker, 2015; Walker
& Lynn, 2013), this study has the potential to inform future research adopting a
networks approach by illuminating the basic network processes that occur within
and between role-based groups. For instance, the consistent association between
role-based assortativity and role-participation may shed light some light on the
mechanisms involved in the reciprocal relationship between network processes,
identity processes, and identity-related behavior. Future research should utilize
longitudinal methods along with social network methods to track how networks
dynamics and identity processes unfold within and between role-based groups
over time.
References
Ajrouch, K. J., Antonucci, T. C., & Janevic, M. R. (2001). Social networks among blacks and
whites: The interaction between race and age. The Journal of Gerontology Series B: Psycho-
logical Sciences and Social Sciences, 56(2), S112–S118.
Borgatti, S. P., & Halgin, D. S. (2011). On network theory. Organization Science, 22(5),
1168–1181.
Borgatti, S. P., & Lopez-Kidwell, V. (2011). Network theory. In J. Scott & P. J. Carrington (Eds.),
The sage handbook of social network analysis (pp. 40–54).
Coleman, J. S. (1988). Social capital in the creation of human capital. American Journal of
Sociology, 94, S95–S120.
DiPrete, T. A., Gelman, A., McCormick, T., Teitler, J., & Zheng, T. (2011). Segregation in social
networks based on acquaintanceship and trust. American Journal of Sociology, 116(4),
1234–1283.
Feld, S. L. (1981). The focused organization of social ties. American Journal of Sociology, 86(5),
1015–1035.
Granovetter, M. (1992). Problems of explanation in economic sociology. In N. Nohria & R. G.
Eccles (Eds.), Networks and organizations: Structure, form, and action (pp. 25–56). Harvard
Business School Press.
Huckfeldt, R. R., & Sprague, J. (1995). Citizens, politics, and social communication: Information
and influence in an election campaign. Cambridge University Press.
Knoke, D. (1990). Networks of political action: Toward theory construction. Social Forces, 68(4),
1041–1063.
Kossinets, G., & Watts, D. J. (2006). Empirical analysis of an evolving social network. Science, 311
(5757), 88–90.
Kossinets, G., & Watts, D. J. (2009). Origins of homophily in an evolving social network. American
Markowski, K. L. (2019). Identity, networks, and mental health: The relationship between struc-
tures and meaning on distress and subjective wellbeing. Doctoral dissertation, Kent State
University.
Marsden, P. V. (1987). Core discussion networks of Americans. American Sociological Review,
52(1), 122–131.
Marsden, P. V. (1988). Homogeneity in confiding relations. Social Networks, 10(1), 57–76.

McFarland, D., & Pals, H. (2005). Motives and contexts of identity change: A case for network
effects. Social Psychology Quarterly, 68(4), 289–315.
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily in social
networks. Annual Review of Sociology, 27(1), 415–444.
Moore, G. (1990). Structural determinants of men’s and women’s personal networks. American
Newman, M. E. J. (2002). Assortative mixing in networks. Physical Review Letters, 89(20), 28701.
Newman, M. E. J. (2003). Mixing patterns in networks. Physical Review E, 67, 026126.
Pugliesi, K., & Shook, S. L. (1998). Gender, ethnicity, and network characteristics: Variation in
social support resources. Sex Roles, 38(3–4), 215–238.
Riley, A., & Burke, P. J. (1995). Identities and self-verification in the small group. Social
Stryker, S. (1968). Identity salience and role performance. Journal of Marriage and Family, 4,
558–564.
Stryker, S. (1980). Symbolic interactionism: A social structural version. Benjamin/Cummings
Publishing Company.
Stryker, S., & Serpe, R. T. (1982). Commitment, identity salience, and role behavior: Theory and
research example. In Personality, roles, and social behavior (pp. 199–218). Springer.
social structures on commitments. In W. Ickes & E. S. Knowles (Eds.), Advances in group
processes (pp. 199–218). Springer-Verlag.
Walker, M. H. (2015). The contingent value of embeddedness: Self-affirming social environments,
network density, and well-being. Society and Mental Health, 5(2), 128–144.
Walker, M. H., & Lynn, F. B. (2013). The embedded self: A social networks approach to identity
Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications. Cambridge
University Press.
Yamaguchi, K. (1990). Homophily and social distance in the choice of multiple friends an analysis
based on conditionally symmetric log-bilinear association models. Journal of the American
Statistical Association, 85(410), 356–366.
Chapter 13
The Structure of Friendship: How Network
Density Influences Identity Verification
Kelly L. Markowski
The Structure of Friendship: How Network Density

Influences Identity Verification
Within sociology, identity theory serves as a key framework for the study of the
relationship between individuals, their social interactions, and the structure of
society (Burke, 1991; Burke & Stets, 2009; McCall & Simmons, 1978; Stryker,
1968, 1980). One well-established area of investigation within identity theory
involves an internalized process in which individuals compare the meanings that
they associate with their identities to the meanings they perceive that others hold for
them in their identities (Burke, 1991; Burke & Stets, 2009). This process is identity
verification. Previous studies have consistently shown that non-verification—or
perceptions that others do not see individuals in the way individuals see them-
selves—is associated with a host of negative self-outcomes (Burke & Stets, 1999;
Cast, 2004; Riley & Burke, 1995; Stets & Burke, 2014a, 2014b). For example, an
individual is likely to feel sadness, anger, or shame when she thinks of herself as a
reliable friend but perceives that her friends see her as unreliable.
The identity verification process is complex and nuanced. Scholars have posited
that outcomes associated with non-verification are contingent on other factors, such
as how important, or prominent, an identity is for the individual (McCall &
Simmons, 1978) as well as how salient, or likely to be enacted, the identity is
(Burke, 1991; Burke & Stets, 2009). The idea is that an individual should feel
more sadness, anger, or shame if she is non-verified in an identity that is very
important (compared to not at all important) to how she sees herself. Similarly, an
individual should also feel more sadness, anger, or shame if she is non-verified in an
identity that she is likely to enact across situations compared to an identity that is
K. L. Markowski (*)
University of Nebraska-Lincoln, Lincoln, NE, USA
266 K. L. Markowski
rarely enacted. While one study finds support for this idea (Ellestad & Stets, 1998), a
recent study further testing these ideas across a broader range of identities and
outcomes finds mixed support for the proposed relationships (Markowski & Serpe,
2018).
In this paper, I suggest that further contextualizing the verification process may
help resolve the mixed support. Toward this end, I examine the role of social
structure in this relationship. I focus on social structure for two reasons. First, a
foundational premise undergirding identity theory is that the self is a reflection of
society. Aspects of identities, like salience and prominence, are themselves
influenced by the social structures in which individuals find themselves (Serpe &
Stryker, 2011; Stryker, 1980). Second, social structure provides information about
the social contexts in which identity processes unfold (Merolla et al., 2012; Serpe
et al., 2019; Stryker et al., 2005). It also provides insight on interpersonal resources,
like trust and social support, that derive from social relationships and that individuals
can use to offset the effects of non-verification (Freese & Burke, 1994; Stets & Cast,
2007). For these reasons, I suggest that the outcomes of identity verification may
vary across features of the social structure.
Though identity theory provides several ways to assess social structure, I expand
these assessments by drawing from social network analysis. Social network analysis
provides a wide range of tools by which to examine social context through the
configuration of social relationships. I focus here on density, which refers to the
proportion of one’s social contacts that are connected to one another (Wasserman &
Faust, 1994). Density offers an advantage over other assessments of structure in
identity theory because instead of measuring aspects of relationships that exist
between an individual and their social contacts, such as how similar social contacts
are to an individual, density measures the connections that exist among one’s social
contacts. This provides unique information about the social context in which identity
verification occurs as well as the resources that are likely afforded by the social
context.
Thus, the goal of this paper is to examine the role of density in the identity
verification process. Given the central role that friendships play in social life (Sias &
Bartoo, 2007), I focus on verification as it unfolds in friendship relationships within
adults’ larger friendship networks. Do individuals with more dense friendships have
better or worse outcomes from non-verification compared to those with less dense
friendships? This is important to address because it gives a clearer understanding of
how social context shapes the non-verification experience in a commonly-held
identity: the friend identity.
Identity Theory
Identity theory (Burke, 1991; Burke & Stets, 2009; McCall & Simmons, 1978;
Stryker, 1968, 1980) derives from structural symbolic interactionism, a perspective
that views social behavior as the product of individuals’ selves, which are, in turn,
13 The Structure of Friendship: How Network Density Influences Identity. . . 267
reflections of the societies in which individuals participate (Serpe & Stryker, 2011).
From this perspective, selves are comprised of identities. Identities are the shared set
of meanings that define individuals as occupants of social positions, as members of
groups, or as people with characteristics that make them unique from others (Burke
& Stets, 2009; Stets & Serpe, 2013). These are referred to as role, group, or person
identities, respectively.
Regardless of the type of identity under consideration, identity theory examines
different aspects of identities as well as the processes in which identities are
implicated. For example, some work examines features of relationships that exist
between an individual and their social contacts, focusing on the effect that these
features have on the way in which individuals cognitively organize their identities
(Stryker, 1968, 1980). In this work, commitment, or the emotional intensity and
interactional frequency between an individual and the people they interact with
because of an identity, has been shown to influence individuals’ rankings of how
important, or prominent, an identity is to how individuals see themselves (Yarrison,
2013). Commitment has also been shown to influence how salient an identity is—or
the likelihood that the identity will be enacted across situations (Stryker & Serpe,
1982; Yarrison, 2013).
Other identity theory work focuses on the internal mechanisms by which identi-
ties guide behavior. The central focus of this work is the perceptual control process
(Burke, 1991). The perceptual control process consists of four components: the
identity standard, perceptual input, the comparator, and behavioral output. First,
individuals associate meanings with their identities, and these meanings are stored in
the identity standard. Individuals have working copies of their identity standards as
they apply to specific situations, and these working copies can be referred to as self-
views. Second, individuals perceive input, or situation-based meanings related to
their identities. An important part of perceptual input is reflected appraisals, which
are the meanings that individuals imagine others hold for them in the identity. Third,
the comparator assesses the degree to which the perceptual input matches the self-
view. When meanings match or are close to matching, one experiences identity
verification. However, the more that meanings are discrepant from one another, the
more one experiences non-verification.
When individuals experience verification, individuals feel good, and when indi-
viduals experience non-verification or discrepancy, they feel bad. The greater the
degree of non-verification, the more negative the feelings. This relationship is robust,
as research has confirmed the link between non-verification and distress (Burke &
Stets, 1999), negative emotions (Stets & Burke, 2014a), and levels of depression and
anxiety (Cast, 2004) as well as lower self-esteem (Markowski & Serpe, 2018; Stets
& Burke, 2014b) and lower role satisfaction (Riley & Burke, 1995). Importantly,
when negative feelings emerge, the perceptual control model posits that individuals
modify their behavior and self-views in an endeavor to both reduce the negative
feelings and bring the meanings in closer alignment (Stets et al., 2020).
Additional work in identity theory adds further nuance to the perceptual control
model. Aside from specifications on how the model operates with multiple identities
simultaneously (Burke, 2006), scholars have long suggested that other aspects of
268 K. L. Markowski
identities matter for the verification process. It has been argued that the degree to
which identity non-verification is associated with negative feelings depends on the
prominence and salience levels of the identity (Burke, 1991; Burke & Stets, 2009;
McCall & Simmons, 1978). Specifically, an individual should feel more sadness,
anger, or shame if she is non-verified in an identity that is very important to how she
sees herself or that is highly salient. The idea is that non-verification becomes more
threatening the more the identity occupies a relevant place in an individual’s social
world.
One study tested this idea in the parent identity (Ellestad & Stets, 1998). The
authors found support for the effect, as mothers whose parent identities were higher
in prominence reported more jealousy in response to a vignette where
non-verification was presumed to occur. A more recent study tested this idea on a
wider range of outcomes and identities using survey data (Markowski & Serpe,
2018). Importantly, this study examined both prominence and salience as modera-
tors for identity verification in the parent, childless, spouse, and single identities but
found few effects on the role-specific worth, efficacy, and authenticy components of
self-esteem. Why might the findings of the two studies differ?
One reason for the lack of findings may be because Markowski and Serpe (2018)
used a measurement of identity verification that examined only one meaning asso-
ciated with each identity. Another reason may be that other variables that were not
included in either of the two studies, such as social structure, may influence the
verification process. This is a possibility for two reasons.
First, identity theory views individuals’ selves as reflections of the societies in
which individuals participate (Serpe & Stryker, 2011; Stryker, 1980). Research has
shown that the prominence and salience aspects of identities are themselves
influenced by the social structures in which individuals find themselves (Stryker &
Serpe, 1982; Yarrison, 2013). Thus, it may be fruitful to examine assessments of
social structure directly as the primary feature that matters for outcomes associated
with non-verification instead of prominence or salience.
Second, social structure provides important information about social context that
is not captured by concepts like salience, prominence, or even commitment, the latter
of which served as the primary assessment of social structure in identity theory until
recently (Merolla et al., 2012; Serpe et al., 2019; Stryker et al., 2005). In clearer
specifications of social structure, some scholars examine what they label proximate
social structure, or the smallest, most immediate social contexts in which individuals
maintain social relationships and enact identities (Merolla et al., 2012; Stryker et al.,
2005). Examples of these proximate contexts include small groups, such as the
family, or a program-based group, like an enrichment program designed to improve
retention among minority students in STEM (Merolla et al., 2012).
One study measures proximate social structure according to whether an individ-
ual is a member of a small group at all (Merolla et al., 2012). Another measures
proximate social structure according to the properties that describe the group.
Specifically, Yarrison (2016) measures proximate social structure as the degree to
which members in a group share the same religious beliefs as the study participant.
These assessments of proximate social structure are in contrast to concepts like
commitment: instead of describing aspects of social relationships that potentially

span across groups and that, on average, exist between an individual and those with
whom they interact because of an identity (e.g, average emotional intensity or
average interaction frequency), proximate social structure involves the larger
group context in which such specific relationships, interactions, and identity pro-
cesses occur. The difference is thus one of scope, with proximate social structure
emphasizing group boundaries and group-level properties.
Information about group-level properties is important because it provides infor-
mation about the resources that individuals have and can use to influence the
verification process (Freese & Burke, 1994). For example, studies have shown that
structural resources, like education and gender, provide status that allows one to
influence meanings and situations in ways that lead to a greater likelihood of
experiencing verification (Cast et al., 1999; Stets & Cast, 2007; Stets & Harrod,
2004). Similarly, interpersonal resources, such as trust and social support, derive
from social relationships themselves and similarly aid in producing the likelihood of
verification (Stets & Cast, 2007). In a related vein, I argue that individuals also use
interpersonal resources to offset the effects of non-verification. Thus, to the extent
that proximate social structure indicates the availability of interpersonal resources, it
is likely that social context indexed by proximate social structure will influence the
extent to which non-verification negatively affects individuals.
As is clear from above, past identity theory work provides several options by
which to examine proximate social structure, and thus, how proximate social
structure influences the identity verification process. However, these options are
limited in their ability to capture access to interpersonal resources: though member-
ship in a group (Merolla et al., 2012) or the degree to which group members are
similar to an individual (Yarrison, 2016) may correspond to important information,
like the degree to which one is liked by others in a specific social context, these
assessments do not necessarily indicate whether interpersonal resources such as
social support are available in that context. A clearer, more direct indication of
social support is thus needed. I draw from social network analysis for this clear index
of social support and focus on the concept of density, proposing it as an alternative
measure of proximate social structure.
Network Density as Proximate Social Structure
In social network analysis, density broadly refers to the level of connectedness

among the members in one’s network or bounded group. It is assessed by calculating
the proportion of social contacts that are connected to one another (Wasserman &
Faust, 1994). Connections (or ties) can be defined in a variety of ways. For example,
a tie can exist if people are emotionally close to one another or perhaps if they
communicate with one another regularly, etc. In any case, it is important to highlight
that density does not assess the ties that exist between the respondent and their
network members; this is because it is assumed that the respondent is connected to
270 K. L. Markowski
all members they list as part of their personal network. Instead, density assesses the
ties that exist among one’s network members. When many ties exist among mem-
bers, the network is high in density. When few ties exist, the network is low in
density.
To illustrate, suppose that Alice lists all her close friends, which includes Brian,
Cate, and Danielle. Also suppose that connections are defined as whether Alice’s
friends are friends with one another. The total possible number of ties that could exist
among Alice’s friends is three: Brian and Cate, Brian and Danielle, and Cate and
Danielle could be friends with one another. If each of these possible friendship dyads
actually existed, Alice’s friend network would be high in density because three ties
exist out of a total possible number of three ties. In this scenario, the density score is
one. If, however, only Brian and Cate are friends, Alice’s friend network is low in
density because only one tie exists out of a total possible number of three tie
connections, and the density score is relatively low at .33.
What makes density a unique alternative measure of proximate social structure?
First, density assesses social context and group dynamics at the network-level, which
serves as the backdrop against which specific relationships unfold. This larger group
dynamic can influence interactions between an individual and their network mem-
bers as well as identity verification. For example, an individual may have tense or
strained friendships if the overall dynamic among one’s friends is contentious, and
this might lead to non-verifying situations (e.g., Hallinan & Kubitschek, 1988).
Similarly, engagement in a network of harmonious ties may lay the foundation for
close friendships and mutually-verifying situations in which individuals verify each
other (Burke & Stets, 1999, 2009). In short, density provides information that can be
used to make predictions on the likelihood that verification will occur.
Second, and perhaps most critically for this analysis, density provides unique
information about the interpersonal resources that individuals can draw upon when
they feel non-verified. Highly dense networks tend to be cohesive networks in which
members share strong emotional bonds that are durable despite adverse events
(Friedkin, 2004; Green et al., 2001). Strong ties like these are likely to benefit
individuals because they confer social support (Berkman et al., 2000; Heaney &
Israel, 2008; Kawachi & Berkman, 2001; Perry et al., 2018). To this point, some-
times researchers use density as a proxy for studying social support in one’s network
(Siebert et al., 1999). Thus, it is likely that individuals in dense networks can buffer
the negative effects of non-verification because they exist in supportive networks
from which they can draw support from specific relationships to mitigate negative
feelings, as needed.
On a final note, other scholars have tested relationships similar to the ones of
interest here (e.g., Walker, 2015). However, the emphasis of the work differed, as
did the predicted relationship. There, the focus was on how the relationship between
network density and self-esteem is moderated by authenticity, or whether individuals
feel that they can ‘be themselves’ in general. It was predicted (and found) that high
density was associated with lower self-esteem when individuals felt inauthentic in
their environments. By contrast, the focus here is on the moderating role of density in
the identity verification process, where I expect that the negative effects of
non-verification in the friend identity can be buffered by the interpersonal resources

associated with high density in one’s friendship network. Thus, I extend this past
work by directly assessing identity verification (instead of assessing authenticity,
which has been found as an outcome of the verification process; Stets & Burke,
2014b) while also limiting my analysis to a specific identity-based network context:
friendship.
Current Study and Hypothesis
To test the idea that network density affects identity discrepancy, I focus on the
friendship identity and do so for several reasons. First, friendships occupy a central
role in social life across all stages of the life course, making it an especially common
identity that is held by members of the general population (Sias & Bartoo, 2007).
Indeed, research finds that virtually everyone has at least one close friend (Gillespie
et al., 2015). Second, although research finds that the number of friends declines
with age (Kalmijn, 2003), there is evidence that friendships become more important
in later life (Fiori et al., 2020), especially if one loses or never had a spouse or partner
to begin with (Demir, 2010). This supports the continued need to examine friend-
ships among adults. Last, friendships are important social relationships. They pro-
vide feelings of safety, security, and support that influence both mental and physical
health (Greco et al., 2015; Taylor et al., 2000). For these reasons, it is important to
understand how network context influences identity verification in the friend
identity.
Consistent with the above, I expect that density will moderate the relationship
between identity discrepancy and negative self-outcomes—specifically, negative
emotions. I examine emotions because theory and empirical work on identity
verification and the expected moderation effects have focused on emotions as a
central outcome (Ellestad & Stets, 1998; Stets & Burke, 2014a). I thus hypothesize
the following:
H1: As non-verification in the friend identity increases, individuals will
experience fewer negative emotions if one’s friendship network density is
high compared to low.
Method
Data and Sample
This study uses data from an elecronic survey administered in 2018 to 500 members
of a non-probability panel. Although generalizations to the larger population cannot
be made with non-probability samples, they can be used to investigate general
272 K. L. Markowski
Table 13.1 Factor analysis Item Factor loading

for negative emotions
Negative emotions (alpha reliability ¼ 0.87)
Sad 0.80
Angry 0.72
Ashamed 0.81
Guilty 0.81
processes within identity theory (Adams & Serpe, 2020; Markowski & Serpe, 2018;
Serpe et al., 2019), which is the purpose of this work. Panelists were eligible to
participate in this survey if they were U.S. citizens that were at least 18 years of age,
employed at least part-time, and had a spouse or partner who also was employed at
least part-time. Part of the survey asked participants about their friendship networks
and friend identity; this is the data of interest here.
The total usable sample for this analysis included participants who identified two
or more friends in their network. This is because a density measure cannot be
calculated for participants who indicated only one or no friends (i.e., at least two
friends need to be listed for a dyadic tie to be possible). Among the 277 participants
who satisfied this criterion, the average age of individuals was 46 years. Just over
half the sample (57%) were women. Just over one-fourth of respondents (27%) were
racial or ethnic minorities. The majority of the sample (66%) had at least one college
degree.
Measures
Negative emotions. Negative emotions served as the dependent variable and were
measured by asking respondents to indicate how often (1 ¼ “None of the time,”
2 ¼ “A little of the time,” 3 ¼ “Some of the time,” 4 ¼ “Most of the time,” or
5 ¼ “All of the time”) in the past 30 days they felt: (1) sad, (2) angry, (3) ashamed,
and (4) guilty. These items were included to represent both primary and secondary
emotions. Sadness and anger are considered primary emotions, or emotions that
serve as primary responses to the environment, and shame and guilt are considered
secondary emotions, or emotions that are socially constructed (Turner & Stets,
2005). Though distinct kinds of emotions, the alpha reliability for all negative
emotion items was high (.87), and each item loaded well onto a single negative
emotions dimension (see Table 13.1).
Density. To assess density, participants needed first to provide a list of individuals
within their social network. A ‘name generator’ (Marsden, 1990) was used to prompt
participants to list these individuals. The prompt for this study was designed
according to guidelines by Perry et al. (2018) and was as follows:
We want to learn about the people in your life, such as your family, friends, and people you
work with. Who do you see or talk to on a regular basis? This may include face-to-face
interaction as well as online or other technology-mediated interactions. Please list those

people below using only first names and last initials. Duplicates are not allowed.
The page was formatted to only show one blank space at a time. Once the
participant indicated a person’s name, they were directed to proceed to the next
page, which displayed that name as well as all other names that had been previously
listed. The following text was shown under the list and was followed by another
single blank textbox:
Who else do you see or talk to on a regular basis? Remember we are interested in the
people in your life, such as family, friends, and people you work with.
Participants were permitted to list up to 20 names in total in response to the name

generator. On average, participants listed names for 8.65 individuals with a range of
2 to 20 individuals in total.
Once participants were satisfied with their list, a ‘name interpreter’ (Marsden,
1990) was used to identify how participants knew or were related to each of the
people they listed. Individuals indicated if each participant was their “Spouse/
partner,” “Child,” “Parent (including step-parents),” Sibling (including step-sib-
lings),” “Part of your family (not spouse/partner, children, parents, or siblings),”
“Close friend (someone that you know and can count on if you need them),” or “Co-
worker (including subordinates, bosses, etc.).” Of interest to this analysis were the
people that participants indicated as a “Close friend.”
To measure density, it was necessary to capture the number of connections that
existed between members of one’s friend network out of all possible connections. To
do this, a matrix was presented that listed all possible dyadic combinations between
the individuals that participants indicated as friends through the name interpreter.
The formula for the number of pairs the participant was shown (and thus, the total
possible number of tie connections that could exist in the friend network) is shown in
Eq. (13.1), where n refers to the number of friends indicated by the participant.
nðn 1Þ
¼ total number of friend pairs=possible tie connections ð13:1Þ
2
For each dyad, participants were asked to indicate whether the individuals
“Would stop and talk to each other if they passed one another on the street.” A tie
or connection exists if participants affirmatively indicated that the pair of individuals
would stop and talk to each other (Walker, 2015; Walker & Lynn, 2013). Though
this type of density prompt implies a geographical closeness requirement, it is
hypothetical in nature such that the focus of the prompt is whether the two individ-
uals are linked intimately enough to recognize and acknowledge each other in the
absence of geographical constraint.
Once this information is indicated for all dyads, density can be calculated by
taking the number of existing ties divided by the total number of possible ties (see
Eq. 13.2). The result is a proportion. Thus, density scores can range from 0 to 1, with
0 indicating the least dense friend network possible, and 1 indicating the most dense
friend network possible.
274 K. L. Markowski
number of existing dyadic ties among friends

