Retirement Employee
Retirement Employee
Retirement Employee
University of Maryland
Kenneth S. Shultz
Despite the obvious importance of retirement to employees, their employing organizations, and
the larger society, the last comprehensive review of employee retirement in the field of organizational science was published more than 20 years ago. As such, the first purpose of this review
is to provide a summary of key theoretical and empirical developments in employee retirement
research since Beehr in 1986. A second purpose of this review is to highlight inconsistent findings revealed by studies that were designed to answer the same research questions. By identifying and scrutinizing those inconsistent findings, this study expects to provide suggestions and
recommendations to further the theoretical development in the field of retirement research to
address these research gaps. As a result, this proposed review would be of interest to scholars
in a wide variety of areas within the organizational sciences, including human resource management, organizational behavior, organizational theory, and research methods.
Keywords: retirement; retirement planning; early retirement; retirement transition and adjustment; bridge employment; aging
Demographic projections have shown that by 2012, nearly 20% of the total U.S. workforce will be age 55 years or older, up from just less than 13% in 2000, leading to a sizable
increase in the number of people who will transition into retirement in the next decade
(Toossi, 2004). This reflects the fact that the population as a whole is getting older due to
several factors, including the aging of the large Baby Boom generation, lower birth rates, and
Corresponding author: Mo Wang, Department of Psychology, University of Maryland, College Park, MD 20742, USA
E-mail: [email protected]
Journal of Management, Vol. 36 No. 1, January 2010 172-206
DOI: 10.1177/0149206309347957
2010 Southern Management Association. All rights reserved.
172
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
longer life expectancies (Alley & Crimmins, 2007). These demographic and labor force
change patterns are also demonstrated by data from other countries and regions (e.g.,
Western Europe, Japan, China, and India; Tyers & Shi, 2007), suggesting retirement is an
area of global significance. Furthermore, retirement poses significant adjustment challenges
for older employees (Quick & Moen, 1998). Maladjustment to retirement has been shown to
lead to increased alcohol use (Perreira & Sloan, 2001) and decreased mental health (Wang,
2007).
Despite the obvious importance of retirement to employees, their families and employing
organizations, and the larger society, the last comprehensive review of employee retirement
in the field of management and organizational scienceBeehr (1986) appeared in Personnel
Psychologywas published more than 20 years ago. Although there have been other reviews
that focused on aging and work issues (e.g., Hedge, Borman, & Lammlein, 2006; Kanfer
& Ackerman, 2004; Shultz & Adams, 2007), none of them has thoroughly examined
employee retirement and incorporated new conceptualizations and theoretical perspectives
that could be applied to understanding retirement processes as well as the consequences of
retirement since Beehr (1986). Guided by these new conceptualizations and theoretical perspectives, researchers have indeed made significant advances in understanding factors that
influence retirement decision making and unfolding the retirement processes to reveal both
long- and short-term consequences of retirement. Nevertheless, there are also inconsistent
findings and research gaps that call for a critical review and creative theoretical and methodological solutions for further examination.
Furthermore, the context within which employees retire has changed significantly in the
past 20 years. For example, employees who retire now have to work longer to receive the full
Social Security benefit. In addition, most employers in the private sector have replaced
defined-benefit pensions with defined-contribution plans, which put investment risks on the
employees side. Most employers are also eliminating health care benefits for retirees. The
U.S. Department of Commerce (2005) also reported that the private saving rates have been
decreasing over the past two decades and reached an all-time low since the Great Depression.
Changes such as these have altered the research questions investigators ask, drawing more
attention to topics such as bridge employment and retirement planning (Cahill, Giandrea, &
Quinn, 2006), which were not covered by Beehrs (1986) review.
As such, the first purpose of this review is to provide a summary of key developments
in employee retirement research since Beehr (1986). Specifically, both theoretical and
empirical advancements since Beehr are reviewed and critically analyzed. A second purpose of this review is to highlight inconsistent findings and research gaps revealed by
empirical studies. By identifying and scrutinizing those inconsistent findings and research
gaps, we provide suggestions and recommendations to further the theoretical development
in the field of retirement research in order to address these issues. As a result, we expect
this review to serve the interest of scholars in a wide variety of areas within the organizational sciences, including human resource management, organizational behavior, organizational theory, and research methods. In addition, this review may also inform research
outside of the management literature, such as research in development psychology, gerontology, sociology, and economics, as retirement is also an important topic in those fields.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Table 1
Retirement Conceptualizations, Corresponding Theories, and Research Examples
Conceptualizations
Corresponding Theories
Research Examples
Hatcher (2003)
Feldman (1994)
Talaga and Beehr (1995)
Cron, Jackofsky, and Slocum (1993)
Kim (2003)
Retirement as an adjustment
process
Wang (2007)
Kim and Feldman (2000)
Adams, Prescher, Beehr, & Lepisto (2002)
Retirement as a career
development stage
hypothesis has received mixed empirical support (see our later review of the empirical findings regarding bridge employment), it highlights the importance of the retirement decision
as a major life event and illustrates some normative motivations for people to retire (Adams
& Beehr, 1998; Shultz, Morton, & Weckerle, 1998).
When conceptualizing retirement as decision making, researchers have typically relied on
the informed decision-making approach to conduct their investigations. This approach
assumes that older workers make their retirement decisions based on information they have
regarding their own characteristics and their work and nonwork environment. They weigh
these factors and evaluate the overall utility of retirement before they reach the decision
about whether to retire. Theoretical mechanisms that link these factors to retirement decisions (i.e., explicating why these factors are important in influencing retirement decisions)
include rational choice theory (Gustman & Steinmeier, 1986), image theory (Beach &
Frederickson, 1989), role theory (Ashforth, 2001; Moen, Dempster-McClain, & Williams,
1992), theory of planned behavior (Ajzen, 1991), and expectancy theory (Vroom, 1964).
Specifically, rational choice theory has typically been used to tie older workers financial
status as well as the external economic environment to their retirement decisions (Hatcher,
2003; Quinn, Burkhauser, & Myers, 1990). It argues that the retirement decision is a result
of comparing the financial resource accumulated and financial resource needed in retirement. Rationally, workers will retire only when they feel that their accumulated financial
resources, as well as the forecast of future economic conditions, would allow them to support
their consumption needs in retirement.
Both image theory and role theory have been used to tie workers demographic status,
work experience, marital life, type of industries, and productivity to their retirement decision making (e.g., Adams et al., 2002; Anson, Antonovsky, Sagy, & Adler, 1989; Feldman,
1994; Mears, Kendall, Katona, Pashley, & Pajak, 2004; Talaga & Beehr, 1995). In general,
all these factors influence how people perceive themselves and their roles in the largersocietal context, which creates comparison standards for workers to evaluate whether the
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
action of retirement matches their self-images or roles (Brougham & Walsh, 2007). If the
match is achieved, then the decision will be made to retire. However, it should be noted that
the foci of these two theories are slightly different. Image theory typically emphasizes that
the retirement decision is reached by upholding a stable self-image in resolving decision
conflicts (i.e., retire or not retire; Feldman, 1994), whereas role theory also underscores the
retirement decision as a result of active pursuit of role change that associates with transition
into retirement (Barnes-Farrell, 2003). Therefore, in comparison to image theory, theoretical mechanisms advanced by role theory tend to be more future-focused than presentfocused.
The theory of planned behavior has typically been used to link workers work attitudes
(e.g., job satisfaction, organizational commitment), retirement attitudes, and workplace
norms to their retirement decision (e.g., Adams & Beehr, 1998; Cron, Jackofsky, & Slocum,
1993; Huuhtanen & Piispa, 1992; Shultz, Taylor, & Morrison, 2003; Wang, Zhan, Liu, &
Shultz, 2008). The general premise of this theory highlights the importance of workers
attitudes toward retirement and its alternativecontinuing workingin influencing their
retirement decisions. Furthermore, it also emphasizes the role of perceived social pressure to
retire (i.e., norms) in influencing an individuals retirement decision.