¼ density ð13:2Þ
total possible dyadic ties among friends
Identity Discrepancy. To assess identity discrepancy, or both verification and

non-verification, participants completed two questions. First, participants were asked
to indicate how they view themselves in the friend identity. This item indicates the
self-view. “Positive” served as the meaning of interest, and participants indicated
their self-view on a bipolar scale ranging from 1 (“Not at all positively”) to
10 (“Very positively”). Though this assessment is not ideal because it does not
incorporate a set of meanings, it is consistent with previous abbreviated assessments
of identity that focus on one overall, evaluative meaning (e.g., Stets & Harrod,
2004).
Second, participants were asked to indicate on the same bipolar scale how they
thought their close friends viewed them in the friend identity. This item indicates the
reflected appraisal. It should be noted that the reflected appraisal item asked
participants to imagine how their close friends in general viewed them as opposed
to how they imagined each separate friend viewed them. Though this is not ideal
because it is possible that variation exists in reflected appraisals across friends, it is
also consistent with previous reflected appraisal assessments in which the item asks
the participant to evaluate their close friends in general (e.g., Stets & Harrod, 2004).
From the two items, an identity discrepancy measure was constructed by
subtracting the self-view from the reflected appraisal. If the self-view and the
reflected appraisal scores are equal, the discrepancy score is zero. If the scores differ,
the discrepancy score is non-zero. The closer the score is to zero, the more likely
individuals are to experience verification. The further away the score is from zero,
the more individuals experience non-verification. A positive discrepancy score
indicates that an individual views their identity less positively than their friends. A
negative discrepancy score indicates that an individual views their identity more
positively than their friends. For affective outcomes, discrepancies are squared
because they theoretically become more negative as the magnitude of the
non-verification, positive or negative, increases (Burke & Stets, 2009; Stets &
Burke, 2014a, 2014b).
Additional variables. Other variables were included in this analysis. They
included age, gender, race/ethnicity, education, network size, friendship length,
frequency of communication with friends, and emotional closeness to friends. Age
was measured in years. Gender was coded as 0 for men and 1 for women. Race/
ethnicity was coded as 0 for whites and 1 for racial/ethnic minorities. Level of
education was coded as 0 for less than high school, 1 for high school graduate, 2 for
some college or technical school, 3 for a college graduate, and 4 for graduate or
professional school. These variables were included because each is associated with
the likelihood of being verified, with higher statuses in general experiencing more
verification and greater psychological well-being (Burke & Stets, 2009).
Network size was measured as the total number of friends included in the
network. Network size was included because of its negative relationship to density:
the larger the network, the more ties that are possible between network members but
the less likely that those greater numbers of possible ties will exist as actual ties due
to various constraints, like time (Carter & Feld, 2004). Friendship length was
measured on a graded scale where less than one year was coded as 1, one to three
years was coded as 2, three to five years was coded as 3, five to ten years was coded
as 4, and more than ten years was coded as 5. Frequency of contact was included as
the proportion of friends that the participant talked to on at least a weekly basis.
Finally, emotional closeness to friends was included, where scores ranged from
1 (“Not at all close”) to 10 (“Extremely close”) and were averaged across friends
listed. These latter variables were included because each is associated with emotions.
Individuals are less happy with friendships that lack closeness and frequent interac-
tion (Demir et al., 2006), and longer-lasting friendships are associated with more
positive feelings (Mendelson & Aboud, 1999).
Analytic Technique
I used a series of multivariate regressions to test the hypothesis. Factor weights from
the factor analysis for negative emotions were used in the regressions. According to
identity theory, only the squared identity discrepancy variable is of interest for
verification effects on affective outcomes (Burke & Stets, 2009). For this reason, I
report estimates from a regression in which only the squared variable is included
both alone and in an interaction term with density. Following guidelines on quadratic
moderation, I did also estimate a regression in which all lower order variables were
estimated alongside the squared term and the interation with the squared term (Baron
& Kenny, 1986; Dawson, 2014). Though I do not report these latter results, it should
be noted that they did not differ substantively from the results reported here and thus
lead to similar conclusions as discussed below.
Results
Table 13.2 presents the means and standard deviations for the variables included in
this analysis. The average level of negative emotions was low among participants
(M ¼ 1.82). In this sample, the average friendship network density score was
moderate (M ¼ .45). Overall, respondents indicated relatively high positive self-
views in the friend identity (M ¼ 8.67) and only slightly higher reflected appraisals
from friends (M ¼ 8.79). Nonetheless, a slight discrepancy was present on average
(M ¼ 1.14). Participants listed 4.25 friends on average, with a range of 2 to
15 friends. Among the friends listed, the majority of participants (74.37%) reported
that the average friendship length was more than five years. Participants communi-
cated weekly or more with a high proportion of their friends (M ¼ .76). Last,
participants were, on average, emotionally close to their friends (M ¼ 7.47).
276 K. L. Markowski
Table 13.2 Descriptive sta- Variables Mean SD Range

tistics (N ¼ 277)
Negative emotions 1.82 0.72 1–5
Density* 0.45 0.39 0–1
Self-view 8.67 1.37 3–10
Reflected appraisal 8.79 1.29 4–10
Identity discrepancy* 1.14 3.41 0–36
Network size 4.25 2.59 2–15
Friendship length 4.27 0.89 1.20–5
Frequency of contact 0.76 0.32 0–1
Emotional closeness 7.47 1.72 1–10
Age 46.44 14.39 21–76
Woman 0.57 0.50 0–1
Racial/ethnic minority 0.28 0.45 0–1
Education 2.75 0.86 0–4
*Items were standardized prior to the analysis
Table 13.3 presents the correlations between all variables included in the analysis.
These correlations reveal that negative emotions are positively related to identity
discrepancy (r ¼ .26) and racial/ethnic minority status (r ¼ .16), though they are
negatively related to both friendship length (r ¼ .41) and frequency of contact
(r ¼ .15). In addition to its positive association with negative emotions, identity
discrepancy is negatively related to frequency of contact (r ¼ .15), age (r ¼ .18),
and education (r ¼ .17). Interestingly, friendship network density is not related to
the total number of friends in the network, nor is it related to the average level of
emotional closeness to friends or to identity discrepancy. It is, however, positively
related to friendship length (r ¼ .22) and frequency of contact (r ¼ .17).1
To understand how density interacts with identity discrepancy on negative
emotions, I estimated two regression equations. The results are shown in
Table 13.4. Model 1 is the baseline model that includes only the main effects of
density and identity discrepancy (plus all additional variables). Model 2 is identical
to Model 1 except with the inclusion of the interaction term between density and the
squared discrepancy term.
As shown in the results from Model 1, density is not directly associated with
negative emotions; only the identity discrepancy term is significant and positive
(β ¼ .22), which reflects that increases in discrepancy are associated with greater
1
Network size tends to be negatively associated with density (Carter & Feld, 2004), and density
tends to be associated with close emotional bonds among network members (Friedkin, 2004; Green
et al., 2001). The lack of associations seen here, however, may be explained by the boundary
specification of the friendship network. The name interpreter specifically asked individuals to
identify their close friends as opposed to friends in general, which would have led participants to
also identify acquaintances with which they are less close. Thus, it is likely that the nature of the
friend networks assessed here obscure these general effects, perhaps allowing the association of
density with other network aspects of friendship, such as length of friendship and frequency of
contact, to be more pronounced.
13
Table 13.3 Correlation matrix (N ¼ 277)

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
1. Negative emotions 1.00
2. Density 0.11 1.00
3. Identity discrepancy 0.26*** 0.08 1.00
4. Network size 0.04 0.06 0.01 1.00
5. Friendship length 0.41*** 0.22*** 0.06 0.03 1.00
6. Frequency of contact 0.15* 0.17** 0.15** 0.07 0.04 1.00
7. Emotional closeness 0.01 0.03 0.03 0.01 0.21*** 0.12 1.00
8. Age 0.31*** 0.01 0.18** 0.03 0.41*** 0.03 0.06 1.00
9. Woman 0.09 0.01 0.05 0.04 0.11 0.10 0.05 0.13* 1.00
10. Racial/ethnic minority 0.16** 0.12 0.10 0.02 0.10 0.11 0.05 0.18*** 0.06 1.00
11. Education 0.06 0.02 0.17** 0.01 0.13* 0.03 0.07 0.08 0.01 0.08
*p < .05
**p < .01
***p < .001
The Structure of Friendship: How Network Density Influences Identity. . .
277
278 K. L. Markowski
Table 13.4 Standardized coefficients for negative emotions (N ¼ 277)

Model 1 Model 2
(Baseline) (Model 1 + Interaction)
Independent variables β β
Density 0.01 0.01
Identity discrepancy 0.22*** 0.11
Density Identity discrepancy – 0.15*
Network size 0.01 0.02
Friendship length 0.36*** 0.36***
Frequency of contact 0.11 0.12*
Emotional closeness 0.08 0.08
Age 0.11 0.11
Woman 0.02 0.03
Racial/ethnic minority 0.08 0.07
Education 0.03 0.04
R2 0.26*** 0.28***
*p < .05
**p < .01
***p < .001
amounts of negative emotions. This is consistent with identity theory. However,

when the interaction term between density and identity discrepancy is included in
Model 2, only the interaction term is significant and negative (β ¼ .15). The
significant and negative interaction term indicates that the form of the relationship
between identity discrepancy and negative emotions is altered by higher compared
lower levels of density (Dawson, 2014).
Interaction Plot
To explore the nature of the interaction, I used graphing tools that followed guide-
lines on quadratic moderation (Miller et al., 2013). Prior to this, all main variables
except negative emotions were standardized (not shown here) such that all means
were 0, and the standard deviations were 1. Both the identity discrepancy items and
the density term were standardized before creating the interaction terms. The graphs
of the significant interaction is shown in Fig. 13.1.
In Fig. 13.1, values of identity discrepancy line the x-axis, and predicted values of
negative emotions line the y-axis. The moderator includes three values of density:
the average density score (represented by the solid line) as well as one standard
deviation above (represented by the dashed line) and below (represented by the
dotted line) the density average. Though these moderator points are referred to as
low, average, and high density, these should not be interpreted in absolute terms.
Instead, the average density score is just below a moderate level of density at .5 (.45),
3.5
Predicted Value of Negative Emotions
2.5
2
Low Density
1.5 Average Density
High Density
1
0.5
0
–2 –1.5 –1 –0.5 0 0.5 1 1.5 2
Value of Identity Discrepancy
Fig. 13.1 Interaction plot between density and identity discrepancy on negative emotions
and the low as well as high density scores are just above the lowest density cutoff
point at 0 (.06) and slightly below the highest density cutoff point at 1 (.84).
Shown in Fig. 13.1, there are varying associations between identity discrepancy
and negative emotions at each of the different levels of density. Greater degrees of
non-verification are associated with higher levels of negative emotions when density
is low compared to medium or high. Compared to the other two density moderator
values, the high density moderator shows the lowest and flattest curve on negative
emotions. In other words, high levels of density buffer the impact that increasing
non-verification has on negative emotions such that individuals who experience any
level of identity non-verification experience fewer negative emotions than if density
were lower. Importantly, this relationship appears to hold even when the number of
friends in the network, the frequency of contact with friends, and the average level of
emotional closeness to friends in the network are included as controls (alongside age,
gender, race/ethnicity, and education). This suggests a unique protective effect of
dense friendship networks that is not reducible to number, interaction frequency, or
intensity of friendship ties. This supports Hypothesis 1.
Discussion
The purpose of this paper was to examine the role of density in the identity
verification process. Scholars have suggested that the negative outcomes associated
with non-verification are contingent on other factors, such as how prominent or
salient an identity is for individuals (Burke, 1991; Burke & Stets, 2009; McCall &
Simmons, 1978). I suggested here that an overlooked factor that identity theorists
280 K. L. Markowski
need to consider in this relationship is social structure. I drew from social network
analysis to assess an alternative aspect of proximate social structure—specifically,
density, or the level of connectedness among the members in one’s network. I
expected that higher levels of density would buffer the negative emotions associated
with identity non-verification. Results supported these expectations in identity of
focus to this study: the friend identity.
This research is important because it adds context to the identity verification
process by suggesting the structural conditions under which non-verification leads to
different outcomes. The results here show that, at high levels of network density, the
relationship between identity discrepancy and negative emotions is very weak; high
levels of density buffer the negative emotions associated with non-verification, while
low density magnifies the effect on negative emotions. Future research should assess
if similar relationships are found among other identities, identity-based networks,
and outcomes. Research should also assess what such findings may mean for the
outcomes of the perceptual control process as it occurs over time. For example, does
the weak or potentially non-existent relationship between identity discrepancy and
negative outcomes for individuals within a highly dense network mean that, when
non-verification occurs, the individual is less motivated to modify their behavior or
self-views aimed toward bringing reflected appraisals in closer alignment with their
identity standard? Though the results in this study cannot directly speak to this
possibilty, they elicit an interesting issue that may serve as a point of inquiry for
future work on the perceptual control process.
This research is also important because it demonstrates the utility of social
network analysis for identity theory, particularly when conceptualizing and measur-
ing social structure. Social network analysis provides a wide range of tools by which
to describe the configuration of individuals’ social relationships. These tools provide
information about the properties associated with the bounded groups organized
around individuals and their particular identities. In identity theory, these small,
immediate social contexts in which individuals maintain social relationships and
enact identities are referred to as proximate social structures (Merolla et al., 2012;
Stryker et al., 2005). The advantage of employing social network analysis tools to
assess proximate social structure is that a wider range of information about group-
level properties can be directly examined above and beyond mere group membership
or notions of individuals’ commitments to specific others. This is important because
group-level properties influence how individuals interact with others and how
identity experiences unfold, meaning that these network-based concepts can thus
help identity theorists better understand identities and their processes.
What makes the network concept of density in particular a useful addition to
identity theory is that density provides insight into the relationships among one’s
social contacts. This is crucial social context to capture because it indexes the overall
dynamic of the network, which serves as the backdrop for relationships that exist
between the individual and their network members. It also serves as an indicator for
the interpersonal resources that derive from social relationships and thus are likely
available for individuals to use to offset the negative effects of non-verification
(Freese & Burke, 1994; Stets & Cast, 2007). For example, high levels of density may
have buffered negative emotions from non-verification in this analysis because dense
networks are often characterized by strong emotional bonds that confer social
support (Berkman et al., 2000; Friedkin, 2004; Green et al., 2001; Heaney & Israel,
2008; Kawachi & Berkman, 2001; Perry et al., 2018). Thus, it is likely that
individuals are able to draw social support from their dense networks to offset the
negative effects of non-verification. Another possibility is that dense networks imply
the existence of many mutually-verifying situations in which individuals have
bonded over the experience of continually verifying each other (Burke & Stets,
1999, 2009). This would suggest that individuals are able to minimize
non-verification from their network as a whole if they are also implicated in
mutually-verifying situations with specific members of the dense network.
Though social support and mutually-verifying situations are likely drivers of the
results found in this research, a limitation of this study is that information on social
support was not captured, nor was verification from each specific listed member of
the network. Thus, future work should assess if social support and/or mutually-
verifying situations drive the effects seen here. This can be done by incorporating
into the survey a more detailed density prompt, such as one that specifically asks
participants to indicate the supportive versus conflictual connections that exist
between their social contacts in addition to assessments of supportive exchanges
that have occurred between the individual and their social contacts. This can also be
done by including a reflected appraisal item for each network member listed in
addition to a reflected appraisal item that assesses the network as a whole. To best
reflect the many meanings that are included in the identity standard, a reflected
appraisal item should be included for each network member for each of the meanings
under examination.
Finally, while it may have been important to control for prominence and salience
in this analysis (since they were originally theorized as moderators for the identity
verification process), neither concept was available in the dataset. However, I did
include in the analysis the proportion of friends that the participant reported talking
to on at least a weekly basis. To the extent that this item is a rough proxy for salience,
post-hoc analyses (not shown here) showed no significant interactions between this
measure and the identity discrepancy item. This suggests that assessing social
structural features more directly as was done with density may be more useful
enterprise for research on contingencies in identity verification. Future research
can use this work as a starting point for better understanding the structural circum-
stances under which identity verification does or does not occur as expected.
References
Adams, R. E., & Serpe, R. T. (2020). Role discrepancies and psychological distress: a partial test of
incorporating identity theory and self-definitions into the stress process model. In R. T. Serpe,
R. Stryker, & B. Powell (Eds.), Identity and symbolic interaction: deepening foundations,
building bridges (pp. 293–316). Springer.
282 K. L. Markowski
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: conceptual strategic, and statistical considerations. Journal of Person-
ality and Social Psychology, 51(6), 1173–1182.
Berkman, L. F., Glass, T., Brissette, I., & Seeman, T. E. (2000). From social integration to health:
durkheim in the new millennium. Social Science & Medicine, 51, 843–857.
836–849.
Burke, P. J. (2006). Identity change. Social Psychology Quarterly, 69(1), 81–96.
Carter, W. C., & Feld, S. L. (2004). Principles relating social regard to size and density of personal
networks, with applications to stigma. Social Networks, 26(4), 323–329.
Cast, A. D. (2004). Well-being and the transition to parenthood: an identity theory approach.
Cast, A. D., Stets, J. E., & Burke, P. J. (1999). Does the self conform to the view of others? Social
Dawson, J. F. (2014). Moderation in management research: what, why, when, and how. Journal of
Business and Psychology, 29, 1–19.
Demir, M. (2010). Close relatnships and happiness among emerging adults. Journal of Happiness
Studies, 11, 293–313.
Demir, M., Ozdemir, M., & Weitekamp, L. A. (2006). Looking to happy tomorrows with friends:
best and close friendships as they predict happiness. Journal of Happiness Studies, 8, 243–271.
Ellestad, J., & Stets, J. E. (1998). Jealousy and parenting: predicting emotions from identity theory.
Fiori, K. L., Windsor, T. D., & Huxold, O. (2020). The increasing importance of friendship in later
life: understanding the role of sociohistorical context in social development. Gerontology, 66
(3), 286–294.
Freese, L., & Burke, P. J. (1994). Persons, identities, and social interaction. In B. Markovsky (Ed.),
Advances in group processes (Vol. 11, pp. 1–24). JAI Press.
Friedkin, N. E. (2004). Social cohesion. Annual Review of Sociology, 30, 409–425.
Gillespie, B. J., Lever, J., Frederick, D., & Royce, T. (2015). Close adult friendships, gender, and
the life cycle. Journal of Social and Personal Relationships, 32(6), 709–736.
Greco, S., Holmes, M., & McKenzie, J. (2015). Friendship and happiness from a sociological
perspective. In M. Demir (Ed.), Friendship and happiness: across the life-span and cultures
(p. 19). Springer.
Green, L. R., Richardson, D. S., Lago, T., & Schatten-Jones, E. C. (2001). Network correlates of
social and emotional loneliness in young and older adults. Personality and Social Psychology
Bulletin, 27(3), 281–288.
Hallinan, M. T., & Kubitschek, W. N. (1988). The effects of individual and structural characteristics
on intransitivity in social networks. Social Psychology Quarterly, 51(2), 81–92.
Heaney, C. A., & Israel, B. A. (2008). Social networks and social support. In K. Glanz, B. K. Rimer,
& K. Viswanath (Eds.), Health behavior and health education: theory, research, and practice
(4th ed., pp. 189–210). John Wiley & Sons, Inc..
Kalmijn, M. (2003). Shared friendship networks and the life course: an analysis of survey data on
married and cohabiting couples. Social Networks, 25(3), 231–249.
Kawachi, I., & Berkman, L. F. (2001). Social ties and mental health. Journal of Urban Health, 78
(3), 458–467.
Markowski, K. L., & Serpe, R. T. (2018). Identity theory paradigm integration: assessing the role of
prominence and salience in the verification and self-esteem relationship. In S. R. Thye & E. J.
Lawler (Eds.), Advances in group processes (Vol. 35, pp. 75–102). Emerald.
Marsden, P. V. (1990). Network data and measurement. Annual Review of Sociology, 16(1),
435–463.
Mendelson, M. J., & Aboud, F. E. (1999). Measuring friendship quality in late adolescents and
young adults: McGill friendship questionnaires. Canadian Journal of Behavioral Science, 31
(2), 130–132.
Merolla, D., Serpe, R. T., Stryker, S., & Schultz, W. (2012). Structural precursors to identity
processes: the role of proximate social structures. Social Psychology Quarterly, 75, 149–172.
Miller, J. W., Stromeyer, W. R., & Schwieterman, M. A. (2013). Extensions of the Johnson-
Neyman technique to linear models with curvilinear effects: derivations and analytical tools.
Multivariate Behavioral Research, 48(2), 267–300.
Perry, B. L., Pescosolido, B. A., & Borgatti, S. P. (2018). Egocentric network analysis: foundations,
methods, and models. Cambridge University Press.
Riley, A., & Burke, P. J. (1995). Identities and self-verification in the small group. Social
Serpe, R. T., & Stryker, S. (2011). Interactionist perspective and identity theory. In S. J. Schwartz,
K. Luyckx, & V. L. Vignoles (Eds.), Handbook of identity theory and research: Volume
1 structures and procedures (pp. 225–248). Springer.
Serpe, R. T., Yarrison, F. W., Stets, J. E., & Stryker, S. (2019). Multiple identities and self-esteem.
In J. E. Stets & R. T. Serpe (Eds.), Identities in everyday life (pp. 53–71). Oxford University
Press.
Sias, P. M., & Bartoo, H. (2007). Friendship, social support, and health. In L. L’Abate (Ed.), Low-
cost approaches to promote physical health and mental health: theory, research and practice
Siebert, D. C., Mutran, E. J., & Reitzes, D. C. (1999). Friendship and social support: the importance
of role identity to aging adults. Social Work, 44(6), 522–533.
Stets, J. E., & Burke, P. J. (2014a). Emotions and identity non-verification. Social Psychology
Quarterly, 77, 387–410.
Stets, J. E., & Burke, P. J. (2014b). Self-esteem and identities. Sociological Perspectives, 57(4),
409–433.
Stets, J. E., & Cast, A. D. (2007). Resources and identity verification from an identity theory
perspective. Sociological Perspectives, 50(4), 517–543.
Stets, J. E., & Harrod, M. M. (2004). Verification across multiple identities. Social Psychology
Quarterly, 67(2), 155–171.
symbolic interaction: deepening foundations, building bridges (pp. 65–88). Springer.
Stets, J. E., & Serpe, R. T. (2013). Identity theory. In J. Delamater & A. Ward (Eds.), Handbook of
Stryker, S. (1968). Identity salience and role performance: the relevance of symbolic interaction
theory for family research. Journal of Marriage and the Family, 30(4), 558–564.
Stryker, S. (1980). Symbolic interactionism: a social structural version. Blackburn Press.
Stryker, S., & Serpe, R. T. (1982). Commitment, identity salience, and role behavior: a theory and
research example. In W. Ickes & E. S. Knowles (Eds.), Personality, roles, and social behavior
Stryker, S., Serpe, R. T., & Hunt, M. O. (2005). Making good on a promise: the impact of larger
social structures on commitments. In S. R. Thye & E. J. Lawler (Eds.), Advances in group
processes (Vol. 23, pp. 93–124). Elsevier.
Taylor, S. E., Klein, L. C., Lewis, B. P., Gruenwald, T. L., Gurung, R. A. R., & Updegraff, J. A.
(2000). Female responses to stress: tend and befriend, not fight or flight. Psychological Review,
107(3), 411–429.
Turner, J. H., & Stets, J. E. (2005). The sociology of emotions. Cambridge University Press.
Walker, M. H. (2015). The contingent value of embeddedness: self-affirming social environments,
network density, and well-being. Society and Mental Health, 5(2), 128–144.
284 K. L. Markowski
Walker, M. H., & Lynn, F. B. (2013). The embedded self: a social networks approach to identity
Wasserman, S., & Faust, K. (1994). Social network analysis: methods and applications. Cambridge
University Press.
Yarrison, F. W. (2013). Normative vs. counter-normative identities: the structural identity model.
M.A. Thesis, Department of Sociology, Kent State University, Kent, OH.
Yarrison, F. W. (2016). Contextualizing proximate social structure in identity theory. In New
directions in identity theory and research (pp. 343–366). Oxford University Press.
Chapter 14
Using Identity Theory to Understand
Homophily in Groups
Jan E. Stets, John Aldecoa, Quinn Bloom, and Joel Winegar
Using Identity Theory to Understand Homophily in Groups
Individuals tend to form social ties based on shared characteristics, known as