Finally, the theoretical linkage from workers productivity, job characteristics, and health
status to their retirement decision is often built by endorsing mechanisms following expectancy theory (e.g., Belgrave & Haug, 1995; Cron et al., 1993; DeVaney & Kim, 2003;
Karpansalo, Manninen, Kauhanen, Lakka, & Salonen, 2004; Kim, 2003). Specifically, it is
argued that when retirees perceive low expectancy for reaching good productivity or being
rewarded for it, they are more likely to retire instead of continuing to work.
It is interesting to note that although conceptualizing retirement as decision making significantly enriched the theoretical mechanisms in explicating predictors of the retirement
decision in the past 20 years, very few studies investigating outcomes of the retirement decision have endorsed the same conceptualization. In other words, very few studies that examine outcomes of retirement have incorporated factors that influenced the original retirement
decision. In some cases, even when these factors (e.g., preretirement financial status and job
satisfaction) were included (e.g., Wang, 2007; Zhan, Wang, Liu, & Shultz, in press), the
theoretical mechanisms that established the links between these factors and retirement outcomes had nothing to do with the original decision-making process of retirement. This creates a logic gap because the reasons why people decide to retire would naturally influence how
they evaluate outcomes associated with their retirement. Addressing this gap by explicitly
incorporating the retirement decision-making process in understanding pathways that lead
to different retirement outcomes could significantly extend the research.
One limitation of the decision-making conceptualization of retirement is that not all
retirement decisions are voluntary (e.g., Gallo, Bradley, Siegel, & Kasl, 2000; Hanisch &
Hulin, 1990; Shultz et al., 1998; Szinovacz & Davey, 2005; van Solinge & Henkens,
2007). As such, the theoretical utility of this conceptualization depends on the extent to
which the retirement decision is a result of personal choice. If the personal choice component is missing, then the informed decision-making approach may not apply. Therefore,
the voluntariness of the retirement decision could be viewed as a boundary condition for
applying the informed decision-making approach in testing predictors of the retirement
decision.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
p reretirement job attitudes, job characteristics, and career standings; Wang, 2007) as well as
social context (e.g., social network and family structure). The social context is particularly
important because resources embedded in social context, such as social support (e.g., Taylor,
Goldberg, Shore, & Lipka, 2008), often help retirees to adjust to retirement. On the other
hand, other factors manifested in social context, such as negative stereotypes toward retirement (e.g., Settersten & Hagestad, 1996), may hinder the adjustment to retirement. Retirees
experience in other life spheres (e.g., marital life; Rosenkoeter & Garris, 1998; Szinovacz &
Davey, 2004) is also important, providing alternative salient identities after retirement and
offering opportunities for postretirement engagement (Bosse, Aldwin, Levenson, &
Workman-Daniels, 1991; Calasanti, 1996; Reitzes, Mutran, & Fernandez, 1996).
Furthermore, the life course perspective provides specific hypotheses on the shape of the
posttransition developmental trajectories. Life course theorists (Levinson, 1986; Levinson &
Levinson, 1996; Super, 1990) suggest the normative later life stages being characterized by
movement to activities and roles that involve less responsibility to others and less rigorous
physical effort (e.g., leisure activities). Accordingly, the resource demands associated with
retirees reorganizing their activities and time structures to achieve adjustment would gradually decrease over time (Reitzes & Mutran, 2004; van Solinge & Henkens, 2008). Thus, the
life course perspective predicts a positive developmental path in postretirement life.