“homophily” (McPherson et al., 2001). Homophily can be characterized as ties to
others based on sharing at least one attribute, such as age, race, social class, or values
(McPherson et al., 2001). Homophily forms the foundation of the social structure
(Blau, 1977), but it can constrain individuals’ opportunities for social interaction by
excluding diverse others.
Despite extensive research on homophily (Lazarsfeld & Merton, 1954;
McPherson et al., 2001), some issues are neglected. One oversight is an analysis
of social processes at the micro level that influence homophily. The social structure
or context may allow for the development of homophily (Mollenhorst et al., 2008).
For example, one’s school or work environment provides the opportunity to interact
with others who may share similar background characteristics, interests, or skills
(Stryker et al., 2005; Turner, 2016). However, there also is agency in the develop-
ment of homophily. While structuralists might dispute this claim (White, 1992),
individuals’ tastes, dispositions, and worldviews contribute to homophily (Lizardo,
2006; Vaisey & Lizardo, 2010). To this we add individuals’ identities.
We think that agency is important in homophily. Agency is individual action that
is oriented toward individually held goals. Goals motivate and give people direction
J. E. Stets (*) · Q. Bloom · J. Winegar

e-mail: [email protected]; [email protected]; [email protected]
J. Aldecoa
Department of Sociology, California State University, Sacramento, CA, USA
286 J. E. Stets et al.
to act. Individuals bind themselves to others who they see as sharing their charac-
teristics or attributes. However, rather than seeing these similarities in terms of a
shared internal worldview or moral orientation as has been argued (Vaisey &
Lizardo, 2010), we need to go beyond individual orientations—the psychological.
We also need to see that the individual and others in the situation play a role in
producing homophily. Further, it is not only individuals’ worldviews that are
relevant in homophily, but individuals’ identities as they interact with others, and
the degree to which each person confirms the identities of the other in the situation.
Thus, agency in homophily is about the individual and the other’s identities being
verified, and the ways in which they achieve this in situations.
Most interactions in situations are not between whole persons, but between
aspects of persons having to do with their identities. Identities are fundamental to
the individual and to interaction (Burke & Stets, 2009). There are various styles of
interaction that are appropriate in different situations for different identities. We
move into and out of these modalities very easily, and generally with very little
thought. For example, we may start our day by collaborating on a project with our
co-workers. After work, we may have a drink with a friend and share the latest
gossip. When we return home, we may discuss with our spouse a problem with one
of our children. Attached to the role of worker, friend, and spouse are meanings that
define oneself in each of these roles. These are one’s identities. In each case, there are
matters that are not discussed because they are not relevant to that identity, and there
are concerns that are more likely to be discussed given the identity that is being
claimed. Often, two or more identities may operate in a setting as in being a worker
and friend on the job or a spouse and parent at home.
Actors’ identities bind each person, one to the other, in a situation. Thus, what
helps foster homophilous groups is not only the sharing of similar positions in the
situation, such as being co-workers, friends, or spouses, but also the sharing of
identity meanings attached to those positions. As we will discuss, when actors’
identity meanings are verified in a situation, with each actor supporting the other’s
identity meanings, this verification should stabilize and reinforce homophily.
Homophily, in turn, should foster continued identity verification.
Because individuals claim multiple identities, we also argue that certain identities
may have a greater formative role in homophilous groups than other identities. Since
multiple identities are organized hierarchically within the self (Burke & Stets, 2009),
those identities that are higher in the hierarchy of identities (known as “master”
identities) may influence the meanings of other identities held by individuals and
may be more important in the formation of homophilous social ties. Master identities
are person identities or characteristics that people internalize that serve to define
them as unique persons such as being more (or less) kind or more (or less) hard-
working, for example. Person identities are often activated across situations. Given
their frequent activation and higher salience among a person’s multiple identities,
person identities are more likely to influence the meanings of one’s role identities
(self-definitions related to the roles people play out) and social identities (self-
definitions as a member of a category or group) (Burke & Stets, 2009). Since master
identity meanings are more common throughout an individual’s system of identity
14 Using Identity Theory to Understand Homophily in Groups 287
meanings, they are more available as the “glue” that binds them to others’ identity
meanings in groups.
More generally, we maintain that identity theory is a fruitful way to conceptualize
the micro processes that tie people’s identities, one to the other, in groups. We
outline the identity processes in what follows. We conclude by providing an example
of how this approach to homophily works in the world using a rather salient feature
of today’s American society: politics. Overall, we see that an identity perspective
advances our understanding of homophily in ways not yet demonstrated in the
literature.
Homophily
Homophily is the principle that similarity breeds connection. Network theory mea-
sures the extent to which homophily exists within a network of social relationships.
From a network approach, the structure is viewed as a complex web of social ties that
connect multiple individuals (“nodes”) (Katz et al., 2004). Nodes that share social
ties are made up of attributes that allow for the existence of shared qualities or
homophily. In terms of homophily, we need to distinguish between network struc-
ture and network composition. Network structure refers to the patterns of relation-
ships that exist among individual nodes, whereas network composition pertains to
the attributes of each individual node, such as a person’s race, gender, or religion
(Frank, 1997). We focus on the compositional features of the social network or social
group(s) to which people identify, and the conditions under which homophily
emerges.
People are more likely to interact with others with whom they share similarities
with respect to status characteristics like gender, race, or class (McPherson et al.,
2001; Smith et al., 2014), and cultural orientations such as morals, values, beliefs,
and attitudes (Lazarsfeld & Merton, 1954; Vaisey & Lizardo, 2010). This is the
distinction between status homophily and value homophily, respectively (Lazarsfeld
& Merton, 1954; McPherson et al., 2001). Status homophily refers to sharing some
status characteristic. It is rooted in structural forces that stratify individuals into
social positions based on ascription such as gender, race, and class, or achievement,
such as educational attainment or religious affiliation. Status homophily is present
among children who prefer friendships with peers of their own gender or racial
category (Shrum et al., 1988), and among high school students who prefer to bond
with others who have similar levels of academic achievement (Smirnov & Thurner,
2017).
Value homophily refers to sharing some cultural orientation such as morals,
values, ideologies, and attitudes. We add to this, identities. While identities are a
response to who one is, identity meanings are rooted in culture, and we are socialized
into culture. This is why we can share identity meanings with others. Because others
are born into the same culture as us, they may define themselves similarly to how we
define ourselves with respect to different identities.
Value homophily is an important factor in choosing friends (Lazarsfeld &

Merton, 1954), intimate partners (Bahns et al., 2016), hiring potential job candidates
(Rivera, 2012), and electing political representatives (Golder & Stramski, 2010).
Close friends often share values and worldviews (Parkinson et al., 2017; Vaisey &
Lizardo, 2010), beliefs, attitudes, and opinions (Byrne, 1969), and religious and
political ideologies (Cheadle & Schwadel, 2012; Poteat et al., 2011).
Value homophily also is evident when persons prefer to live near ideologically
similar others (Motyl et al., 2014), to watch ideologically consistent news networks,
and to cut ties with ideologically dissimilar others (Mitchell et al., 2014). It is
suggested that social relationships based on value homophily may form stronger
and longer-lasting bonds than those based on status homophily (Lizardo, 2006;
Vaisey & Lizardo, 2010).
Given both status and value homophily, we can say that homophily develops
from both the influence of the social structure and individual selection. The social
structure sets the opportunities to interact with similar others, thus fostering status
homophily. For example, we are born into a family that shares a similar background.
We enter school with those who share our age and perhaps our social class. Thus, the
social composition of the contexts we enter help set the initial conditions for
homophily (Mollenhorst et al., 2008). A certain amount of similarity is important
as it lowers the costs associated with interacting with others who may be very
different from us and thus not support us; and similarity helps foster trust (Homans,
1950). However, homophily also emerges through individual selection. Beyond
sharing status characteristics, people may further select others with whom to rou-
tinely interact within and across situations based on shared identity meanings.
Homophily is important for individual functioning because it satisfies many
individual social and psychological needs. For example, interacting with similar
others can provide people with social support that satisfies the need for belonging
and feeling secure within themselves (Hogg, 2007; Giddens, 1991; Suitor & Keeton,
1997; Wright, 2012). In turn, social support can raise self-esteem (Levendosky et al.,
2004). Homophily also has benefits for people’s health and emotional well-being by
reducing stress and depression and enhancing the endocrine and immune systems
(Sani, 2012). These health benefits reduce mortality rates and increase the quality of
life. To begin to examine how identities play a role in homophily, we turn to identity
theory.
Identity Theory
In identity theory, an identity is a set of meanings that describes individuals as a

member of a social category (social identity) or group (group identity), as a role
player (role identity), or as a unique person (person identity) (Burke & Stets, 2009).
Meaning is central to understanding identities, situations, and homophily. Meaning
is a mediational response to a stimulus; it mediates between an initial perception of a
stimulus and a response to the stimulus (Burke & Stets, 2009; Osgood et al., 1957).
When the self is the stimulus, as when we take ourselves as an object and think of
ourselves in terms of our identities, meanings then characterize the self. They are
descriptors, such as adjectives, which represent a part of the self-view or one’s
identity standard that is set by the individual. For example, a person may describe
herself as “independent” in her gender social identity, “trustworthy” in her friend
role identity, and “persuasive” in her person identity of being dominant.
When the situation rather than the self is the stimulus, as when we enter settings
and assess how we are expected to behave, meanings also are activated. For
example, when we enter a church, we are expected to be “reverent.” When we
enter a party, we are expected to be “friendly.” What is key in understanding
meanings in a situation is that they influence the identity that is activated. The
identity becomes relevant in the situation because there is a correspondence between
identity meanings and situational meanings. For example, upon entering a church,
the identity that shares the meaning of “reverence” should be activated, that is, the
religious social identity. Upon entering a party, expectations of being “friendly”
should overlap with the meaning tied to the “sociable” person identity; thus, this
identity should be activated.
Which identity is enacted in a situation also is a function of how salient an identity
is for individuals compared to all the identities they may claim (Stryker, 2002
[1980]). Identity salience is the readiness to enact an identity in a situation, with
more salient identities more likely to be brought into situations. Thus, persons
invoke some identities more than others, although some situations provide little
opportunity to choose, and thus are more “closed” than “open” to choice (Serpe,
1987). For example, one’s work environment may provide little opportunity to enact
a salient parent identity given the cultural view that work and home life are separate
spheres.
The prominence or importance of an identity has been shown to influence the
salience of the identity in situations (Brenner et al., 2014). Essentially, those
identities that are more important are more likely to be performed in situations.
However, this relationship is not absolute. There are situations in which a highly
prominent identity may not be enacted because the situation is “closed” to the
performance of that identity. Conversely, a salient identity may be played out in a
situation but not highly valued because the situation requires its enactment (Thoits,
2020). For example, a manager may be compelled to enact her dominant identity at
work even though she does not highly value being a dominant person. Overall, the
evidence suggests that what lies behind salient identities are identities that are
important or valued to individuals, but more work is needed.
Given that there is correspondence between identity meanings and situational
meanings, there also is correspondence between identity meanings and meanings
implied by one’s behavior so that individuals can fulfill situational meanings. Thus,
the identity standard and behavior are linked through a common system of meaning
as well (Burke, 1980; Burke & Reitzes, 1981). Perhaps most importantly, persons
are motivated to verify who they are in situations, and they will do so by acting to
control meanings in the situation so that their identity meanings align with the
meanings in the situation. Identity theorists conceptualize the identity verification
process in terms of a perceptual control system (Burke, 1991). The perceptual

control system operates as a negative feedback loop that controls perceptions in
situations to match perceptions held in one’s identities.
The perceptual control system has five central aspects: (a) the identity standard;
(b) output; (c) input; (d) the comparator, and (e) emotions. The identity standard is
the set of meanings that define who the person is in an identity, and it is activated
when the person perceives meanings in a situation as relevant to an identity. The
perceived situational meanings (input) include perceptions of self in the situation,
perceptions of how one thinks others see him/her in the situation, signs, and symbols
in the situation which give off identity relevant expectations of how one is to behave,
and the meanings implied by the behaviors of others. Each one of these meanings is
compared to the identity standard meanings by the comparator.
A perceived “match” between identity meanings and meanings from others in the
situation indicates that an identity has been verified. Persons will then feel positive
emotions, and behavior (output) will continue, uninterrupted (Burke & Stets, 2009).
A perceived “mismatch” will alert individuals that their identity has not been
verified, which arouses negative emotions that motivate adjustments by modifying
behavior (output) and slowly changing identity standard meanings to better “match”
situational meanings. Identity non-verification can exist because evaluations by
others exceed an identity standard, or because evaluations by others fall short of
an identity standard. In either case, small discrepancies are resolved in a relatively
automatic fashion through learned patterns of controlling the situational meanings.
Larger discrepancies prompt more intense negative emotions, and a more delibera-
tive process of adjustments, until the disturbance causing the non-verification is
resolved (Stets, 2016).
We maintain that identity verification is an important mechanism by which
homophily is accomplished in social groups. As a person attempts to verify an
identity in a situation, others in the situation are also attempting to verify their
identities (Burke & Stets, 1999). Identity verification will arouse positive emotions
that will increase interaction and foster homophilous groups. As we discuss below,
verification in these groups will involve individuals confirming identities that are
like their own, or confirming different, but complementary identities. If people are
not verified, they are more likely to cease interaction with those others, thereby
reducing the likelihood of a homophilous group forming with those others. They will
likely seek out others who will verify them, and if verified, forge a homophilous
group. Identity verification also may make more prominent and salient the identity
that is activated in the situation. As each verifies the identity of the other in the
situation, that identity may be made more prominent as a source of verification and
made more salient in the future. Thus, we offer the following initial hypotheses:
H 1: Identity verification will increase homophily in social groups.
H 2: Identity verification in homophilous groups will strengthen the prominence
and salience of identities.
Identity and Homophily
In identity theory terms, individual selection or agency refers to a person’s prefer-

ence to interact with others who share meanings for an identity. For this to occur, the
social structural context needs to allow individuals to sort through others in the
setting who may share meanings with them (Mollenhorst et al., 2008). This sorting
occurs at primarily the intermediate (meso) and proximate (micro) social structural
levels (Stryker et al., 2005; Turner, 2016).
The intermediate or meso social structural level is the groups, organizations, and
communities that individuals place themselves into, and the meso social structure is
embedded in the larger macro social structure of institutions, the stratification
system, and the broader culture (Stryker et al., 2005; Turner, 2016). The meso
level is the arena that individuals routinely frequent such as their workplace, school,
church, and neighborhood. These structures bring many people together in one
setting, allowing interaction to take place and relationships to develop. They provide
important social boundaries that influence the likelihood of shared identities to exist
(Merolla et al., 2012; Stets & Serpe, 2013; Stryker et al., 2005; Yarrison, 2016).
They initially may be organized around status characteristics such as education
(in the workplace), age (in a school), religion (in a church), and social class (in a
neighborhood).
The proximal or micro social structural level is the closest arena to the individual
that provides the resources and opportunities to claim and enact identities. These are
the daily interactions or encounters of individuals (Turner, 2016). It captures more
intimate relationships, such as those with family and friends. While closer relation-
ships at the proximal or micro level can be characterized by individuals who also are
likely to be similar on status characteristics such as race, age, education, and social
class, there is a further sorting of individuals to create stronger, closer bonds.
Within intermediate/meso and proximal/micro social structures, we suggest that,
while initial connections to others are based on one or more status characteristics,
individuals will go through a further sorting process and select others to bond with
based on shared identities. By shared identities, we mean that individuals will share
with others the same dimension of meaning and the same location on that dimension
of meaning. For example, recent research studied the degree to which individuals, in
the student identity, perceived that other students who were their close friends shared
the same dimension of meaning on the student identity such as being “motivated,”
“hardworking,” and “responsible” (Grindal & Trettevik, 2019). Further, people’s
location on each of these dimensions of meanings likely needs to be the same as in
being high, middle, or low on each dimension. Finally, people may share the same
dimensions and locations of meanings across multiple identities. For example,
individuals might be very “responsible” and very “hardworking” in the student
identity, worker identity, and parent identity.
In interactions, people learn the meanings that characterize each other’s identities.
When individuals share identity meanings, identify verification is more likely to
emerge because the responses of each are not only in the service of their own
identities, but also in the service of others’ identities. For example, if a person (P) and
other (O) agree that being a student means being very “responsible” and very
“hardworking,” and both have an upcoming exam, P may choose to study rather
than go to a party, to which O also is invited. This provides the opportunity for P to
experience identity verification because the meaning of P’s behavior is consistent
with the meanings in P’s identity standard. P’s decision to study, and even inviting O
to study with P rather than attend the party, may remind O of O’s own meanings of
being very “responsible” and “hardworking.” P’s invitation calls forth the opportu-
nity for O to also study for the exam, and thus verify O’s student identity as well.
Both P and O’s decision to study rather than attend the party, fosters identity
verification by way of direct, self-appraisals. That is, P and O monitor the meanings
of their own behavior and determine whether their behavior is consistent with their
identity standard. Thus, identity verification can occur not only indirectly, through
how people think others see them in an identity (reflected appraisals), but also
directly, through self-appraisals. If P routinely declined to attend social gatherings
when there was pressing schoolwork, and O routinely chose a social life over
schoolwork, over time, the tie between P and O likely would loosen unless they
were to find other meanings that they shared across other identities such as the friend
identity or worker identity.
Identity verification can create a social tie in the absence of a shared location of
meaning. This occurs when individuals assume complementary identities, for exam-
ple, a teacher-student or parent-child relationship is characterized by each having
different, yet corresponding meanings that are attached to the different roles in the
relationship. The differences allow the identities to be tied in a collaborative manner.
For example, the teacher identity cannot exist without a corresponding student
identity to which it is oriented. One cannot demonstrate the meaning of being very
“knowledgeable” by lecturing to students if there are no students present to receive
the lecture and learn from the teacher’s knowledge. In complementary identities,
what is important is the accepted position of each partner on the shared location of
meaning. This, too, can produce homophily. Thus, our next hypothesis is:
H 3: Homophily will emerge when identity meanings are on the same
(or complementary) location of meaning.
Identity verification may not activate a social tie initially, even if there is a shared
location of meaning. Identity verification may have to occur several times before a
social tie begins to emerge. This is because the initial stages of a relationship show
low levels of trust, and there is uncertainty because individuals do not know in
advance what are the intentions of others, and whether their intentions are benevo-
lent. As the frequency of interactions increases and identity verification occurs,
uncertainty subsides, trust increases, and a social bond between individuals emerges
(Burke & Stets, 1999; Stets et al., 2018).
If a social tie emerges, it is possible that identity verification is unilateral, that is, P
may verify O, but O may not verify P. Such social ties are not uncommon. We can
think of relationships in which P is dependent upon O for resources such that P
verifies O’s identity, without receiving identity verification in return. Such a
relationship is unlikely to be as deep or long-lasting than if each was verifying the