In addition to the life course perspective, researchers also often use other theories to study
retirement as an adjustment process, such as continuity theory and role theory. Specifically,
continuity theory emphasizes that human beings general tendency is to maintain consistency
in life patterns over time and to accommodate transitions without the experience of disruption (Atchley, 1989, 1999). Therefore, continuity theorists view retirement adjustment as an
opportunity for retirees to maintain and continue their social relationships and lifestyle patterns into their retirement. Consequently, continuity theory predicts that only severe difficulty in maintaining life patterns would lead to undesirable transition quality and
unsuccessful adjustment to retirement. Variables that have been investigated within this theoretical framework include health and financial status change during retirement transition (e.g.,
Gallo et al., 2000; Wang, 2007), bridge employment (e.g., Kim & Feldman, 2000; Zhan et al.,
in press), functional capacity (Chirikos & Nestel, 1989), retirement planning (Fretz, Kluge,
Ossana, Jones, & Merikangas, 1989; Taylor-Carter, Cook, & Weinberg, 1997), and transferability of skills (Spiegel & Shultz, 2003).
Role theory emphasizes the importance of role exit and role transition embedded in the
retirement process. According to Ashforth (2001), to the extent that one is highly invested in
a particular role, feelings of self-worth tend to be associated with the ability to carry out that
role in an effective manner. Therefore, being retired can be characterized as a role transition
(Moen et al., 1992; Riley & Riley, 1994), which may include the processes of losing or
weakening work roles, while strengthening the family and community member roles
(Barnes-Farrell, 2003). Furthermore, role theory argues that the role transition could lead to
either positive or negative adjustment consequences, depending on whether the role transition
is desirable or matches the individuals values and goals (e.g., Adams et al., 2002; Thoits,
1992;). Following role theory, antecedents that have been examined for retirement transition
and adjustment include role stressors (e.g., Lin & Hsieh, 2001), role identities (e.g., Taylor,
Shultz, Morrison, Spiegel, & Greene, 2007), and values and goals (e.g., Shultz et al., 1998).
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
At the organizational level, factors such as perceived organizational climate related to age
bias and discrimination (e.g., Finkelstein & Farrell, 2007; Posthuma & Campion, 2009),
organizational downsizings (e.g., Gallo et al., 2000), and declines in demand for certain
types of labor (e.g., AARP, 2005) are also likely to influence retirees career pursuit. For
example, whether the organization treats older workers with respect and dignity would
largely influence the retirees motivation to continue their career pursuit. Organizational
downsizing and declines in demand of traditional jobs may also discourage retirees careerrelated pursuit. Nevertheless, retirees who can function across different job roles, organizational hierarchies, and national boundaries may thrive in this business environment (Tyers &
Shi, 2007).
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Empirical Findings
We next review studies that answer different research questions about retirement. These
studies are organized around the major empirical themes of retirement planning, retirement
decision making, early retirement, retirement transition and adjustment, and bridge employment. Within each empirical theme, we discuss the antecedents and consequences associated
with the empirical phenomenon as well as analyze how the studies tie to the aforementioned
four theoretical conceptualizations. At the end of reviewing each empirical theme, we
also summarize the research gaps and provide recommendations for future investigations.
Figure 1 presents a model that summarizes the diverse issues considered by the empirical
studies reviewed below. The left side of the model has the various factors that impact the
retirement process. The right side of the model depicts how the retirement process unfolds
over time, and how the components of retirement process influence each other.
Retirement Planning
Research on retirement planning over the last two decades has primarily focused on how
planning for retirement influences retirement timing, postretirement satisfaction, and adjustment. As a result, the theoretical conceptualizations most employed are retirement as decision making and retirement as adjustment. For example, retirement planning predicts higher
levels of postretirement adjustment across a variety of occupational settings, such as the
public sector (Taylor & Doverspike, 2003), private sector (Mutran, Reitzes, & Fernandez,
1996), and military settings (Spiegel & Shultz, 2003). In addition, those who have prepared
for retirement and feel ready to make the transition are more likely to exit the workforce at
an earlier age (Reitzes, Mutran, & Fernandez, 1998; Taylor & Shore, 1995).