other’s identity, that is, if mutual verification occurred. Mutual verification can occur
between individuals who do not share a location of meaning as when the meaning of
the partner identity for P may mean being “dominant” and “gregarious” while for O
the meaning of the partner identity may mean being “submissive” and “shy,” yet a
relationship is still forged because the meanings needed to verify each identity are
produced by the other.
If individuals develop a social tie based on their location of shared meaning, it
may be that, over time, individuals will explore other identity meaning locations that
they share. Individuals should form a stronger social bond the more they share
identity meanings, as these shared meanings present additional opportunities for
homophily. Again, people can share the same identity dimensions and locations of
meanings (on those dimensions) across multiple identities. Indeed, the more one has
in common with another, the closer the two individuals should feel. The additional
identities that are explored may be about other person, role, or social/group identities
that individuals claim. For example, while P and O may share the meaning of being
very “hardworking” in the student identity, worker identity, and parent identity, they
also may discover that they share meanings of being very “caring” in the moral
person identity, very “loyal” in the friend role identity, and not very “devout” in the
religious social identity. Here, the dimensions of meaning have changed across
identities, but P and O’s location on these different dimensions is the same.
The discovery of other shared identities may be guided by the prominence and
salience of one’s own identities. People may select on those identities that they value
or frequently activate as a way of helping them assess what further they have in
common with others. The prominence and salience of people’s identities may
provide a “short cut” as to what to select on when they may not have the time to
carefully sift through additional identities to explore.
In general, we anticipate that the more locations of meaning that individuals have
in common across multiple identities, the stronger should be the homophilous ties.
This is because all individuals have more opportunities for their identities to be
verified. Verification should lead to more positive feelings, increased liking and trust
for similar others, and greater feelings of unity with similar others (Burke & Stets,
1999; Stets et al., 2018). These outcomes should help maintain the relationship for
continued and future verification of all individuals. Individuals should experience a
greater sense of “we-ness” (Burke & Stets, 1999), identifying with their
homophilous social ties and dis-identifying with non-homophilous ties. Over time,
this may increase the social distance between similar others and dissimilar others, or
the in-group and the out-group. An “us” vs. “them” orientation may lead to a
polarization of social ties and the formation of an echo chamber wherein the identity
meanings become local, held among those within the homophilous group, and
through the verification process, the meanings are routinely reinforced and solidified
within the group. Therefore:
H 4: The more individuals share multiple locations of identity meanings across
multiple identities, the stronger will be the homophilous ties.
If individuals do not share locations of meanings on multiple identities, two

things may happen. Either the social tie is broken because there are too few meanings
in common to bind the individuals in the long run, or dissimilar locations in identity
meanings between individuals become similar over time. In the latter case, identity
standard meanings are slowly shifting in the direction of the meanings that are
attributed to the self by others in the situation (Cast & Cantwell, 2007; Cast et al.,
1999; Stets et al., 2020). However, while identity standard meanings are slowing
changing in the direction of reflected appraisals, people also are acting to change the
reflected appraisals (Burke, 2006; Savage et al., 2017; Stets et al., 2020). In this way,
people are resisting the influence of others’ meanings, hoping to change these
meanings in the direction they desire, while at the same time, they are shifting in
the direction of others’ meanings. Thus, both processes are going on over time. More
generally, these dynamics are known as identity change. Identity change is always
occurring within individuals. It is often very gradual with people unaware that the
change is taking place. For example, in one study on changes in the meanings of
people’s spousal identity, research revealed that lower status spouses changed their
identity meanings in the direction of the higher status spouse, but this change
occurred slowly over a period of one to three years (Cast et al., 1999).
Person Identities and Homophily
Since developing social groups involves sampling on other’s identities and the
meanings associated with these identities, multiple identities are very much involved
in homophily. In considering one person in the situation who has multiple identities
contained within herself, the behavior and perceptions of the individual must meet
the goals of several identities at once. A person must adjust the perceived meanings
so that multiple identities can be verified simultaneously. When individuals forge
ties with others across shared locations of meaning involving multiple identities, this
fosters the very verification that a person needs. For example, two people may share
the overlapping location of meaning of being very “caring” in the parent identity and
friend identity. Thus, controlling the meanings in one identity will simultaneously
control the meanings in the other identity, and verifying one identity also will verify
the other identity. It is when the meanings across identities are along different
dimensions or in different locations along those different dimensions that it is a
problem because controlling the meanings in one identity may prevent the control of
meanings in another identity, causing conflict and perhaps the abandonment of one
identity over another. But, how are identity meanings related to each other within the
individual, since this has implications for how they would be related to each other
within social groups?
Within the self, identities are arranged within a hierarchical control system that
consists of multiple levels, where the outputs of higher levels are the standard of
lower levels. Among identities, we distinguish two important perceptual levels of
control: control at the principle level and control at the program level (Burke & Stets,
2009; Tsushima & Burke, 1999). At the higher, principle level are more abstract
standards and meanings, such as the meanings associated with one’s identities. At
the lower, program level are more concrete standards and meanings associated with
specific goals to be accomplished in the immediate situation that are in the service of
one’s identities. Principle level meanings guide program level meanings to ensure
verification. For example, a parent who values education and the meaning of being
academically responsible (a principle level standard) will make sure that her child is
at school on time, completes his homework, and adequately prepares for exams
(program level standards) (Tsushima & Burke, 1999).
At the principle level, some identities may be conceptualized as ranked higher
than others. As higher ranked identities, they are “master identities” in that they
influence other identities within and across situations. Given that there are different
bases of an identity: social/group, role, and person identities, person identities are
presumed to be higher in the hierarchical control system of identities. Given their
higher ranking, they are key to defining social/group identities and role identities. By
extension, person identities become central in the development of homophilous
groups.
For example, the moral person identity, that is, how individuals define themselves
in terms of being moral (Stets, 2010; Stets & Carter, 2006, 2011, 2012) is tied to
individuals rather than to individuals’ roles and groups. The meanings of being
moral likely are activated within and across social/groups, roles, and social situa-
tions. For instance, if individuals see themselves as high on the “care” and “justice”
dimensions of being moral, they may choose roles and join groups that reflect these
meanings. They may take on the role identity of Peace Corps volunteer, trial lawyer,
or social worker, and they may assume social/group identities such as membership in
a local church or city council. In turn, the meanings of these role and social/group
identities may, over time, reinforce the meanings of the moral person identity
(Aldecoa, 2019).
Given that master identities such as the moral identity have influence in defining
the nature of role and social/group identities, master identities should be more
influential in the formation of social group ties than other identity meanings. This
builds upon a growing body of research that demonstrates the importance of value
homophily as an organizing feature of social networks. Specifically, research reveals
that sources of value homophily such as one’s moral orientation may be more
influential in shaping social network ties than sources of status homophily, such as
those found with social/group (for example, age, race, class) and role identities (for
example, spouse, worker, or student) (Vaisey & Lizardo, 2010). Therefore, our final
hypothesis is:
H 5: Person identities will have a stronger influence on homophily than role and
group identities.
One type of moral similarity that appears to particularly drive homophily is moral
purity (Dehghani et al., 2016). Physical and spiritual purity is linked to concerns
about physical contamination, avoidance, and feelings of disgust. In turn, such
reactions amplify perceptions of moral wrongness and create social distance from
others. Moral purity, as Dehghani and his associates argue, may be frequently used
as the basis for religious and political division in our society. To develop this idea
more fully, we apply identity theory to homophily in groups by discussing how
moral meanings may be an important basis of partisan politics in the United States.
Essentially, while some may see one’s political identity to be an important basis of
homophily, what may be a more important basis of homophily is one’s moral
identity.
Applying Identity Theory to Homophily: Morality and Politics
Partisan politics exists in American society. How Democrats and Republicans vote
in elections largely can be explained by their liberal and conservative ideologies
(Jost, 2006; Jost et al., 2008; Pew Research, 2017). Thus, one’s political identity is
intimately tied to their political ideology. Further, there is a high degree of homo-
geneity among the political views of individuals who share a similar political identity
(Boutyline & Vaisey, 2017; Brandt et al., 2019). However, people’s morals may be a
more fundamental force that drives the political divide (Graham et al., 2009; Haidt,
2012; Hunter, 1991; Miles & Vaisey, 2015).
In early research, it was argued that one’s moral orientation was related to one’s
political identification (Haidt & Joseph, 2004). Liberals tended to rely on individu-
alizing morals, that is, morals that focused on promoting/protecting the individual’s
welfare such as “care/harm” and “fairness/cheating,” while conservatives tended to
rely on collectivizing morals, that is, morals that promoted/protected the group such
as “loyalty/betrayal,” “authority/subversion,” and “sanctity/degradation” (Haidt &
Joseph, 2004). In later research, it was argued that while the concern for care,
oppression, and fairness was the liberal agenda, issues of care, oppression, and
fairness alongside issues of loyalty, authority, and sanctity were the conservative
agenda (Haidt, 2012). Thus, as Haidt argued, conservatives showed a wider array of
“moral tastes” than liberals.
Haidt argued that when political issues become moralized, people will “bind”
together into political groups along their perceived moral similarities. Political
groups become galvanized along a shared morality and act according to the goals
that they view as morally good. Differing goals among political groups can eventu-
ally clash, creating moral conflict. For example, moralized political issues such as
abortion that raises issues of individual rights including life itself; immigration that
introduces concerns of social welfare and individual protection; and economic and
racial inequality that presents problems related to discrimination and fairness places
individuals into different political groups.
People are motivated to “choose sides” by not affiliating with persons of the
opposing side, and instead, associate only with those who have similar morals
(Dehghani et al., 2016; DeScioli & Kurzban, 2012). Indeed, persons may choose
to “de-friend” those who do not share similar moral or political views (Schwarz &
Shani, 2016), and even decide to live in geographical regions in which the commu-
nity ideologies are consistent with their own (Motyl et al., 2014). An “echo cham-
ber” emerges in which only certain ideas and beliefs are shared (and others are
omitted), and this is maintained through selective media consumption that only
reinforces the chamber, allowing for increasing political polarization (Jamieson &
Cappella, 2008). However, some have argued that partisan echo chambers are
overstated, given the availability of diverse content and perspectives in different
media outlets (Dubois & Blank, 2018).
An identity theory approach can enhance the research on political polarization by
conceptualizing the moral person identity as that which influences individuals’
political identity. As discussed earlier, person identities like the moral identity are
master identities whose meanings impact other identity meanings that individuals
claim, such as the political (social) identity. As a master identity, the moral identity is
more likely to influence the formation of social group ties than other identities
(Hypothesis 5). Thus, the moral person identity becomes an important guide for
the development and participation in homophilous political groups.
Recall that homophily will emerge when identity meanings are on similar loca-
tions of meaning (Hypothesis 3), and that an important identity mechanism that
facilitates homophily is identity verification (Hypothesis 1). When persons identify
with the moral meaning of “purity/sanctity,” for example, they may find others who
similarly identify with this moral meaning either through social structural opportu-
nities in proximate (micro) or immediate (meso) social contexts, or through a sorting
process of selecting individuals in those contexts. Upon finding similar others, this
will create the conditions for identity verification in which each supports expressions
of “sacredness” in the other. As mutual verification is repeated over space and time,
it should influence the formation of social ties among individuals. The verification
also should reinforce the moral identity, strengthening its value and likelihood of
being activated in other situations. Thus, the prominence and salience of the moral
identity should strengthen over time in the homophilous group (Hypothesis 2).
Individuals may explore other identity meanings that they share with
homophilous others. The more individuals share multiple identity meanings, the
stronger will be those homophilous ties (Hypothesis 4). They may use their prom-
inent and salient identities as “short cuts” as to what other identities to explore. If
membership in a political group is a prominent and/or salient identity, and it likely is
for many people, individuals may find that they share with those who value
“sacredness,” the political meanings associated with being Republican. For example,
when political issues emerge that are high in moral content such as issues about
sexuality, those who embrace the moral meaning of the “sacred” may be strongly
opposed to a gay lifestyle and will thus support Republican policies such as
opposition to same-sex marriage (Koleva et al., 2012). Importantly, what links the
moral person identity and political social identity are meanings that are held in
common.
The moral conflict that arises within political issues such as the abortion issue can
increase the potential for receiving, from political opponents, non-verifying feed-
back for one’s moral identity. As a result, individuals will be more inclined to “seek
out” opportunity structures for moral identity verification among political allies to
increase the likelihood of verifying their moral identity within the political context
(DeScioli & Kurzban, 2012). As the moral identity of individuals is verified within
the political context, in turn, they can provide verifying feedback of their own to
political allies. This mutual verification can strengthen the social bond among
individuals, fostering a sense of “we-ness.
The risk in homophily is the creation of echo chambers, social distance from
dissimilar others, a consolidation of views among in-group members, and further
polarization between in-group and out-group members. Intergroup contact can help
temper stereotypes and biases that produce discrimination and other social injustices.
Thus, we need to be careful. In our desire to obtain identity verification at the
individual level, which fosters homophily, we may be encouraging intergroup
conflict and unrest at the societal level.
Summary
We have used identity theory to explain homophily. We have argued that identity
meanings that are on the same (or complementary) locations of meaning for indi-
viduals mark the emergence of homophily. And, the more individuals share multiple
locations of identity meanings, the stronger should be the homophily. What we have
maintained is that what is key in shared meanings creating homophily is identity
verification. Verification arouses positive emotions that will increase interaction and
should foster homophily. Identity verification also should strengthen the prominence
and salience of identities in homophilous groups. As each verifies the identity of the
other in a situation, it reinforces the identity, strengthening the value of the identity
and the likelihood of being activated in other situations. We also have argued that
person identities are more likely to be influential features of homophilous groups
than other identities because they serve as master identities that help shape and
reinforce role identities and social/group identities that individuals claim. Because
person identities are central to the formation of other identities, they should be
central in the development of homophilous groups.
We used the moral identity within the context of partisan politics to contextualize
how person identities shape social groups. Homophily can provide a site for identity
verification. Identity verification activates positive feelings, builds trust and certainty
among individuals, and generates a feeling of “we-ness” (Burke & Stets, 1999; Stets
et al., 2018). On the other hand, homophily can create echo chambers in which
certain identity meanings are reinforced in a relatively closed system that does not
allow alternative or competing meanings to be discussed (McPherson, 2004). This
can fuel ethnocentrism, intergroup conflict, divisiveness, and sharpen emotions and
rhetoric of the out-group as “bad” and the in-group as “good.” If people act on this
divisiveness, it can lead to injury and unrest. This may help us understand some of
American politics today.
Acknowledgements We thank members of the social psychology seminar at the University of

California, Riverside, for the comments on an earlier version of this paper.
References
Aldecoa, J. (2019). Toward a theory of moral identity development. In J. E. Stets & R. T. Serpe
(Eds.), Identities in everyday life (pp. 115–135). Oxford University Press.
Bahns, A. J., Crandall, C. S., Gillath, O., & Preacher, K. J. (2016). Similarity in relationships as
niche construction: Choice, stability, and influence within dyads in a free choice environment.
Journal of Personality and Social Psychology, 112(2), 329–355.
Blau, P. (1977). Inequality and heterogeneity: A primitive theory of social structure. Free Press.
Boutyline, A., & Vaisey, S. (2017). Belief network analysis: A relational approach to understanding
the structure of attitudes. American Journal of Sociology, 122(5), 1371–1447.
Brandt, M. J., Sibley, C. G., & Osborne, D. (2019). What is central to belief system networks?
Personality and Social Psychology Bulletin, 45(9), 1352–1364.
836–849.
Burke, P. J. (2006). Identity change. Social Psychology Quarterly, 69(1), 81–96.
Byrne, D. (1969). Attitudes and attraction. In L. Berkowitz (Ed.), Advances in experimental social
psychology (pp. 35–89). Academic Press.
Cast, A. D., & Cantwell, A. M. (2007). Identity change in newly married couples: Effects of
positive and negative feedback. Social Psychology Quarterly, 70(2), 172–185.
Cheadle, J. E., & Schwadel, P. (2012). The ‘friendship dynamics of religion,’ or the ‘religious
dynamics of friendship’? A social network analysis of adolescents who attend small schools.
Social Science Research, 41(5), 1198–1212.
Dehghani, M., Johnson, K., Hoover, J., Sagi, E., Garten, J., Parmar, N. J., Vaisey, S., Iliev, R., &
Graham, J. (2016). Purity homophily in social networks. Journal of Experimental Psychology:
General, 145(3), 366.
DeScioli, P., & Kurzban, R. (2012). A solution to the mysteries of morality. Psychological Bulletin,
139(2), 477–496.
Dubois, E., & Blank, G. (2018). The echo chamber is overstated: The moderating effect of political
interest and diverse media. Information, Communication & Society, 21(5), 729–745.
Frank. (1997). Composition and structure in social networks. Mathématiques et Sciences Humaines,
137, 11–23.
Giddens, A. (1991). Modernity and self-identity: Self and society in the late modern age. Stanford
University Press.
Golder, M., & Stramski, J. (2010). Ideological congruence and electoral institutions. American
Journal of Political Science, 54(1), 90–106.
Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and conservatives rely on different sets of
moral foundations. Journal of Personality and Social Psychology, 95(5), 1029–1046.
Grindal, M., & Trettevik, R. (2019). Perceived similarity in identity meanings, identity verification,
and positive emotions. In J. E. Stets & R. T. Serpe (Eds.), Identities in everyday life (pp. 35–52).
Haidt, J. (2012). The righteous mind: Why good people are divided by politics and religion. Vintage
Books.
Haidt, J., & Joseph, C. (2004). Intuitive ethics: How innately prepared intuitions generate culturally
variable virtues. Daedalus, 133(4), 55–66.
Hogg, M. A. (2007). Uncertainty-identity theory. In M. P. Zanna (Ed.), Advances in experimental
social psychology (pp. 69–126). Elsevier.
Homans, G. C. (1950). The human group. Harpers.
Hunter, J. D. (1991). Culture wars: The struggle to control the family, art, education, law, and
politics in America. Basic Books.
Jamieson, K. H., & Cappella, J. N. (2008). Echo chamber: Rush Limbaugh and the conservative
media establishment. Oxford University Press.
Jost, J. T. (2006). The end of the end of ideology. American Psychology, 61(7), 651–670.
Jost, J. T., Nosek, B. A., & Gosling, S. D. (2008). Ideology: Its resurgence in social, personality,
and political psychology. Perspectives on Psychological Science, 3(2), 126–136.
Katz, N., Lazer, D., Arrow, H., & Contractor, N. (2004). Network theory and small groups. Small
Group Research, 35(3), 307–332.
Koleva, S. P., Graham, J., Iyer, R., Ditto, P. H., & Haidt, J. (2012). Tracing the threads: How five
moral concerns (especially purity) help explain culture war attitudes. Journal of Research in
Personality, 46(2), 184–194.
Lazarsfeld, P., & Merton, R. K. (1954). Friendship as a social process: A substantive and
methodological analysis. Freedom and Control in Modern Society, 18(1), 18–66.
Levendosky, A. A., Anne Bogat, G., Theran, S. A., Trotter, J. S., von Eye, A., & Davidson, W. S.,
II. (2004). The social networks of women experiencing domestic violence. American Journal of
Community Psychology, 34(1–2), 95–109.
Lizardo, O. (2006). How cultural tastes shape personal networks. American Sociological Review, 71
(5), 778–807.
McPherson, M. (2004). A Blau space primer: Prolegomenon to an ecology of affiliation. Industrial
and Corporate Change, 13(1), 263–280.
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily in social
networks. Annual Review of Sociology, 27, 415–444.
149–172.
Miles, A., & Vaisey, S. (2015). Morality and politics: Comparing alternative theories. Social
Science Research, 53, 252–269.
Mitchell, A., Gottfried, J., Kiley, J., & Matsa, K. E. (2014). Political polarization and media habits.
Pew Research Center. Retrieved August 9, 2018, from http://www.journalism.org/2014/10/21/
political-polarization-media-habits/
Mollenhorst, G., Völker, B., & Flap, H. (2008). Social contexts and personal relationships: The
effect of meeting opportunities on similarity for relationships of different strength. Social
Networks, 30(1), 60–68.
Motyl, M., Iyer, R., Oishi, S., Trawalter, S., & Nosek, B. A. (2014). How ideological migration
geographically segregates groups. Journal of Experimental Social Psychology, 51, 1–14.
Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement of meaning. University
of Illinois Press.
Parkinson, C., Kleinbaum, A. M., & Wheatley, T. (2017). Similar neural responses predict
friendship. Nature Communications, 9, 332.
Pew Research. (2017). The Partisan divide on political values grows even wider: Sharp shifting
among democrats on aid to needy, race, immigration. http://www.people-press.org/2017/10/05/
the-partisan-divide-on-political-values-grows-even-wider/
Poteat, V. P., Mereish, E. H., Liu, M. L., & Sophia Nam, J. (2011). Can friendships be bipartisan?
The effects of political ideology on peer relationships. Group Processes & Intergroup Relations,
14(6), 819–834.
Rivera, L. A. (2012). Hiring as cultural matching: The case of elite professional service firms.
Sani, F. (2012). Group identification, social relationships and health. In J. Jetten, C. Haslam, & S. H.
Alexander (Eds.), The social cure: Identity, health, and well-being (pp. 21–37). Psychology
Press.
Savage, S. V., Stets, J. E., Burke, P. J., & Sommer, Z. L. (2017). Identity and power use in exchange
networks. Sociological Perspectives, 60(3), 510–528.
Schwarz, O., & Shani, G. (2016). Culture in mediated interaction: Political defriending on facebook
and the limits of networked individualism. American Journal of Cultural Sociology, 4(3),
385–421.
Serpe, R. T. (1987). Stability and change in self: A structural symbolic interactionist explanation.
Shrum, W., Cheek, N. H., & Hunter, S. M. D. (1988). Friendship in school: Gender and racial
homophily. Sociology of Education, 61(4), 227–239.
Smirnov, I., & Thurner, S. (2017). Formation of homophily in academic performance: Students
change their friends rather than performance. PLoS ONE, 12(8), e0183473.
Smith, J. A., McPherson, M., & Smith-Lovin, L. (2014). Social distance in the United States: Sex,
race, religion, age, and education homophily among confidants, 1985 to 2004. American
Stets, J. E. (2010). The social psychology of the moral identity. In S. Hitlin & S. Vaisey (Eds.),
Handbook of the sociology of morality (pp. 385–409). Springer.
Stets, J. E. (2016). Rationalist vs. intuitionist views on morality a sociological perspective. In
C. Brand (Ed.), Dual-process theories in moral psychology (pp. 345–366). Springer
Fachmedien Wiesbaden.
Stets, J. E., Burke, P. J., & Savage, S. V. (2018). Exchange, identity verification, and social bonds.
Stets, J. E., & Carter, M. J. (2006). The moral identity: A principle level identity. In K. A.
McClelland & T. J. Fararo (Eds.), Purpose, meaning, and action (pp. 293–316). Palgrave
MacMillan.
Quarterly, 74(2), 192–215.
symbolic interaction: Deepening foundation, building bridges (pp. 65–88). Springer.
social psychology (pp. 31–60). Kluwer-Plenum.
Stryker, S. (2002 [1980]). Symbolic interactionism: A social structural version. Benjamin/Cum-
mings Publishing.
social structures on commitment. Advances in Group Processes, 22, 93–123.
Suitor, J., & Keeton, S. (1997). Once a friend, always a friend? Effects of homophily on women’s
support networks across a decade. Social Networks, 19(1), 51–62.
Thoits, P. A. (2020). The relationship between identity importance and identity salience: Context
matters. In R. T. Serpe, R. Stryker, & B. Powell (Eds.), Identity and symbolic interaction:
Deepening foundations, building bridges (pp. 37–63). Springer.
Tsushima, T. M., & Burke, P. J. (1999). Levels, agency, and control in the parent identity. Social
Turner, J. H. (2016). The macro and meso basis of the micro social order. In S. Abrutyn (Ed.),
Handbook of contemporary sociological theory (pp. 123–148). Springer.
Vaisey, S., & Lizardo, O. (2010). Can cultural worldviews influence network composition. Social
Forces, 88(4), 1595–1618.
White, H. C. (1992). Identity and control. Princeton University Press.
Wright, K. B. (2012). Emotional support and perceived stress among college students using
Facebook.com: An exploration of the relationship between source perceptions and emotional
support. Communication Research Reports, 29(3), 175–184.
Yarrison, F. W. (2016). Contextualizing proximate social structure in identity theory. In J. E. Stets
& R. T. Serpe (Eds.), New directions in identity theory research (pp. 343–365). Oxford
University Press.
Chapter 15
Relational Positioning As Intermediate
Social Structure in Identity Theory
Timothy R. Rose and Philip S. Brenner
Introduction
A key aspect of social interaction, and one that has received relatively little attention
in research on identity, is the physical location in which it occurs. Embodied
individuals inhabit locations in the physical environment where they engage in
interaction and travel through space in between interaction sites. Mead (1934)
acknowledged the impact of physical location on interaction in the context of two
physical laws of the universe: objects possessing mass require the expenditure of
energy to change position and are unable to occupy the same position as other
objects with mass. We revisit and expand on this early formulation of symbolic
interactionism by examining how the layout of the physical environment facilitates
and constrains access to the locations where social interaction occurs and in which
identities are enacted.
The layout of the physical environment determines, at least partially, opportuni-
ties for social interaction. The physical locations we inhabit in our towns and
neighborhoods, the others who are present in them, and the distance between these
locations, encourage or constrain interaction. When the layout of the physical
environment requires fewer resources to reach a place associated with a given
identity, individuals will form stronger bonds with others also in that location due
to an increased level of exposure to these individuals (Proshansky et al., 1983; Tuan,
1975). When more resources are needed reach a place, the quantity and quality of
bonds with those at that hard-to-reach location may suffer (Freese & Burke, 1994).
T. R. Rose (*)
Kent State University, Kent, OH, USA
P. S. Brenner
304 T. R. Rose and P. S. Brenner
The distance between places in the physical environment where interaction

occurs, conceptualized as relational positioning, may act as form of social structure
that can facilitate and restrict interaction and encourage or suppress identity forma-
tion and enactment. If the arrangement of places in the physical environment and the
resources required to move between them influence interaction by increasing or
decreasing the likelihood of contact between individuals, then the physical environ-
ment also plays a role in individuals’ collections of identities. Religious places, such
as cathedrals and mosques, and religious identities provide a good illustration. The
place of one’s birth influences one’s religious affiliation not only through the religion
of one’s parents, but also the religious habitus of those in the individual’s community
(Brenner, in this volume). An accessible religious community as what might be
found at a local church, temple or other religious locale, provides a place for
individuals to enact their religious identities. The ease of travel to the religious site
allows more opportunities for the religious identity to be enacted. Through this
increased potential for interaction with others who share a religious identity, the
individual’s own religious identity may be strengthened. As one knows more people
through this identity and those connections are felt more deeply (i.e., increased
interactional and affective commitments), the value placed on the religious identity
(religious identity prominence) increases, in turn increasing the likelihood of its
enactment (religious identity salience).
We introduce relational positioning as a form of intermediate social structure,
conceptualized as localized social forces that shape the probability of coming into
regular contact with a stable and identifiable group of individuals (Stryker et al.,
2005). We argue that relational positioning, as a form of intermediate social struc-
ture, shapes the structure of the self by way of the identities the individual holds,
values, and enacts (Serpe & Stryker, 2011; Stryker, 1980). Integrating the relational
positioning into identity theory as an intermediate social structure will add context to
processes already identified within identity theory, such as proximate social structure
and identity commitment, and it will further ground identity theory in the physical
environment in which individuals are located and in the social networks in which
individuals are embedded.
We test this idea using the religious identity because its frequent and regular
enactment, such as weekly attendance at church services, facilitates measurement.
We use a sample of religious individuals from different types of communities based
on population size. Including communities of differing sizes will permit us to assess
variation in relational positioning, operationalized as resources expended to traverse
the physical environment between individuals’ homes and their religious meeting
places. We anticipate these differences in relational positioning will be reflected in
variation in the opportunities to interact with others in the religious identity. Rela-
tional positioning is used to predict the prominence and salience of the religious
identity.
15 Relational Positioning As Intermediate Social Structure in Identity Theory 305
Background
Identity theory is rooted in a set of basic principles of interaction. Individuals use

language to convey meaning, act on the basis of shared meanings, have the ability to
take the position of another person, and reflexively see themselves as others see them
(Mead, 1934). These ideas were later expanded on by early symbolic interactionists
and role theorists, who were primarily concerned with the ways in which meanings
and behavioral expectations were associated with a social position (Stryker, 1980;
Burke & Stets, 2009). This expansion of Mead’s original premises led to the
examination of identities, defined as internalized meanings and the correspondence
of identity meanings with behavior meanings, a primary focus of identity theory.
Indeed, the initial concerns of the theory focused on why one identity would be
enacted rather than another. Why would one person decide to spend free time with
their children, while another decides to play golf with friends (Stryker, 1968, 1980)?
Identity theory conceptualizes behavior as an outcome of a process that includes
three key concepts. The first, commitment, is the link the individual has to others via
an identity, operationalized by the number of others one knows through an identity
and the affective strength of these connections. One’s identity as a Catholic or a
parishioner in their local church is strengthened by the number of others one knows
through that identity, such as one’s fellow Catholics or parishioners, and the strength
of those connections. These two types of commitment, interactional and affective,
are influenced by relational positioning in the physical environment. Systematic
differences in the opportunities to meet and know others caused by social structures,
such as residential segregation, may contribute to systematic variation in commit-
ments (Gottdiener & Hutchinson, 2011; Hawley, 1971; Sigelman et al., 1996).
The second key concept, prominence, is the value or importance placed on the
identity (Brenner et al., 2014). Identities are arranged in a prominence hierarchy
given how central they are to how individuals see themselves. Affective commit-
ment has a positive effect on the prominence of an identity as the deeply felt,
intensive connections to others through an identity increases the importance placed
on that identity (Burke & Stets, 2009; Owens et al., 2010; Serpe & Stryker, 2011).
The third key concept, identity salience, is the likelihood of an identity being
acted upon across situations (McCall & Simmons, 1966; Stryker & Serpe, 1994,
1983; Stryker, 1968, 1987). Identities also are arranged in a salience hierarchy
regarding how likely they are to be acted upon across situations. Interactional
commitment has a direct, positive effect on the salience of an identity as the
extensiveness of the numerous connections to others through an identity increases
the likelihood of enactment of that identity (Burke & Stets, 2009; Owens et al., 2010;
Serpe & Stryker, 2011).
Identities are influenced by social structure which is conceptualized in three
levels—large, intermediate, and proximate—that shape the affective and interactive
commitments individuals form (Stryker et al., 2005). Large social structures are
those that order or stratify society, such as race, gender, and class. These structures
operate as guiderails or barriers that limit and constrain individuals to connections
with others, by promoting racial or class-based homogamy or gendered employment

norms (Stryker, 1980). Large social structure influences intermediate social struc-
ture, which is reflected in individuals’ experiences as they move through their daily
lives in their neighborhoods, schools, and churches. This level of social structure
includes networks of individuals with whom one interacts in a variety of contexts,
such as neighbors and acquaintances, schoolmates and their parents, and fellow
congregants and pastors (Stryker et al., 2005). Proximate social structures are those
closest to individuals as they enact identities in their everyday lives, composed of
one’s spouse or partner, children, close friends, and so on. The homogeneity of the
proximate social structure has a positive association with the salience of the associ-
ated identities such as spouse or partner, parent, and friend (Merolla et al., 2012;
Stets et al., 2017; Stets & Serpe, 2013; Yarrison, 2016).
Relative to the advancements in the understanding of proximate social structure
and its influence on identities, the potential influence of intermediate social structure
remains largely under-theorized and under-investigated. We argue here that the
physical environment shapes individuals’ opportunities to engage in identity-related
interaction. Moreover, understanding the physical layout of the social world and the
effort required to traverse space can benefit our understanding of identities. A
primary focus of identity theory is the influence of social structure on human
behavior (Serpe & Stryker, 1987; Stryker, 1980). A conceptualization of social
structure that includes the physical environment allows us to understand its potential
influence on behavior through the lens of identity theory.
Identities in the Physical Environment
Understanding interaction requires understanding the physical space within which it

emerges (Mead, 1934). Human beings inhabit physical bodies that are constrained
by two important properties of physical space. First, two objects with mass, such as
the bodies of two people, cannot jointly occupy the same space. The relational
component of the physical environment, the space or distance between distinct
locations between which an individual can travel, focuses on the position of one
location, such as an individual’s church, given that of another, such as an individ-
ual’s home (Gans, 2002; Logan, 2012). Relational positioning is operationalized in
terms of distance, such as the number of kilometers between one’s home and one’s
church, but it may also be characterized in terms of time, such as the number of
minutes it takes to drive between locations.
Second, changing an object’s relative position requires an expenditure of
resources, such as transportation costs or time (Brenner, 2017a; Kwan, 1998; Pucher
& Renne, 2005). Monetary costs, such as car payments, insurance, and fuel, create
structural barriers that may severely limit travel, especially in areas underserved by
or lacking public transit (Millward & Spinney, 2011; Pucher & Renne, 2005). Larger
distances between positions and the increased time required to traverse distances
contribute to these barriers (Handy & Neimeier, 1997; Kwan, 1998). Even when
absolute distances are small, the costs of changing relative positions can be high in
the face of physical barriers, such as traffic congestion and poor transportation
infrastructure for motorists (Cervero & Gorham, 1995), cyclists, and pedestrians
(Grannis, 1998). The resources necessary to change one’s position relative to a
specific location, such as traveling to church from one’s home, may act as a
structural constraint (Brenner, 2017a; Freese & Burke, 1994). Therefore, relational
positioning considers not only time and distance but also resources required to move
through the physical environment.
The spatial layout of the physical environment and the resources necessary to
traverse it structure human behavior (Grannis, 1998; Hawley, 1971; Kwan, 1998;
Miller, 2005; Sigelman et al., 1996). Relational positioning operates in a manner
consistent with the conceptualization of intermediate social structure in identity
theory by facilitating and constraining contact between interaction partners. This
proposition draws from the foundations of identity theory and symbolic
interactionism. Individuals must envision themselves in the physical environment
in order to understand the degree to which the environment facilitates or inhibits
interaction goals (Burke & Stets, 2009). Relevant features of the physical environ-
ment are those factors that widen or narrow relational positioning by increasing or
decreasing the difficulty of traversing the space between locations, such as one’s
home and church, where identities are enacted. Like other intermediate social
structures, relational positioning may influence identity prominence and salience
through its effect on identity commitments by facilitating or constraining contact
with others enacting counter-roles.
Using the religious identity as an example provides an opportunity to test this
proposition. Enactment of the religious identity typically occurs at a designated
place, such as a church or temple that is recognized by members of that religious
organization and others in the community. Individuals with shared religious identi-
ties regularly and frequently assemble to enact their religious identities. Yet, the
physical environment exerts a strong influence on the individual’s ability to attend
through the relational positioning of home and church, which determines the effort
and resources that must be expended to reach one’s place of religious practice. We
argue that the greater the effort and resources required to reach a location, such as
one’s place of religious practice, the lower the frequency and regularity of religious
behavior at the location, including attendance at religious services (Brenner, 2017a;
Miller, 2005). Therefore, we anticipate that:
H1 The effort individuals must expend to change their relational positioning to
that of their primary house of worship will be negatively related to attending
that place of worship.
The frequency and regularity of church attendance with others who share the
religious identity influences the individual’s relationships. Frequent attendance may
shift the composition of one’s proximate social structure toward one that is more
religiously homogeneous. This potential has increased in the past couple of decades
as churches expanded to add programs and services beyond worship and religious
education. New programs and services, such as coffee shops, bookstores, and group
travel encourage the quantity and quality of bonds with fellow congregants, poten-
tially to the detriment of connections with others outside of the religious organization
(Wade, 2016). Frequent attendance also increases interactional commitment to the
religious identity as one knows more people through church and other religious
activity. Affective commitment may also be affected through the strengthened bonds
brought about by increased familiarity with fellow congregants in their shared
identities (Dovidio et al., 2003).
Thus, those who attend frequently will form close attachments to others in the
church, bringing them into their proximate social structure. In turn, these bonds
should encourage interactional and affective commitments to others in the church
and further anchor the individual into their religious identity. Consequently, we
expect that:
H2: The frequency with which one attends their primary house of worship will be
positively related to the proportion of the individual’s proximate social
structure who share in the religious identity.
H3: A homogenous proximate social structure will be positively associated with
one’s interactive commitment to the religious identity.
H4: A homogenous proximate social structure will be positively associated with
one’s affective commitment to the religious identity.
Proximate social structure also directly affects identity prominence and salience
(Burke & Stets, 2009; Merolla et al., 2012; Stryker & Serpe, 1994). The character-
istics of those in one’s proximate social structure, those with whom one frequently
and regularly interacts, shape the individual by increasing the prominence and
salience of the identity being enacted (Stryker et al., 2005; Yarrison, 2016). The
higher proportion of others in one’s proximate social structure who share an identity,
the more value will be placed on that identity, and the more likely it will be enacted.
Thus, as one attends religious services more frequently, and fellow congregants
become a higher proportion of one’s proximate social structure, the religious identity
will become more prominent and more salient.
H5: A homogenous proximate social structure will be positively related to the
prominence of the religious identity.
H6: A homogenous proximate social structure will be positively related to the
salience of the religious identity.
Relationships between religious identity commitments, religious identity promi-
nence, and religious identity salience are assumed to mirror those from prior
research. Affective and interactive commitments to the religious identity positively
shape the location of that identity in the prominence and salience identity hierarchies
(Stryker & Serpe, 1994). As individuals’ links to those in their church are higher in
quantity and quality, they will place more value on their religious identity and
become more likely to enact it. Further, prominence is posited to positively affect
salience (Brenner et al., 2014). Individuals are more likely to enact identities that are
important to the person’s sense of self, such as the religious identity. Given the
above, we predict that:
Fig. 15.1 Heuristic model of relational positioning as Intermediate Social Structure
H7: Affective commitment will be positively associated with the prominence of

the religious identity.
H8: Affective commitment will be positively associated with the salience of the
religious identity.
H9: Interactive commitment will be positively associated with the prominence of
the religious identity.
H10: Interactive commitment will be positively associated with the salience of the
religious identity.
H11: Prominence will be positively associated with the salience of the religious
identity.
A heuristic model illustrates these hypotheses as a series of paths in a structural
model (see Fig. 15.1).
Data and Methods
Data collection occurred in 2018 and 2019 at the Survey Research Lab at Kent State
University. A web survey of a non-probability sample was conducted. Although
these data are not generalizable, they are useful for investigating identity processes
and assessing new measures (Long, 2016; Markowski, 2019; Yarrison, 2016).
Panelists were screened to meet the eligibility criteria: adult Christians 18 years of
age or older who attend a place of worship more than once a year on average and at
least once in the past 12 months. Eligibility criteria were used given the predomi-
nantly Christian population under study and to ensure that survey respondents
attended readily recognizable houses of worship. Limiting eligibility to those attend-
ing in the past 12 months ensured that respondents were able to answer the questions
about the effort required to reach their primary church. A quota was used to ensure a
relatively equal distribution by sex.
Eligibility was limited to panelists living in Indiana, Kentucky, and Ohio com-
munities of 20,000 residents or fewer that were not connected to another community
(at least one mile away from neighboring communities) or a larger metropolitan area.
These states were purposively selected to ensure diversity in community size given
the presence of many small municipalities. We sample from smaller communities as
they offer fewer choices for religious practice. Communities meeting these require-
ments were identified in the 2010 Census. The final sample size was 1216.
Measures The survey included questions identifying the primary place of worship,
frequency of attendance, relational positioning, proximate social structure, religious
identity commitment, prominence, and salience. Respondents were asked to report
the number of churches they attended in the last year, naming the primary and up to
five additional churches they attended. The name of the primary church was auto-
matically inserted into subsequent questions. Frequency of attendance was
operationalized using a single item that measured the frequency of attendance in
the last 12 months at one’s primary place of worship using a 7-point scale: “once a
year or less frequently,” “several times a year,” “at least once every two months,”
“about once a month,” “two or three times a month, once a week,” or “more than
once a week.”
Measures of relational positioning as an intermediate social structure included
geographic distance, as well as the tangible and intangible resources required to
travel these distances (Kwan, 1998). Respondents rated how much each of the
following influenced their frequency of attending their primary place of worship:
how much does distance (e.g., miles to the place) factor into how often you attend
your place of worship; how difficult is traveling to your place of worship; how much
does travel time factor into how often you attend your place of worship; and how
tiring is traveling to your place of worship. Each question used a 5-point scale, from
(1) “not at all” to (5) “a great deal.” This scale formed a single factor with good
internal consistency (α ¼ 0.82).
Measures of proximate social structure mirror recent research, focusing on
religious homogeneity in the respondent’s everyday contacts (Brenner et al., 2014;
Merolla et al., 2012; Yarrison, 2016). The proportion of others who share the
respondent’s religious identity was scored on a 5-point scale, from “almost none”
to “almost all” across three groups: close friends, friends, and others with whom one
interacts with on a daily basis. These three items formed a single factor with good
internal consistency (α ¼ 0.85).
Interactive commitment to the religious identity was operationalized as the
frequency of contact, duration of contact, and amount of discretionary income
spent with those who share one’s religious identity. Although others who do not
share one’s religious identity can still verify that identity, these measures assess the
degree to which individuals feel committed to an identity given their connections
with others who share the identity. Frequency of contact with others who share the
respondent’s religious identity in an average week was measured on a 7-point scale
from “never” to “daily.” Duration of contact with others who share the respondent’s
religious identity in an average week was measured on a 5-point scale, from “less
than 5 hours a week” to “more than 30 hours a week.” The amount of discretionary
income spent on doing things with those who share their religious identity was
measured on a 5-point scale, from “almost none” to “almost all.” These three items
formed a single factor with acceptable internal consistency (α ¼ 0.65). Although this
level of internal consistency is consistent with previous research, it suggests mea-
surement improvements are needed (Owens & Serpe, 2003).
Affective commitment to the religious identity was operationalized as closeness
to and importance of others who share one’s religious identity, and how much they
would be missed. Personal and emotional closeness to others known through one’s
religious identity, and the importance of these others, were both measured on 4-point
scales, from “not at all” to “very.” How much respondents would miss these others if
they were not able to see them anymore was measured on a 4-point scale from “not at
all” to “a great deal.” All three items formed a single factor with good internal
consistency (α ¼ 0.81).
Religious identity prominence was measured using a four-item index adapted
from previous research (Merolla et al., 2012; Stets et al., 2017; Yarrison, 2016).
Respondents were asked to rate their agreement on a 4-point scale, from strongly
agree to strongly disagree, with four statements: (1) being religious is an important
part of my self-image, (2) being religious is an important reflection of who I am, (3) I
have come to think of myself as a religious person, and (4) I have a strong sense of
belonging to the community of religious people. These items loaded on a single
factor with good internal consistency (α ¼ 0.88).
Religious identity salience was operationalized as the likelihood of contextual
disclosure of one’s religious identity (Yarrison, 2016). In the context of meeting the
friend of a close friend for the first time, respondents were asked to rate the
likelihood from “almost certainly would” to “almost certainly would not” that they
would disclose their religious identity (1) if the other person asked about their
religion, or (2) mentioned being religious. Respondents were then asked how likely
they would be (3) to ask the other person about their religion, or (4) invite them to get
together in the future if the other person mentioned their religion. These items loaded
onto a single factor with good internal consistency (α ¼ 0.79).
Demographics A set of demographic variables are included to control for factors
which shape how individuals view themselves and engage with their environment
and large social structure (Brenner et al., 2017; Owens & Serpe, 2003). Age was
measured in years (mean ¼ 45). Gender was measured as female, male, and some
other gender, but later collapsed as female (60%) or not female. Race was measured
as White, Black or African American, American Indian or Alaskan Native, Asian,
Pacific Islander, Multi-racial, and some other race. As over 90% identified as White,
race was excluded from the final model. Annual household income was measured in
a set of bracketed incomes with a median income between $25,000 and $34,999.
Educational attainment was measured as highest level of education completed, from
Table 15.1 Descriptive statistics