Taylor-Carter et al. (1997) found that preparation for retirement both formally and informally increased subjects confidence in their abilities in making the retirement transition (for
similar findings, see Elder & Rudolph, 1999, and Anderson & Weber, 1993). Specifically,
formal retirement planning works because it contributes to improving peoples actual financial and activity planning for retirement through formal planning seminars, whereas informal
planning works as it sets up the psychological expectations about retirement.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Figure 1
Summary of Issues and Relationships Considered in Empirical Studies on the
Retirement Process
Individual Attributes
Demographic Characteristics
Needs and Values
Personality
Knowledge, Skills, and Abilities
Attitudes Toward Retirement
Health and Financial
Circumstances
Job and Organizational Factors
Employment History
Job Characteristics
Job Attitudes
Career Attachment
Age Stereotypes at Work
Flexible Job Options
Financial Incentives
Family Factors
Family Support
Marital and Dependent
Care Status
Marital Quality
Spouses Working Status
Socio-Economic Factors
Social Norms about Retirement
Current Economic Conditions
Future Economic Trends
Social Security System
Government Policies and
Programs
Retirement Process
Retirement Planning
Formal vs. Informal Planning
Financial Planning
Retirement Goal Setting
Voluntary Retirement
Involuntary Retirement
Retirement Intention
Bridge Employment
Career Bridge Employment
Bridge Employment in a Different Field
Volunteer Work
Full Retirement
Retirement Transition and Adjustment
Postretirement Leisure Activities
Retirement Satisfaction
Life Satisfaction
Postretirement Physical and Mental Health
Retirement Adjustment Trajectories
Spiegel and Shultz (2003) examined whether preparing for retirement and having transferable knowledge, skills, and abilities would affect retirement adjustment for retired naval
officers. Their findings indicated that retirement planning benefited those individuals with
higher retirement satisfaction and better adjustment to civilian life. It should be noted that
the vast majority of these military retirees were transitioning into another job within their life
course, which would be in line with the continuity theory and life course perspective discussed earlier and the concept of retirement as a career development stage.
Retirement planning may also facilitate goal setting on the part of older workers planning
for retirement. Hershey, Jacobs-Lawson, and Neukam (2002) noted that having clear goals for
retirement is a critical determinant of life satisfaction and adjustment during the postemployment transition period. In a recent empirical study, Stawski, Hershey, and Jacobs-Lawson
(2007) found that retirement goal clarity significantly predicted retirement planning practices,
which in turn predicted saving tendencies. In addition, Kim, Kwon, and Anderson (2005)
found that retirement preparation and planning as well as workplace financial planning, were
strong predictors of feelings of retirement confidence. Hence, setting and clarifying
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
retirement related goals are important components of the retirement planning process. This
would be in line with conceptualizing retirement as decision making, which typically takes
a more purposeful and motivated choice behavior perspective.
A recent line of research by Hershey and his colleagues (e.g., Hershey, 2007;
Hershey, Henkens, & Van Dalen, 2007; Hershey, Jacobs-Lawson, McArdle, & Hamagami,
2007; Hershey et al., 2002; Hershey, Mowen, & Jacobs-Lawson, 2003) has focused on
the psychological foundations for financial planning for retirement. This research has
explored the factors associated with designing effective financial planning seminars. For
example, Hershey et al. (2007) found that Dutch workers were less involved in retirement planning activities and had lower goal clarity scores than U.S. workers. This is not
surprising given that the vast majority of older Dutch workers are covered by guaranteed
defined benefit pension plans, whereas most U.S. organizations now offer much more
volatile and uncertain defined contributions plans. However, their research does highlight the need to look at potential cross-cultural differences with regard to the antecedents and outcomes of retirement planning.
There is another line of research that looks at the antecedents to retirement planning. For
example, Kim and Moen (2001) reported that unfavorable attitudes toward retirement were
associated with absence of retirement planning and failure to seek information about retirement, which in turn were related to unsuccessful adaptation to retirement. Ekerdt, Kosloski,
and DeViney (2000) found that the closer the perceived proximity of retirement, the more
motivated workers were to engage in both formal and informal retirement planning activities. Taylor and Geldhauser (2007) note that older workers from lower income brackets,
which also have higher proportions of women and minorities, are less likely to engage in
both informal and formal retirement planning.