Variable Mean (SD) Range N
Effort to change relational Physical exhaustion from travel 1.47 (1.03) 1–5 1206
position Perceived burden of travel 1.52 (1.10) 1–5 1207
Travel time 1.90 (1.40) 1–5 1206
Travel distance 2.12 (1.55) 1–5 1201
Travel Frequency traversing space/ 4.71 (1.86) 1–7 1209
attendance
Proximate social structure: Friends who are religious 3.67 (1.17) 1–5 1195
Identity Homogeneity Close friends who are religious 3.57 (1.14) 1–5 1192
Everyday contacts who are 3.26 (1.15) 1–5 1136
religious
Affective identity commitment How much would you miss 3.27 (0.88) 1–4 1150
others
How close in personal/emo- 3.15 (0.81) 1–4 1175
tional terms
Importance of identity-related 3.38 (0.76) 1–4 1169
others
Interactive identity How often do you interact 4.24 (1.63) 1–7 1174
commitment monthly
How many interaction hours 1.88 (1.14) 1–5 1120
weekly
Income spent on identity 1.96 (0.97) 1–5 1074
interaction
Identity prominence Religion is important part of 3.28 (0.77) 1–4 1186
self-image
Being religious is important 3.34 (0.74) 1–4 1186
reflection of me
I think of myself as religious 3.23 (0.73) 1–4 1171
Strong sense of belonging to 3.16 (0.77) 1–4 1165
religious community
Identity salience Likely to tell unknown other 5.69 (1.72) 1–7 1170
they are religious
Likely to tell religious other they 6.09 (1.45) 1–7 1194
are religious
Ask Religious other more about 5.06 (1.84) 1–7 1167
being religious
Ask Religious other to get 4.99 (1.62) 1–7 1115
together in future
less than high school, to graduate degree or professional degree, with 24% reporting
a high school diploma or GED and 28% reporting some college.
Table 15.1 provides the summary statistics for each measure. Table 15.2 presents
a zero-order correlation matrix for all non-dichotomous variables.
Analysis Structural equation models allowed estimation of a measurement model
for relational positioning, proximate social structure, affective and interactional
Table 15.2 Zero order correlations
15
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1 Relational
positioning 1
2 Relational 0.62
positioning 2
3 Relational 0.54 0.57
positioning 3
4 Relational 0.44 0.52 0.61
positioning 4
5 Attendance 20.15 20.13 20.15 20.11
6 Prox. Social 20.07 20.06 20.06 20.07 0.27
structure 1
7 Prox. Social 0.03 0.04 0.05 0.05 0.24 0.81
structure 2
8 Prox. Social 0.02 0.02 0.01 0.03 0.16 0.55 0.58
structure 3
9 Affective 20.11 0.04 20.10 0.05 0.26 0.31 0.29 0.18
commitment 1
10 Affective 0.04 0.01 20.07 0.01 0.27 0.33 0.35 0.23 0.62
commitment 2
11 Affective 20.06 20.06 20.12 0.05 0.23 0.29 0.29 0.20 0.58 0.61
commitment 3
12 Interactive 20.06 0.04 20.07 20.06 0.27 0.39 0.38 0.36 0.36 0.40 0.31
Commitment 1
Relational Positioning As Intermediate Social Structure in Identity Theory
13 Interactive 0.08 0.06 0.02 0.07 0.07 0.20 0.22 0.23 0.14 0.25 0.13 0.49
Commitment 2
14 Interactive 0.11 0.12 0.06 0.10 0.07 0.19 0.19 0.18 0.20 0.31 0.22 0.32 0.42
Commitment 3
15 Prominence 1 0.01 0.03 0.01 0.01 0.26 0.30 0.30 0.22 0.26 0.31 0.29 0.18 0.08 0.19
16 Prominence 2 0.03 0.01 0.02 0.02 0.27 0.29 0.32 0.24 0.26 0.31 0.31 0.22 0.10 0.18 0.78
313
(continued)
314
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
17 Prominence 3 0.05 0.01 0.04 0.05 0.24 0.27 0.28 0.20 0.23 0.28 0.30 0.24 0.13 0.14 0.64 0.68
18 Prominence 4 0.03 0.01 0.04 0.02 0.38 0.33 0.34 0.22 0.37 0.45 0.40 0.30 0.15 0.25 0.59 0.62 0.60
19 Salience 1 20.09 20.09 20.07 20.07 0.19 0.21 0.19 0.17 0.31 0.33 0.33 0.28 0.13 0.13 0.31 0.30 0.30 0.32
20 Salience 2 20.12 20.12 20.07 20.11 0.25 0.27 0.21 0.15 0.34 0.30 0.32 0.25 0.07 0.09 0.35 0.34 0.34 0.32 0.71
21 Salience 3 0.04 0.05 20.06 0.03 0.20 0.22 0.21 0.16 0.32 0.35 0.31 0.23 0.14 0.20 0.27 0.27 0.27 0.32 0.51 0.53
22 Salience 4 0.01 0.01 0.02 0.01 0.16 0.17 0.15 0.14 0.29 0.31 0.27 0.20 0.15 0.20 0.21 0.22 0.19 0.29 0.37 0.42 0.41
23 Age 214 20.13 20.16 20.14 0.15 0.05 0.07 0.02 0.04 20.06 0.04 0.05 20.23 20.14 0.04 0.03 0.03 0.06 0.05 0.05 20.09 0.06
24 Education 0.02 0.04 0.02 0.09 0.03 0.12 0.11 0.10 0.03 0 0.05 0.11 0 0.01 0.04 0.01 0.02 0.01 20.08 20.07 20.07 20.08 0.11
25 Income 20.07 20.08 0.04 20.06 0.06 0.12 0.11 0.01 0.01 0 0.04 0.06 0.02 0.03 0.02 0.07 0.02 0.08 0.04 0.04 0.05 20.10 0.08 0.37
Bolded correlations indicate significance at p < 0.05 level
T. R. Rose and P. S. Brenner
commitments to the religious identity, and religious identity prominence and

salience (Brenner et al., 2017; Yarrison, 2016). Effects of these latent constructs
are then estimated given predicted pathways illustrated in the heuristic model in
Fig. 15.1. Standardized coefficients are presented.
Results
Measurement Model In Table 15.3, the measurement model shows that all latent
constructs included in the structural model are reflected in their manifest variables
well. All four manifest variables reflecting relational positioning load on their latent
construct well. Coefficients range from 0.70 to 0.78 and all are statistically signif-
icant. All three manifest variables reflecting proximate social structure load on their
Table 15.3 Measurement model

Parameter estimate Standardized SE
Effort to change relational position
Physical exhaustion 0.72* 0.02
Perceived burden 0.78* 0.02
Travel time 0.77* 0.02
Travel distance 0.70* 0.02
Proximate social structure
Friends 0.88* 0.01
Close friends 0.92* 0.01
Everyday contacts 0.63* 0.02
Affective commitment
How much would you miss others 0.77* 0.02
How close in personal/emotional terms 0.81* 0.02
Importance of identity-related others 0.76* 0.02
Interactive commitment
How often do you interact monthly 0.77 0.03
How many interaction hours weekly 0.64 0.03
Income spent on identity interaction 0.49 0.03
Identity prominence
Religion is important part of self-image 0.86* 0.01
Being religious. Is important reflection of me 0.89* 0.01
I think of myself as religious 0.77* 0.01
Strong sense of belonging to religious community 0.72* 0.02
Identity salience
Likely to tell unknown other they are religious 0.82* 0.02
Likely to tell religious other they are religious 0.86* 0.01
Ask religious other more about being religious 0.63* 0.02
Ask religious other to get together in future 0.56* 0.03
latent construct well. Coefficients range from 0.63 to 0.92 and all are statistically
significant. All three manifest variables reflecting affective commitments to the
religious identity load on their latent construct well. Coefficients range from 0.76
to 0.81 and all are statistically significant. All three manifest variables reflecting
interactional commitment to the religious identity load on their latent construct.
Coefficients range from 0.49 to 0.77 and all are statistically significant. All four
manifest variables reflecting religious identity prominence load on their latent
construct well. Coefficients range from 0.72 to 0.89 and all are statistically signif-
icant. All four manifest variables reflecting religious identity salience load on their
latent construct. Coefficients range from 0.56 to 0.86 and all are statistically
significant.
Structural Model In Table 15.4, wider relational positioning and increased effort
required to overcome it is negatively related to the frequency of traveling to one’s
primary church (β ¼ 0.17). This finding supports Hypothesis 1. The frequency of
attending church is positively associated with the religious identity homogeneity of
one’s proximate social structure (β ¼ 0.38). This effect remains significant even after
accounting for the significant indirect effect relational positioning has on proximate
social structure (β ¼ 0.06) by way of the frequency of attendance. This supports
Hypothesis 2. Taken together, these two findings support the inclusion of relational
positioning as a form of intermediate social structure in identity theory.
Proximate social structure is positively associated with affective commitment to the
religious identity (β ¼ 0.61), interactive commitment to the religious identity
(β ¼ 0.70), and religious identity prominence (β ¼ 0.38). The direct effect of
proximate social structure on religious identity salience (β ¼ 0.01) does not reach
significance, although there are significant indirect paths between proximate social
structure and salience. The total indirect effect of proximate social structure on
salience (β ¼ 0.42) is significant and comprises 98% of the total effect (β ¼ 0.43).
These findings support Hypotheses 3, 4, and 5. Hypothesis 6, however, is only
supported through indirect associations by way of commitment and prominence.
Affective commitment to the religious identity is positively associated with both
religious identity prominence (β ¼ 0.32) and salience (β ¼ 0.31). These findings
support Hypotheses 7 and 8. Interactive commitment to the religious identity,
however, is not significantly associated with either religious identity salience or
prominence, even when taking account of the indirect effect of interactive commit-
ment on identity salience. Thus, Hypothesis 9 and 10 are not supported. However,
the frequency of attending one’s place of worship significantly and positively
influences prominence (β ¼ 0.21) and salience (β ¼ 0.16), which may explain
why these hypotheses are not supported. Finally, religious identity prominence has
a direct positive influence on religious identity salience (β ¼ 0.32). This finding
supports Hypothesis 11.
Overall, these findings support most of our hypotheses. The hypotheses not
supported involve the direct links from interactional commitment to the religious
identity to religious identity prominence and salience. The association between
proximate social structure and religious identity salience is significant but indirect.
15
Table 15.4 Structural Model Direct Effects (DE), Indirect Effects (IE), and Total Effects (TE) and Standard Errors
Relational Proximate Social
positioning Attendance Structure Affective Commitment Interactive Commitment Prominence Salience
DE IE TE DE IE TE DE IE TE DE IE TE DE IE TE DE IE TE DE IE TE
Structural
model
Relational 20.17 -- 20.17 -- 20.06 20.06 -- 20.04 20.04 -- 20.04 20.04 -- 20.03 20.03 -- 20.03 20.03
positioning (0.09) (0.09) (0.02) (0.02) (0.01) (0.01) (0.02) (0.02) (0.01) (0.01) (0.01) (0.01)
Frequency 0.38 – 0.38 – 0.23 0.23 – 0.26 0.26 – 0.21 0.21 –– 0.16 0.16
of (0.01) (0.01) (0.01) (0.01) (0.02) (0.02) (0.01) (0.01) (0.02) (0.02)
attendance
Proximate 0.61 – 0.61 0.70 – 0.70 0.38 0.17 0.55 0.01 0.42 0.43
social (0.04) (0.04) (0.09) (0.09) (0.05) (0.03) (0.03) (0.11) (0.09) (0.07)
structure
Affective 0.32 – 0.32 0.31 0.10 0.41
commitment (0.04) (0.04) (0.09) (0.04) (0.09)
Interactive 0.04 -- 0.04 0.08 0.01 0.07
commitment (.03) (.03) (0.06) (0.02) (0.07)
Prominence 0.32 – 0.32
(0.10) (0.10)
Covariates
Age 20.19 – 20.19 0.11 0.03 0.14 0.01 0.05 0.07 20.06 0.04 0.02 20.24 0.05 20.19 0.03 0.03 0.05 0.05 0.01 0.01
(0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01)
Female 0.01 – 0.01 0.01 0.01 0.01 0.21 0.01 0.21 0.04 0.13 0.09 20.19 0.15 0.04 20.08 0.11 0.03 0.04 0.04 0.04
(0.05) (0.05) (0.11) (0.02) (0.11) (0.06) (0.02) (0.06) (0.04) (0.03) (0.05) (0.09) (0.07) (0.09) (0.04) (0.03) (0.04) (0.09) (0.07) (0.07)
Relational Positioning As Intermediate Social Structure in Identity Theory
Income 20.07 – 20.07 0.03 0.01 0.05 0.07 0.02 0.08 0.06 0.05 0.01 0.02 0.06 0.03 0.04 0.03 0.07 0.04 0.02 0.02
(0.01) (0.01) (0.03) (0.01) (0.03) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.02) (0.01) (0.02) (0.01) (0.01) (0.01) (0.02) (0.01) (0.01)
Education 0.01 – 0.01 0.01 0.01 0.01 0.14 0.01 0.13 0.11* 0.08 0.03 0.01 0.09 0.09 0.05 0.04 0.01 20.07 0.01 0.01
(0.02) (0.02) (0.04) (0.01) (0.04) (0.02) (0.01) (0.02) (0.02) (0.01) (0.02) (0.03) (0.02) (0.03) (0.01) (0.01) (0.01) (0.03) (0.02) (0.02)
Note: Statistically significant coefficients ( p < 0.05) in bold. Standard errors in parentheses. RMSEA ¼ 0.048; CFI ¼ 0.94; CD ¼ 0.30; Chi2 ¼ 253 ( p > Chi2 ¼ 0.000)
317
It is worth emphasizing that both hypotheses associated with the introduction of

relational positioning and intermediate social structure are supported.
Discussion
A primary goal of this research was to further the development of intermediate social
structure in identity theory. This level of social structure facilitates and constrains the
degree to which individuals come into regular contact with others who share an
identity. It is larger in scope than the proximate social structure, which encompasses
a narrow set of immediate family and very close friends, creating conditions under
which proximate social structure operates. Intermediate social structure is smaller in
scope than large social structure, which encompasses society-level forces, such as
race, gender, and class, which push their influence down intermediate social struc-
tures, shaping society in ways that determine individuals’ likelihoods of encounter-
ing others, such as neighborhood segregation and the placement of physical
structures in cities, towns, and neighborhoods (Gottdiener & Hutchinson, 2011;
Sigelman et al., 1996).
While there are likely many forms of intermediate social structure, this research
examined how relational positioning in the physical environment operates as one of
them. We find that relational positioning, operationalized in terms of distance, time,
or effort to traverse a distance, structures individuals’ behavior (Kwan, 1998; Logan,
2012). Moreover, relational positioning is relevant to identity enactment. As rela-
tional positioning widens, the frequency of reaching the location of identity enact-
ment is reduced (Gottdiener & Hutchinson, 2011; Grannis, 1998; Sigelman et al.,
1996).
We used the religious identity as an example, employing the relational position-
ing of the respondent’s home and primary church as a measure of intermediate social
structure. Church is a key place where the individual’s religious identity can be
called up, enacted, and verified by others who share their identity (Burke & Stets,
2009). Indeed, those who frequently and regularly attend religious services have
proximate social structures that are more homogeneously religious. Sharing a space
with other religious individuals presents opportunities that later produce interaction
with an even greater number of religious others through secondary contacts.
While most of our hypotheses were supported, a few unexpected results warrant
comment. The predicted effects of interactive commitment were not supported in
this research. Measurement problems led to a poor fit between the manifest variables
and the latent construct for interactive commitment. As this problem has been noted
in prior research, the results here support a call for a revision to these measures.
Interactive commitment also was found to not have a significant impact on either of
the identity hierarchies, as predicted in prior identity theory research (Stryker &
Serpe, 1994). This finding mirrors that found in other identity theory research that
has examined social structure in regards to various levels, and may be due to the
inclusion of proximate social structure into the identity model. The strong effects of
proximate social structure on affective and interactive commitment, along with the
significant indirect effect of proximate social structure on prominence through both
forms of commitment, supports the conclusion that proximate social structure is now
capturing a social relationship effect previously only measured by interactive com-
mitment. The significant indirect effect of attendance may also have somewhat
captured this effect of social contact on the identity process.
Surprisingly, proximate social structure failed to directly affect identity salience.
It may be that the influence of proximate social structure on identity salience will
wane as the homogeneity of the proximate social structure increases to the point of
saturation. For the religious identity, once a proximate social network nears religious
uniformity, the individual reaches a saturation point of opportunities to enact the
religious identity. Examining the indirect and total effects, in addition to the direct
effects, extends our understanding of the relationship between these concepts. The
total effect of proximate social structure on religious identity salience was significant
when accounting for several indirect pathways by which proximate social structure
shaped the identity process. The theorized link between relational positioning as
intermediate social structure and proximate social structure was supported by their
significant indirect association. This effect shows that relational positioning shapes
the composition of one’s proximate social structure beyond the effect of relational
positioning on church attendance.
Our findings suggest a fruitful avenue for future research: examining indirect as
well as direct effects of social structures in the identity model. Evidence of the
potential for this future line of research are the significant indirect effects of all
preceding identity theory concepts (with the exception of interactive commitment)
on the identity self-hierarchies in this research. Intermediate social structure, exam-
ined here as relational positioning, was found to have a small yet still detectable
impact on the composition of the individual’s self.
Overall, our results support the inclusion and theoretical refinements of interme-
diate social structure in identity theory and open future directions for research.
Incorporating relational positioning into identity theory connects the theory with a
mechanism of social inequality. Relational positioning brings the social inequalities
that influence and reflect the spatial layout of our cities, towns, and rural areas into
the purview of identity theory. This extension of the theory connects self-structure to
both intermediate and large social structures, which shape society spanning systems
of inequality and which are a primary focus of sociological research. This connection
makes a strong case for future research linking self-structure by way of identity
theory to research on intermediate social structures to large social structures, such as
that linking neighborhood characteristics with systems of stratification, such as race
(Sampson & Sharkey, 2008; Sampson & Wilson, 1995) and socio-economic status
(Gieryn, 2000; Villemez, 1980). This connection would allow future research to
more fully integrate sociological theory of inequality into identity theory. One
particularly fruitful line of future examination of this connection would be the
ways in which individuals attempt to advantageously position themselves in the
intermediate social structures into which they are placed by larger social structure.
This line of inquiry reflects the previously established identity theory principle that
individuals not only are shaped by their environment, but actively shape it and there
position therein in order to enact an identity (Burke & Stets, 2009). Future research
examining the manner in which this is accomplished would offer a further link to
wider sociological research on social inequalities by demonstrating links through
which individuals are able to push back against social structures that result in
disadvantageous circumstances.
There are a few limitations to this study. Retrospective reporting of behavior,
such as the frequency of common occurrences is error prone for survey respondents.
Random errors, such as simple forgetting, can add statistical noise to estimates.
Systematic errors, such as the tendency of survey respondents to overreport norma-
tive states and behaviors, can bias survey estimates, including reported religious
behavior (Brenner, 2017b). Although these data likely include some amount of
random and systematic error, the relationships examined here are likely robust to
these measurement artifacts as any bias was likely of the same magnitude and
direction for all estimated concepts. Nevertheless, future research should investigate
use of new technologies (i.e., app-based data collection programs) to capture behav-
ioral and location data in near real-time to increase validity and more effectively
analyze navigation of the physical environment.
Using a nonprobability sample from purposively selected small towns allowed a
clear focus on relational positioning but harmed external validity. The minimum
attendance eligibility requirement also omits religious nonattenders from these
analyses. To what extent relational positioning influences their decision not to attend
should be investigated in future research. Moreover, the focus on a primary church
overlooks the small minority of religious individuals who regularly attend multiple
places of worship, and how this affects the identity process. These limitations are
unlikely to undermine the findings of this study as relational positioning was found
to have an effect on the religious identity for those who at least occasionally attend
one primary place of worship. Future research will need to examine these concepts,
as well as how relational positioning functions in the context of counter-normative
identities, which have been found to operate somewhat differently than normative
identities (Long, 2016; Yarrison, 2016).
Since its inception, identity theory research has been concerned with the ways in
which the social world structures the internal life of the individual. This focus has led
to numerous discoveries, many of which reinforce the foundational principle that the
structure of the self reflects the structure of society (Stryker, 1980). This research
sought to extend the reach of identity theory to include the physical environment in
which much human interaction occurs. In sum, self-structure was found to reflect the
layout of physical environment. When relational positioning reflects constraints on
engaging in interaction with others who share the religious identity, such as distance,
time, or money, the structure of the self will reflect this in the reduced prominence
and salience of the religious identity. Thus, the physical environment is a significant
part of the process of symbolic interaction and identity enactment. This extension to
identity theory highlights the need for future research linking large social inequalities
to the identity process. Physical environment and relational positioning as a form of
intermediate social structure provides an opportunity to reconceptualize social and

economic inequalities from within the framework of identity theory.
References
Brenner, P. S. (2017a). Differential effects of time constraints on athletic behavior and survey
reports of athletic behavior. Sociological Spectrum, 37(2), 97–110.
Brenner, P. S. (2017b). Narratives of response error from cognitive interviews of survey questions
about normative behavior. Sociological Methods and Research, 46(3), 540–564.
identity theory: An empirical examination. Social Psychology Quarterly, 77(3), 231–252.3.
Brenner, P. S., Serpe, R. T., & Stryker, S. (2017). Role specific self-efficacy as precedent and
Cervero, R., & Gorham, R. (1995). Commuting in transit versus automobile neighborhoods.
Journal of the American Planning Association, 61(2), 210–225.
Dovidio, J. F., Gaertner, S. L., & Kawakami, K. (2003). Intergroup contact: The past, present, and
the future. Group Process and Intergroup Relations, 6(1), 5–21.
Freese, L., & Burke, P. (1994). Persons, identities, and social interaction. In B. Markovsky (Ed.),
Advances in group processes (pp. 1–24). JAI Press.
Gans, H. J. (2002). The sociology of space: A use-centered view. City and Community, 1(4),
329–339.
Gieryn, T. F. (2000). A space for place in sociology. Annual Review of Sociology, 26, 463–496.
Gottdiener, M., & Hutchinson, R. (2011). The new urban sociology (4th ed.). Westview Press.
Grannis, R. (1998). The importance of trivial streets: Residential streets and residential segregation.
American Journal of Sociology, 103(6), 1530–1564.
Handy, S. L., & Neimeier, D. A. (1997). Measuring accessibility: An exploration of issues and
alternatives. Environment and Planning A, 29(7), 1175–1194.
Hawley, A. H. (1971). Urban Society: An ecological approach. The Ronald Press Company.
Kwan, M.-P. (1998). Space-time and integral measures of individual accessibility: A comparative
analysis using a point-based framework. Geographical Analysis, 30(3), 191–216.
Logan, J. R. (2012). Making a place for space: Spatial thinking in social science. Annual Review of
Sociology, 38, 507–524.
Long, B. (2016). Stigmatized identities: Choice, accessibility, and authenticity. In J. E. Stets & R. T.
Press.
Markowski, K. L. (2019). Identity, network, and mental health: The relationship between structures
and meaning on distress and subjective wellbeing. Doctoral Thesis, Department of Sociology,
Kent, OH, Kent State University.
McCall, G., & Simmons, J. (1966). Identities and interactions. Free Press.
Mead, G. H. (1934). Mind, self, and society. University Press.
149–172.
Miller, H. J. (2005). A measurement theory for time geography. Geographical Analysis, 37(1),
17–45.
Millward, H., & Spinney, J. (2011). Time use, travel behavior, and the rural-urban continuum:
Results from the Halifax STAR project. Journal of Transportation Geography, 19, 51–58.
Owens, T. J., & Serpe, R. T. (2003). The role of self-esteem in family identity salience and
commitment among African-Americans, Latinos, and whites. In P. J. Burke, T. J. Owens,
R. Serpe, & P. A. Thoits (Eds.), Advances in identity theory and research (pp. 85–104). Kluwer
Academic/Plenum.
Owens, T., Robinson, D. T., & Smith-Lovin, L. (2010). Three faces of identity. Annual Review of
Sociology, 36, 477–499.
Proshansky, H. M., Fabian, A. K., & Kaminoff, R. (1983). Place identity: Physical world social-
ization of the self. Journal of Environmental Psychology, 3, 57–83.
Pucher, J., & Renne, J. L. (2005). Rural mobility and mode choice: Evidence from the 2001
National Household Travel Survey. Transportation, 32, 165–186.
Sampson, R. J., & Sharkey, P. (2008). Neighborhood selection and the social reproduction of
concentrated radical inequality. Demography, 45(1), 1–29.
Sampson, R. J., & Wilson, W. J. (1995). Toward a theory of race, crime, and urban inequality. In
J. Hagan & R. D. Peterson (Eds.), Crime and inequality (pp. 37–56). Stanford University Press.
Serpe, R. T., & Stryker, S. (1987). The construction of self and the reconstruction of social
relationships. Advances in Group Processes, 4, 41–66.
Serpe, R. T., & Stryker, S. (2011). The symbolic interactionist perspective and identity theory. In
S. J. Schwartz, K. Luyckx, & V. L. Vignoles (Eds.), The handbook of identity theory and
research (pp. 225–248). Springer.
Sigelman, L., Bledsoe, T., Welch, S., & Combs, M. W. (1996). Making contact? Black-white social
interaction in an urban setting. American Journal of Sociology, 101(5), 1306–1332.
Stets, J. E., Brenner, P. S., Burke, P. J., & Serpe, R. T. (2017). The science identity and entering a
science occupation. Social Science Research, 64, 1–14.
Stets, J. E., & Serpe, R. T. (2013). Identity theory. In J. Delamater & A. Ward (Eds.), Handbook of
Stryker, S. (1968). Identity salience and role performance. Journal of Marriage and the Family, 4,
558–564.
Stryker, S. (1987). Identity theory: Developments and extensions. In K. Yardley & T. Honess
(Eds.), Self and identity: Psychosocial perspectives (pp. 89–103). Wiley.
Stryker, S., & Serpe, R. T. (1983). Toward a theory of family influence in the socialization of
children. In A. Kerckhoff (Ed.), Research in the sociology of education and socialization (Vol.
4, pp. 47–74). JAI Press.
Tuan, Y.-F. (1975). Place: An Experiential Perspective. Geographical Review, 65(2), 151–165.
Villemez, W. J. (1980). Race, class, and Neighborhood: Differences in the residential return on
individual resources. Social Forces, 59(2), 414–430.
Wade, M. (2016). Seeker-friendly: The Hillsong megachurch as an enchanting Total institution.
Yarrison, F. (2016). Contextualizing proximate social structure in identity theory. In J. E. Stets &
University Press.
Index
A Anxiety, 160, 267

Abortion, 296, 297 Apathy, 89
Academics, 1, 104, 172, 287 Apology, as reparation, 3, 86, 91, 93, 95–98,
Acculturation, 149 102
Adolescence, 68, 161 Appraisal
Adolescents, 62, 187, 189, 192, 195, 196, 226, actual, 223
230 reflected, 13, 42, 64, 88, 136, 155, 203, 228,
Adulthood, 16, 68, 161, 195 267, 292
Adults, 6, 57, 63–65, 68, 70, 78, 119, 161, 186, self, 292
208, 225–228, 230, 231, 236, 239, Ascription, causal, 217
266, 271, 309 Asencio, E.K., 182, 224, 226, 228, 230
Affect control theory, 224, 238 Asian-American, 87, 91, 226
Affirmation, 42, 49, 124 Asians, 3, 19, 20, 23–29, 31, 32, 47, 51–54, 91,
African American Museum of History and 143, 144, 226, 311
Culture (AAMHC), 85 Assimilation, 11, 17
African Americans, 12, 14, 15, 19, 27, 28, 32, Assortativity, 245–262
47, 51–54, 61, 63–66, 85–87, 89–91, Attachment, 40, 140, 308
111, 117, 120–122, 125, 143, 144, Attitudes, 3, 16, 21, 31, 33, 48, 50, 79, 86,
311 88–92, 95, 104, 105, 117, 118,
Age, 46, 50–54, 65, 68, 71, 76, 78, 79, 94, 96, 120–123, 134, 138, 141, 149, 204,
98, 99, 102, 116, 117, 120–122, 140, 208, 219, 225, 227, 238, 249, 287,
143, 144, 146, 147, 160, 177, 189, 288
191, 192, 195, 196, 206, 210–212, Attribution, 5, 201–219
214, 231–233, 235, 252–254, 256, external, 203, 204, 206, 207, 213, 216–218
257, 271, 272, 274, 276–279, 285, internal, 203, 204, 206, 207, 213, 216–218
288, 291, 295, 309, 311, 314, 317 process, 203, 217
Agency, 105, 135, 285, 286, 291 Authenticity, 15, 23–25, 31, 34, 237, 239, 270,
Aging, 195 271
American Mosaic Project (AMP), 118, 120
Ancestry, 20
Anger, 265, 268, 272 B
Antiblack, 114, 117, 120–122 Behavior, 5, 13–16, 20, 21, 30, 42, 43, 49, 63,
Anti-racist/antiracist, 90, 105, 113 79, 87–89, 91, 92, 97, 100, 104, 105,
Research 6, https://doi.org/10.1007/978-3-030-76966-6
324 Index
111–113, 133, 134, 138, 148, 159, Children, 5, 14, 17, 66, 67, 75, 78, 87, 134,
163, 179, 180, 182–185, 187–196, 136–140, 142, 143, 147, 156, 159,
203, 204, 216, 219, 223–225, 228, 161, 163, 169, 172, 173, 175, 177,
238, 245, 247, 252, 260, 262, 289, 179–197, 226, 273, 286, 287, 295,
290, 292, 294, 305–307, 318, 320 305, 306
Behavioral expectations, 87, 305 Chile, 91
Behavioral outcomes, 56, 104 Christianity, 67
Behavior, devotional, 189 Christians, 67, 309
Beliefs, 21, 22, 31, 40, 50, 78, 89, 100, 104, Christie-Mizell, A., 4, 61, 64, 112, 113
113, 114, 121, 122, 179, 180, 182, Church/churches, 6, 71, 111, 139, 188, 196,
184, 224, 238, 287, 288, 297 246, 260, 289, 291, 295, 304–308,
Believers, 180, 181 310, 316, 318–320
Belonging, 4, 14, 16, 18, 20, 25, 32, 45–47, 49, Civil rights, 3, 68, 69, 71–74, 76, 78, 85, 114
62, 64, 68, 78, 113, 149, 170, 209, Civil Rights Movement, 68, 71, 78, 85, 114
220, 236, 237, 288, 311, 312, 315 Civil War, 111, 115, 125, 126
Belongingness, 4, 45, 138, 145, 147, 148 Closeness
Bible, 180 emotional, 247, 274–279, 311
Biracial, 226 group, 89
Birth, 68, 71, 137, 138, 304 Cognition, 43
Birth cohort, 68, 71 Cognitive, 40, 136, 140
Birthplace, 116 Cohabitation/cohabitiing, 66, 73, 78
Birthrate, 180 Collectivistic orientation, 148
Black/blacks, 17, 61, 86, 112, 170, 189, 231, College, 2, 3, 40, 42, 43, 56, 57, 62, 64, 65, 94,
311 119, 134, 177, 210, 226, 230, 253,
Black Lives Matter (BLM), 68, 69, 105 260, 261, 272, 274, 312
Blumer, H., 65, 88 Collett, J.L., 160, 163, 227
Bobo, L.D., 63, 89–92, 95, 117 Colonialism, 86
Bonding/binding mechanism, 247, 261 Colorblind/color-blind/colorblindness, 64, 78,
Bonilla-Silva, E., 12, 17, 32, 79, 89, 105, 112, 79, 89
113, 126 Communion, 136, 161
Brenner, P.S., 5, 7, 45, 62, 63, 88, 97, 112, 119, Community/communities, 5, 20, 21, 30, 32, 67,
139, 143, 179–196, 202, 207, 289, 79, 86, 116, 121, 122, 157, 167, 179,
303–320 180, 184, 186, 187, 193, 291, 304,
Burke, P.J., 12, 13, 40, 86, 111, 133, 155, 181, 307, 310–312, 315
201, 223, 245, 265, 286, 303 Competent/competence, 4, 133–135, 145,
Busing, school, 91 147–149, 164, 209
Confederacy, 111, 115, 116, 125
Confederate flags, 114, 116, 124, 125
C Confederate monuments, 4, 111–126
Caregiving, 226 Confederate symbols, 114, 116, 117
Caretaking, 162 Confirmation, 45, 184
Cast, A.D., 13–15, 17, 100, 134–136, 138, 140, Congruence, 64, 163
148, 155, 161, 163, 181, 227, 228, Connectedness, 269, 280
265–267, 269, 280, 294 Connections, network, 253, 269, 270, 273, 280,
Catholic, 181, 305 287
Caucasian, 20, 176, 177 Conservatism, political, 94
Cemeteries, 125 Conservatives, 90, 92, 94, 96, 97, 102, 103,
Charleston church shooting, 111 105, 296
Charlottesville protests, 126 Constraints, structural, 307
Childcare, 134, 138, 140, 172 Cooley, C.H., 13, 225, 227, 228
Childhood, 16, 161, 227 Core self, 150
Childlessness, 206 Costs, 91, 288, 306, 307
Childrearing, 155, 162, 163, 176 Counter-normality, 217
Index 325
Counter-normative, 5, 204–208, 216–219, 226, Efficacy-based esteem, 4, 44, 133–135, 140,

320 141, 144–149
Counter-performances, 181 Egocentric network, 249, 251
Counter-roles, 159, 307 Emotions, 6, 14, 42, 43, 92, 133, 136, 155, 160,
Co-workers/coworkers, 6, 65, 157, 224, 228, 162, 202, 207, 227, 229, 267, 271,
230, 233, 246, 251, 261, 273, 286 272, 275–281, 290, 298
Criminal self-view, 226, 228 Employment, 70, 116, 119, 140, 143–147,
Cultural artifacts, 4, 112, 114, 126 201–219, 233, 306
Culture, 2, 4, 12, 17, 18, 20, 21, 30–32, 41, 42, Enslaved people, 86, 97, 103
44, 46, 51, 55, 85, 121–123, 136, Equal Justice Initiative (EJI), 114–116
145, 149, 157, 161, 162, 165, 229, Esteem, worth-based, 4, 44, 45, 134, 135,
287, 291 138–141, 144–149
Cycle coding, 158, 159 Ethnic background, 12, 16, 18–22, 32, 49–51
Ethnic groups, 3, 4, 12, 14–20, 22, 25–33,
39–42, 44–52, 54–57, 63, 113, 134,
D 147, 149, 231
Dads, 166, 170, 172, 174 Ethnic identity, 2, 3, 7, 11, 12, 14–23, 25–33,
Daughters, 116, 125, 157, 174, 189 39–46, 48–52, 54–58, 66
Daycare, 171 Ethnicity, 1, 2, 4, 11, 12, 14, 15, 17–23, 25–33,
Deaux, K., 14, 15, 32, 40, 66, 78, 79, 88 39, 46, 49–51, 54, 56, 65, 149, 177,
De-friend, 296 181, 274, 279
Dehumanization, 125 Ethnocentrism, 298
DeLamater, J., 119, 182, 185, 189 Evaluations, 3, 39, 44–46, 48, 57, 66, 77, 133,
Delinquency, 133 136, 162, 203, 207, 231, 233, 237, 290
Democrats, 296 Evangelism, 180
Demo, D.H., 66, 67, 70, 78, 189 Expectations, 13, 16, 18, 31, 66, 68, 75, 87,
Density 104, 136–138, 149, 160, 163, 202,
intragroup, 248 205–208, 215, 224, 229, 237, 238,
network, 248, 265–266, 269–271, 275, 276, 247, 280, 289, 290
280 Experience sampling method (ESM), 187, 189
within-group, 248
Depression, 43, 196, 267, 288
Desegregation, 116 F
Development Faith, 180, 182, 184
child, 140 Family/families, 21, 43, 66, 118, 134, 155, 179,
identity, 2, 14, 41, 54, 57, 180, 186, 196 201, 224, 246, 268, 288, 318
Developmental process, 40, 41 Fares, P., 17, 65, 78, 87, 88, 113, 114, 116, 117,
Discrimination, 11, 39, 41, 56, 57, 65, 66, 114, 122, 125, 294
118, 149, 296, 298 Fatherhood, 156–158, 161–163, 170, 172–174,
Disenfranchisement, 125 176
Dissortativity, 252, 255 Fathers, 4, 5, 136, 138, 155–164, 166, 169–176,
Diversity, 90, 117, 118, 120, 122, 123, 149, 310 179–197
Domesticity, cult of, 147 Feedback, 5, 6, 13, 18, 57, 160, 202, 203, 205,
Durkheim, E., 138 216, 224, 228, 229, 236–239, 290,
Dyadic ties, 272, 274 297, 298
Feedback loop, 13, 57, 160, 290
Felson, R., 226, 229
E Female, 47, 51, 52, 54, 71, 91, 94, 120–122,
Echo chambers, 293, 297, 298 135, 137, 149, 162, 189, 191,
Education, 47, 51, 65, 69–74, 76, 77, 91, 94–96, 210–212, 214, 231–233, 235,
98, 99, 101, 116, 117, 121, 123, 140, 250–254, 257, 311, 317
177, 202, 206, 210, 231–233, 235, Feminine, 135, 137, 138
269, 274, 276–279, 291, 295, 307, Femininity, 137
311, 314, 317 Fertility, 149
326 Index
Festivals, 20, 21, 30, 50 Heider, F., 203, 204

Floyd, G., 105, 118, 126 Helms, J.E., 113
Friend identity, 233–236, 266, 271, 272, 274, Heritage dimension, ethnic identity, 2, 21, 22,
275, 280, 292, 294 26, 27, 30–32, 49
Friendship, 66, 249, 265, 266, 270–272, Heterogeneous, 86
274–279, 287 Hierarchical control, 55, 294, 295
Hiring, preferential, 3, 86, 94–96, 100
Hispanic/Latino, 118
G Hispanics, 4, 66, 70, 71, 134, 143, 144,
Gallagher, M., 5, 6, 160, 223–239 146–149
Gay, 297 Historical narratives, 116
Gecas, V., 88, 135, 203 Holidays, 20, 21, 30, 115, 117, 180, 191
Gender, 1, 21, 22, 46, 47, 50–54, 56, 65, 66, 69, Homogamy, 306
71, 77, 86, 88, 94, 116, 117, 121, Homogeneity, 248, 296, 306, 310, 312, 316,
134, 137, 138, 148, 162, 181, 191, 319
202, 206, 233, 246, 251, 252, Homogenous social structure, 3, 43, 55, 57, 308
254–256, 258, 259, 269, 274, 279, Homophilous ties, 43, 44, 293, 297
287, 289, 305, 311, 318 Homophily, 2, 6, 43, 57, 249–251, 285–298
General others, 5, 6, 224, 225, 228–230, Hughes, M., 3, 51, 61, 64–67, 69, 78, 89, 91,
233–235, 238, 239 114, 117, 125, 189
General Social Survey (GSS), 3, 62, 70, 93, 94, Hunt, M.O., 3, 12, 17, 43, 57, 61, 70, 85–105,
187, 251 228, 266, 285, 291
Generations, 3, 15, 16, 18, 23, 25–29, 31, 155, Husband identity, 137
162 Hymns, 184
Group affiliation, 116, 181, 226
Group-based identity, 12, 14, 16, 18, 25, 31, 88,
148 I
Group closeness, 89 Iconography, 111
Group distinctiveness, 63 Identity
Group identification, 14, 112, 113 abandoned, 2, 160, 174
Group identity, 4, 7, 29, 45, 55, 85, 89, 97, 112, achievement, 39–58
114, 116, 125, 137, 139, 181, 231, acquisition, 156, 175
236, 237, 288, 293, 295, 298 activation, 159
Group members, 14, 15, 18, 22, 31, 46, 61–65, attainment, 44, 46, 71, 119, 211
77, 79, 88, 248, 269 calling-up, 186
Group membership, 20, 30, 39–41, 49–51, 55, centrality, 68
66, 90, 97, 111, 137, 138, 145, 180, change, 5, 171, 172, 174, 175, 219, 245,
181, 248, 251, 254, 255, 280 248, 294
Group norms, 14, 15 commitment, 184, 195, 304, 307, 308, 310,
Group ties, 6, 295, 297 312
Guilt, 77, 207, 272 affective, 195, 308, 312
extensive, 79
intensive, 148, 195
H interactional, 195
Haidt, J., 296, 297 competition, 195
Happiness, 137 complementary, 290, 292
Harrod, M., 2, 5, 11–33, 43, 46, 48, 49, 51, 56, constructs, 50, 183, 189, 190, 195
64, 65, 136, 148, 201–219, 224, 228, counter-normative, 5, 204, 205, 207, 208,
231, 232, 237, 269, 274 216, 217, 219, 226, 320
Hays, S., 137, 141 crisis, 39
Head Start program, 135, 140 development, 2, 14, 16, 41, 54, 57, 180,
Health 186, 196
mental, 61, 78, 230 discrepancy/discrepancies, 49, 141, 142,
psychological, 39–41, 56 144–149, 224, 239, 271, 274–281
Index 327
enactment, 7, 12, 15, 104, 181, 184, 186, social, 3, 40, 44–46, 48, 55, 56, 62, 64, 66,
203, 318, 320 68, 77, 78, 86, 88, 90, 112, 134, 138,
engagement, 2, 5, 13, 158, 186, 201, 306 159, 181, 286, 288, 289, 293, 297
exploration, 16 standard, 13–15, 22, 42, 43, 224, 237, 267,
extensiveness, 68, 184, 247, 305 280, 281, 289, 290, 292, 294
formation, 39, 41, 57, 160, 193, 196, 304 theory, 12, 40, 62, 86, 111, 134, 155, 180,
group, 4, 30, 45, 55, 89, 97, 112, 114, 116, 201, 223, 245, 265, 287, 304
125, 137, 139, 231, 288, 290 verification, 2–4, 6, 13, 15–18, 27, 28,
group-based, 12, 14, 16, 18, 25, 31, 88, 148 30–32, 39–58, 64–70, 72, 73, 75–77,
hierarchy, 63, 113, 182, 203, 286, 305 134, 136, 138, 139, 148, 156, 160,
importance, 62, 63, 67, 69, 70, 72, 73, 75, 163, 223, 237, 286, 289, 290, 292,
76, 78 297, 298
invoking, 145 white, 91, 117
maintenance, 12, 65, 112, 126, 181, 227, Immigrants, 14–17
236 Immigration, 11, 32, 90, 105, 296
meanings, 6, 13, 19, 20, 22, 25–28, 31, 40, Impact identity, 202
42, 43, 45, 46, 55–57, 65, 133, 134, Income, 47, 51, 70, 94, 96, 98, 99, 101, 116,
136, 139, 148, 155, 162, 163, 201, 117, 119–122, 134, 137, 149, 202,
230, 286–290, 292–295, 297, 298, 206, 210–212, 214, 252–254, 257,
305 259, 310–312, 314, 315, 317
moral, 44, 55, 295–298 Inequality, 62, 134, 137, 319–321
multiple, 111, 182, 267, 286, 291, 293, 294, In-group/Ingroups, 6, 14, 42, 43, 55, 66, 70, 73,
297 88, 105, 114, 115, 125, 138, 236, 237
non-verification/nonverification, 2, 6, 14, Injustices, 64, 105, 298
15, 31, 32, 49, 143, 160, 224, 279, Intensity, 87, 267, 269, 279
280, 290 Interaction partner, 148, 307
normative, 5, 216 Interactional commitment, 184, 195, 247, 305,
parental, 139, 189 308, 312, 316
performance, 15 Interactive commitment, 305, 308–310, 313,
pride, 20, 67–70, 73, 75–78, 88, 100, 117 315–319
processes, 44, 45, 139, 176, 183, 185, 196, Intergenerational transmission, 5, 180, 181, 183
202, 224, 225, 229, 230, 245, 246, Intergroup conflict, 4, 114, 126, 298
248, 260–262, 266, 269, 287, 309, Intergroup contacts, 298
319, 320 Intergroup ties, 248, 259, 261
prominence, 90, 112, 135, 160, 181, 186, Interracial contacts, 117
202
prominence hierarchy, 113, 182, 203, 305
racial, 3, 4, 20, 22, 61–62, 75, 76, 85–105, J
111–126, 226 Jardina, A., 61, 64, 65, 69, 77, 79, 86, 89, 92,
religious, 5–7, 179–196, 233–236, 304, 104, 105
307–311, 315, 316, 318–320 Jim Crow, 114, 125
role, 4, 13, 17, 44, 55, 65, 66, 79, 86, 87, Jimenez, T.R., 11, 12, 14–16, 18, 20, 30, 32
133–136, 139, 147, 150, 159, 181,
236, 237, 245, 247, 248, 286, 288,
289, 293, 295, 298 K
salience, 63, 66–70, 72, 73, 75, 76, 88, 139, Kalkhoff, W., 224, 229, 238
143, 144, 146, 147, 159, 160, 165, Khanna, N., 14–16, 18, 30, 32, 89, 226
182, 184–187, 190, 192–196, Kiecolt, J.K., 3, 61–79, 114, 125
245–248, 259, 260, 289, 304, 305, Kinch, J.W., 223, 225, 226, 239
311, 312, 315, 316, 319 Knowles, E.D., 89, 90, 103
salience hierarchies, 87, 113, 182, 196, 305 Kuhn, M.H., 225, 227, 228
328 Index
L Mothers, 4, 5, 51, 133–150, 159, 163, 170,

Latinos/Latinas, 3, 12, 19, 26, 28, 31, 32, 51, 52 179–197, 226, 268
Leisure, 134, 138 Mt. Rushmore, 116
Liberals, political, 94, 97, 296 Multidimensional Model of Racial Identity
Loneliness, 68, 71–76, 78, 79 (MMRI), 63
Luck, 202, 204, 207–219 Multigroup Ethnic Identity Measure (MEIM),
Lynching, 85 16, 30, 49, 50
Lynn, F.B., 245, 247, 248, 251, 259–262 Multiplexity, 245, 261
Museums, 85, 125
M
Macro-sociological, 161, 162 N
Macrostructural, 62 Nationalism, 117, 125
Marcussen, K., 224, 229, 238 Native Americans, 47, 87, 143, 144
Marital status, 140, 143, 233, 252 Navy, 167, 172
Markowski, K.L., 6, 44, 56, 147, 202, 205–207, Network analysis, 6, 247, 249, 260, 261, 266,
219, 245, 261, 262, 309 269, 280
Martin, T., 66, 69 Network buffer, 270
Masculinity, 117, 162 Network clustering, 248
aggressive, 162 Network clusters, 246, 248
Master identities, 7, 286, 295, 297, 298 Network context, 271
Mastery, 33, 135, 237 Network density, 248, 265–281
McCall, G.J., 62, 88, 112, 159, 182, 203, 265, Network ties, 295
266, 268, 279, 305 Node (network), 252, 287
Mead, G.H., 13, 30, 87, 113, 223, 225, 227, Normative, 5, 32, 88, 136, 160, 182, 185, 189,
229, 303, 305, 306 204–208, 216–219, 225, 320
Meanings, 4–7, 12–16, 18–22, 25–28, 30–32, Norms, 15, 31, 88, 179, 184, 218, 224, 227,
39–46, 49, 50, 55–57, 62–65, 68, 77, 229, 237, 247, 248, 306
79, 87, 111–114, 133–142, 144, 145,
148, 149, 155, 158–160, 162, 163,
180, 181, 201, 202, 216, 217, 223– O
225, 230, 236–239, 245, 261, 265, Obama, B., 68, 71, 90
267, 269, 274, 281, 286–298, 305 Obligatory identities, 4, 5, 155–176, 231
internalized, 202, 217, 305 Occupation, 177, 202
Mechanical integration, 138, 139 Occupational prestige, 70
Medicare, 90 Opinions, 22, 103, 226, 231, 249, 288
Memorial for Peace and Justice (MPJ), 85 Oppression, 85, 296
Memorials, 85, 86, 93–98, 102–104 Organizations, 63, 66, 115, 121, 125, 161, 249,
Men, 3–5, 47, 50, 51, 69, 76, 77, 117, 156, 160– 259, 291, 307, 308
163, 169, 227, 253, 256, 274 Osgood, C.E., 13, 288
Mental health, 61, 78, 230 Out-group/outgroup, 14, 88, 293, 298
Merolla, D.M., 43, 228, 266, 268, 269, 280, Owens, T.J., 201, 202, 205, 216–218, 305, 311
291, 306, 308, 310, 311
Meta-analysis, 204, 216
Millennials, 68, 69, 71–74, 76, 78, 79 P
Minority groups, 17, 27, 63, 64, 77, 78, 113, Parenthood, 66, 73, 75, 78, 134, 186
134 Parenting, 137, 162, 163
Minority, racial, 63, 64, 77, 78, 87 Parents, 4, 17, 23, 44, 66, 67, 71–75, 78, 86, 87,
Moral identities, 44, 55, 56, 295, 297, 298 136, 137, 139, 140, 157–160, 163,
Morals, 7, 44, 65, 103, 137, 142, 143, 145, 181, 169, 179–187, 189, 190, 192–196,
286, 287, 293, 295–297 206, 226, 227, 229, 273, 286, 289,
Mortality, 288 291, 293–295, 304, 306
Mosques, 304 Parishioners, 181, 305
Motherhood, 136–138, 147–149 Partisan politics, 7, 296, 298
Mothering, 4, 134, 136–138, 142, 145, 148, 149 Pastors, 306
Index 329
People of color, 17, 39, 41, 51, 56, 57 Reflected appraisals, 13, 22, 42, 43, 49, 55, 56,
Perceptions, 13, 15, 22, 78, 86, 88, 91, 93, 104, 64, 78, 88, 136, 141–143, 155, 203,
105, 112, 117, 118, 160, 201, 204, 223–239, 274, 276, 281, 292, 294
217, 219, 223, 225, 229, 232, 237, Reichelmann, A.V., 3, 61, 70, 85–105, 117
239, 265, 288, 290, 294, 296 Relational positioning, 7, 304–307, 309, 310,
Perceptual control processes, 13–14, 229, 236, 312, 313, 315–320
239, 267, 280 Religions, 2, 180, 247, 251, 252, 255, 256, 258,
Perceptual control system, 13, 290 260, 287, 291, 304, 311, 312, 315
Perpetrator, 96, 102 Religiosity, 180, 181, 186, 196
Pew Research Center, 162 Religious, 5, 7, 20, 21, 44, 50, 66, 67, 72–76,
Phinney, J.S., 15, 16, 18–20, 30, 39–41, 49–51, 78, 179–197, 206, 225, 231–235,
55, 56, 61, 89 248, 249, 251, 252, 254, 256, 260,
Physical environments, 2, 6, 7, 303–309, 318, 261, 268, 287–289, 293, 296, 304,
320 307, 308, 310–312, 315, 318–320
Physical health, 168, 271 Religious congregants, 249
Physical locations, 303 Religious congregation, 184
Polarization, 293, 297, 298 Religious identities, 5–7, 179–197, 233–236,
Political groups, 296, 297 304, 307–311, 315, 316, 318–320
Political identities, 7, 104, 296, 297 Religious service attendance, 252, 258
Political party affiliation, 116, 117 Reparations, 91, 93
Politics, 2, 104, 287, 296–298 Republicans, 117, 120–122, 296, 297
Portes, A., 11, 14, 17, 23, 31 Resilience, 39, 57
Poverty, 122, 140 Resources, 17, 27, 62, 64, 65, 90, 114, 125, 148,
Prayer, praying, 180, 184, 189, 193 173, 227, 266, 269–271, 280, 291,
Preferential hiring, 3, 86, 94–96, 100 292, 303, 304, 306, 307, 310
Prejudice, 39, 41, 56, 57, 88–90, 117, 120–122 interpersonal, 148, 266, 269–271, 280
Presidency, 68, 71, 90 Rituals, 14, 180, 184, 193, 195
President, 69, 71, 116 Role-based identities, 138
Private regard, 51, 64, 72 Role behavior, 137, 255, 256, 259
Private self-regard, 87, 88, 90, 93, 96, 97, 100, Role conflict, 163
101, 103, 104 Role differentiation, 149
Psychological centrality, 40, 113 Role identities, 4, 13, 17, 44, 55, 65, 66, 79, 86,
87, 133–136, 138, 139, 147, 150,
159, 181, 236, 237, 245, 247, 248,
R 286, 288, 289, 293, 295, 298
Race/ethnicity, 1, 11, 14, 15, 17, 19, 22, 26–30, Role-involvement, 246, 252, 253, 255, 256, 260
32, 65, 274, 279 Role-memberships, 246
Races, 1–4, 11, 12, 14–17, 19, 22, 26–30, 32, Role-participation, 260, 262
33, 62–65, 68, 70–72, 75–77, 79, 86, Role partners, 227, 228, 230, 232, 236–239
88, 93, 100, 113, 134, 140, 143, 148, Role playing, 288
149, 177, 181, 189, 191, 202, 226, Role-taking, 225
233, 235, 238, 246, 251, 252, 254, Rosenberg Self-esteem (RSE) scale, 5, 33, 140,
255, 258, 259, 274, 279, 285, 287, 208
291, 295, 305, 311, 318, 319 Rosenberg, M., 5, 14, 33, 45, 57, 62, 64, 87,
Racial groups, 4, 13, 61–67, 69, 70, 73, 75–79, 113, 133, 135, 139, 140, 203, 204,
86, 89, 93, 97, 100, 103, 112–114, 207, 208, 217, 218
116 Rumbaut, R.G., 11, 14, 17, 23, 31
Racial identities, 3, 4, 20, 22, 61–73, 75–79,
85–105, 111–126
Racial inequalities, 3, 4, 61, 62, 64, 67, 68, 70, S
78, 79, 90, 93, 94, 100, 103, 115, 296 Schemas, 114
Racial integration, 91 Schooling, 233
Racism, 62, 89, 126 Schoolmates, 6, 246, 306
330 Index
Schools, 47, 56, 91, 94, 111, 119, 121, 125, Self-structure, 261, 319, 320
162, 168, 172, 210, 231, 246, 249, Self-understandings, 30, 103, 105
251–253, 260, 274, 285, 287, 288, Self-verification, 205
291, 295, 312 Self-view, 15, 187, 203, 216, 217, 232–234,
Schoolwork, 255, 258, 259, 292 237, 267, 274, 276, 289
Scripture, 193, 195 Self-worth, 3, 14, 23–25, 31, 33, 39–58, 134,
Second World War (WWII), 91, 196 135, 140, 237, 239
Segregation, 117, 259, 305, 318 group-specific, 40, 42, 45, 46, 48, 51, 52,
Self, 1–6, 13–33, 39–57, 61–78, 86–104, 55, 56, 58
111–124, 133–150, 155–176, 181– self-worth, global, 40, 42, 46, 48, 50–57
187, 201–219, 223–237, 245–261, Serpe, R., 5–7, 12, 43–56, 62–65, 86–97, 112–
266–280, 286–294, 304–320 119, 135–147, 156–176, 180–195,
Self-appraisals, 292 201–220, 245, 266–272, 289–291,
Self-assessments, 18 304–318
Self-attribution, 201–204, 215 Sex, 143, 189, 210, 214, 310
Self-authenticity, 44, 55 Sex-role, 149
Self-behavior, 207 Sexuality, 181, 297
Self-categorization, 61 Shared expectations, 87, 202
Self-concept, 39, 42, 62, 63, 87, 88, 104, 126, Siblings, 159, 195, 273
202, 203 Simmons, J., 305
self-concept, global, 44, 46 Situational, 139, 203, 207, 289, 290
Self-confidence, 5, 205–208, 211, 212, Situational meanings, 139, 289, 290
214–220 Situations, 1, 2, 13, 14, 32, 42, 43, 63, 88, 112,
Self-consistency, 204, 205 136, 138, 139, 141, 145, 147, 148,
Self-definition, 225 159, 160, 165, 166, 181, 182, 185,
Self-denigrating, 205, 219 187, 193, 202, 223, 265, 267, 269,
Self-deprecation, 205–209, 211, 212, 215–220 270, 281, 286, 288–290, 294, 295,
Self-development, 155, 156, 159, 176 297, 298, 305
Self-efficacy, 14, 23, 33, 44, 135, 140, 237, 239 Skin color, 17
Self-enhancement, 136, 204, 205 Skin pigmentation, 17
Self-esteem, 2, 4, 5, 14, 17, 23, 25, 31, 39–46, Slavery, 3, 4, 85–105, 115, 125
48, 52, 56, 57, 61, 77, 78, 93, Sociability, 133
133–136, 138, 139, 141, 147, 148, Social class, 134, 285, 288, 291
155, 172, 201–219, 236, 237, 267, Social contacts, 6, 7, 245, 246, 248–250, 253,
268, 270, 288 261, 266, 267, 269, 280, 319
self-esteem, global, 44–46, 57, 204 Social context, 6, 18, 112, 195, 266, 268–270,
Self-evaluation, 205, 230 280, 297
Self-feelings, 139, 218 Social desirability bias, 78
Self-hierarchy, 319 Social distance, 88, 90, 117, 293, 296, 298
Self-identification, 90 Social groups, 12–14, 30, 31, 45, 48, 64, 88,
Self-image, 205, 209, 220, 224, 311, 312, 315 113, 114, 179, 247, 287, 290, 294,
Selfish/selfishness, 5, 142, 156, 159, 173, 174 295, 297, 298
Selfless/selflessness, 175 Social identity theory, 3, 48, 62, 64, 66, 77, 88
Self-liking, 136 Socialization, 66, 67, 78, 89, 116, 179–197, 229
Self-meanings, 12–14, 44, 87, 223 Social networks, 6, 18, 43, 65, 66, 228,
Self-motivations, 205 245–254, 259–262, 266, 269, 272,
Self-motives, 218 280, 287, 295, 304, 319
Self-outcomes, 265, 271 Social positions, 202, 227, 287, 305
Self-presentation, 30 Social psychology, 19, 105, 150, 181
Self-reflections, 174 Social structure
Self-regard, 87, 88, 93, 96, 97, 100, 101, 103–105 intermediate, 57, 201, 291, 303–321
Self-relevant, 13, 42, 87, 202, 203, 205, 217, large, 201, 305, 311, 318–320
236 proximal, 201
Index 331
proximate, 43, 44, 57, 268, 269, 280, 304– Stryker, S., 2, 12–32, 42–57, 61–79, 86–97,
308, 310, 315, 316, 318, 319 111–119, 139, 180–195, 201–218,
Social ties, 68, 78, 245–247, 249, 250, 255, 227–228, 245–261, 265–280,
259–261, 285–287, 292–294, 297 285–291, 304–320
Societies, 1, 4, 7, 12, 13, 17, 32, 44, 62, 65, 68, Students, 2, 3, 6, 19, 20, 40, 42–44, 47, 57, 62,
78, 79, 86, 87, 91, 104, 111, 113, 64–66, 86, 87, 94, 157, 159, 160,
121, 125, 134–138, 148–150, 155, 168, 181, 195, 202, 205, 210, 246,
159, 176, 201, 202, 223, 229, 238, 247, 249, 251–253, 255, 256, 260,
245–262, 265, 266, 287, 296, 305, 261, 268, 287, 291–293, 295
318–320 Substance abuse, 133
Socioeconomic status (SES), 46, 50–54, 56, Sullivan, H., 223, 225
116, 171, 172, 175, 259 Surfer-dads, 4, 156–176
Sociology, 2, 15, 19, 30, 47, 63, 105, 157, 180, Surfers/surfing, 4, 5, 155–176
265 Survey bias, 320
Solidarity, 4, 20, 137, 149 Survey Sampling International (SSI), 92
Sons, 115, 116, 125, 168, 170, 173, 189, 196 Surveys, 19, 47, 70, 86, 92, 118, 125, 135, 140,
Southern, 76, 115, 116, 122, 157, 162 142, 182, 185, 208, 230, 231, 239,
Southerners, 68, 78, 116 253, 261, 309, 310, 320
Southern Poverty Law Center, 111, 115, 118, Swann, W., 135, 136, 205
121, 124, 125 Symbolic interactionism, 87, 88, 234, 303, 307
Spanish (language), 15 Symbolic interactionists, 6, 12, 19, 87, 111,
Sport/sports, 173 225, 227, 228, 230, 236, 305
Spouse, 4–6, 66, 136, 156, 159, 160, 163, 164, Symbols, 115, 124, 125, 290
167–170, 174, 175, 224–229, 231,
232, 234–236, 268, 271–273, 286,
294, 295, 306 T
Spouse/spousal identity, 4–6, 66, 136, 156, 159, Talbert, R.D., 4, 112, 114–117
160, 163, 167–170, 174, 175, 224– Task difficulty, 5, 202, 204, 206–211, 213–218
229, 231, 232, 234, 236, 271–273, Taylor, B., 118, 126
286, 294, 306 Teachers, 13, 87, 195, 226, 292
Statues, 111, 112, 124–126 Teaching, 136, 180, 184
Status, 12, 17, 23, 27, 28, 32, 56, 61, 63–65, 70, Teammates, 226
75, 77, 85, 86, 89–91, 113, 116, 119, Telephone, 118, 208, 230
121, 134, 137, 143, 148, 149, 158, Television, 196
170, 177, 201–219, 224, 228, 233, Thoits, P., 138, 139, 156, 160, 231, 289
238, 252–254, 257, 269, 274, 276, Tichenor, V., 134, 149
287, 288, 291, 294, 295, 319 Traditions, 2, 11, 15, 18, 20–22, 28–30, 32,
characteristics, 287, 288, 291 49–51, 111, 180, 193
hierarchy, 113 Travel, 172, 250, 303, 304, 306, 307, 310, 312,
Status-based, 56 315
Step-parents, 273 Trump, D., 85
Step-siblings, 273 Trust, 78, 105, 248, 266, 269, 288, 292, 293, 298
Stereotypes, 14, 15, 41, 46, 55, 64, 65, 114,
138, 298
Stets, J., 1–7, 12–32, 40–57, 62–78, 86–88, U
111–125, 133–150, 155–176, 181, Undergraduates, 19, 47, 57
201–219, 223–238, 265–281, Unemployed, 5, 117, 143, 202, 203, 205–220,
285–298, 305–318 233
Stratification, 12, 89, 91, 291, 319 United States (U.S.), 7, 70, 85, 86, 89, 100, 103,
Structural symbolic interactionism, 42, 245, 105, 115, 117, 122, 125, 134, 231,
266 296
332 Index
Universities, 2, 3, 6, 19, 47, 66, 94, 96, 100, Whiteness, 61–63, 89, 90, 97, 104, 105,
105, 125, 150, 168, 172, 246, 112–119, 121, 123–126
249–253, 260, 261, 309 White racial identities (WRI), 86, 87, 89–92,
U.S. government, 93, 94 95, 97, 100, 103–105
Whites, 3, 4, 7, 12, 16, 17, 19, 23–29, 31, 32,
47, 50–54, 56, 57, 61–80, 85–105,
V 111–126, 143, 144, 147, 149, 176,
Values, 16, 40, 41, 50, 62, 65, 71, 94, 97, 177, 189, 191, 192, 231–233, 250,
121–123, 135, 139, 143–145, 149, 252–254, 256, 257, 259, 274, 311
179, 180, 182, 184–186, 192–196, White supremacy, 4, 67, 111, 112, 114, 115,
209, 210, 213, 227, 231, 233, 247, 118, 125
252–254, 278, 279, 285, 287–289, Women, 3, 4, 47, 69, 72, 74, 76, 77, 134–138,
293, 295, 297, 298, 304, 305, 308 144, 145, 147, 149, 253, 256, 272,
Victims, 85, 86, 93–99, 102, 104 274
Violence, 85, 105 Worker identities, 4, 5, 167, 168, 171–172, 174,
Visual, 92, 94–96, 98, 99, 102, 103 175, 201–203, 205, 207, 218, 224,
Voluntary identities, 4, 156, 159–161, 167, 169, 231, 233, 234, 291–293
174, 176, 231 Workers, 4–6, 65, 86, 87, 136, 143, 156, 159,
Voting, 90, 105 160, 164, 169, 174, 206, 209, 210,
224, 225, 228, 231, 232, 234–236,
286, 295
W Working, 15, 33, 137, 143, 157, 165–167, 202,
Walker, M.H., 245, 247, 248, 251, 259, 260, 207, 210, 213, 217, 267
262, 270, 273 Workplaces, 86, 100, 249, 256, 291
Waves, 85, 157, 158, 162, 164–166, 173, 174 Worship, 67, 180, 184, 307–310, 316, 320
Wealth, 65
Weiner, B., 202, 209
Welfare, 140, 296 Y
Well-being, 13, 41, 79, 155, 160, 163, 245, 274, Yarrison, F., 43, 267–269, 291, 306, 308–311,
288 315, 320
We-ness, 293, 298 Youths, 57, 63

Identities in Action: Philip S. Brenner Jan E. Stets Richard T. Serpe Editors

Uploaded by

Copyright:

Available Formats

Identities in Action: Philip S. Brenner Jan E. Stets Richard T. Serpe Editors

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Identities in Action: Philip S. Brenner Jan E. Stets Richard T. Serpe Editors

Uploaded by

Copyright:

Available Formats

Frontiers in Sociology and Social Research 6

Series Editor: Richard T. Serpe

More information about this series at http://www.springer.com/series/8690

ISSN 2523-3424 ISSN 2523-3432 (electronic)

1 Introduction: An Overview of Identities in Action . . . . . . . . . . . . . . 1

Part I Ethnicity and Race

Part II Family, Religion, and Work

9 Religious Identity in a Proximate Social Structure: Mothers,

Part III Networks, Homophily, and the Physical Environment

John Aldecoa Department of Sociology, California State University, Sacramento,

Ashley V. Reichelmann Virginia Tech, Blacksburg, VA, USA

Philip S. Brenner, Jan E. Stets, and Richard T. Serpe

“Identity” is an important concept in the social sciences and humanities, widely

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 1

self-esteem; whether the effects of non-veriﬁcation occur uniformly for different

veriﬁcation also should increase the prominence and salience of identities in

Peter J. Burke and Michael M. Harrod

The social psychological components of actors’ racial/ethnic1 identity represent a

Prepared for the 2018 Identity Conference, Boston, MA.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 11

Perceptual Control Process

Identity theory provides an in-depth explanation of the multifaceted nature of an

identity veriﬁcation, which produces positive emotions as well as increased feelings

The Research Questions

We explore two aspects of ethnic identity informed by identity theory. First, we

negative impact of nonveriﬁcation on the three components of self-esteem and

third-generation respondents (Gans, 1992; Jimenez, 2004; Kasinitz et al., 2004;

eating/preparing ethnic foods; looking like and being correctly identiﬁed as a

Table 2.2 Means and stan- Variable Mean Std. dev.

Table 2.3 Zero-order correlations among variables (n ¼ 775)

Table 2.5 Mean heritage Generation

Table 2.6 Tests of effects β p

Table 2.7 Mean personal Generation

Table 2.8 Tests of effects β p

Table 2.9 Mean ethnic iden- Generation

by generation on the average. There are, however, signiﬁcant interaction effects of

Table 2.10 Tests of effects β p

Table 2.11 Mean ethnic Generation

Table 2.12 Tests of effects p

Table 2.13 Mean ethnic Generation

Table 2.14 Tests of effects p

Discussion and Conclusion

Factor Analysis of Self-Worth Scale Items

Factor Analysis of Self-Efﬁcacy Scale Items

Factor Analysis of Authenticity Scale Items

Matthew Grindal, Melanie Kushida, and Tanya Nieri

People of color have historically experienced structural impediments as well as

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 39

Research examining the impact of ethnic identity achievement conﬁrms that it

Ethnic Identity Achievement

Identity Theory, Veriﬁcation, and Self-esteem

Rooted in structural symbolic interactionism (Stryker, 1980), identity theory has

Self-esteem has traditionally been deﬁned and studied as a unidimensional construct

religious category, or membership in a group such as membership in a union or local

Summary and Hypotheses

Fig. 3.1 Theoretical Model

We included respondents in our analysis if they identiﬁed as white, Asian, African

Table 3.1 Factor Loadings for Global Self-Worth Items

* p < 0.05, p < 0.01, * p < 0.001