Several studies have also looked at demographic differences in predicting engagement in
retirement planning related behaviors. For example, Glass and Kilpatrick (1998) note that
men are more likely to not only save more for retirement, but also invest in more aggressive
financial mechanisms. Hershey et al. (2002) found that men were more likely to have specific
concrete retirement goals (e.g., buy a motor home and travel to Alaska), whereas women
reported more general and abstract goals (e.g., be happy). A recent study by Phua and
McNally (2008) showed that younger men were much less likely to be saving for retirement
and they also made a much stronger distinction between preretirement planning and financial
planning for retirement, whereas older men saw these two forms of planning as more closely
aligned. Although much has been discussed on womens inherent disadvantages when it
comes to retirement planning (e.g., more disrupted careers, lower and secondary incomes,
more part time work), little empirical research exists that confirms this theorizing.
After reviewing the literature on retirement planning, several gaps are evident. First, the
long-term influence of retirement planning on postretirement outcomes (e.g., adjustment,
satisfaction) are relatively unknown. Taylor and Doverspike (2003), for example, suggest
that retirement planning may have its strongest impact on these outcomes early in the retirement process. However, there are few longitudinal empirical studies to support this (for an
exception, see Spiegel & Shultz, 2003). Given retirements are lasting much longer than they
did just a generation ago, more longitudinal research is needed to determine both the shortand long-term impact of retirement planning on important postretirement outcomes.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Second, research studies on how retirement planning influences outcomes other than
retirement timing, satisfaction, adjustment, and savings behaviors are badly needed (Hayslip,
Beyerlein, & Nicolas, 1997). As noted earlier, retirement no longer means just the cessation
of work (Beehr & Bennett, 2007). Instead, the concept of retirement has broadened to
include numerous options beyond leisure activities, including paid and unpaid bridge
employment, pursuit of avocations and entrepreneurial activities, second careers, and continued education. To date, however, there is little research that explicitly addresses how retirement planning may influence these other outcomes of the retirement transition process.
Therefore, future research on retirement planning that examines a wider variety of postretirement outcomes would better account for the changing nature of retirement.
A third major area in need of research is the development of more comprehensive models
of the relationship between retirement planning and various postretirement outcomes. For
example, Taylor and Doverspike (2003) suggest that goal setting and having realistic expectations for retirement may be important mediators of the relationship between retirement
planning and the postretirement outcomes of satisfaction and adjustment. However, no
empirical studies have been carried out to examine their supposition. Thus, more comprehensive models of the interplay between retirement planning, various potentially mediating
and/or moderating variables, and a variety of postretirement outcomes is needed to better
understand how best to structure retirement planning.
Finally, issues around the timing of the retirement planning process (i.e., when to begin)
need to be examined in more detail. Traditionally, most older workers do not seriously start
planning for retirement until very close to the actual retirement decision. However, Ekerdt
(2004) has noted that retirement is no longer a concern only for the second half of life, especially given the precipitous shift of the risk of funding retirement from the employer to the
individual employee. Thus, retirement planning needs to not only start sooner in ones life,
but also the focus of retirement planning may need to be substantially different during various life phases (Phua & McNally, 2008). This is becoming increasingly important as both
the nature of and financial security associated with retirement continues to change rapidly.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
classified the predictors of the decision to retire as falling into two broad categories:
(a) microlevel personal factors and (b) mesolevel work-related factors, which occur in a
larger macrolevel social, economic, and political context. However, we will also discuss a
third macrolevel category of factors that influence retirement on it own (e.g., societal norms)
which are often the focus of the sociological literature on retirement decision making.
Some of the most salient personal predictors of the decision to retire are at the individual
level. For example, the older the individual is, the more likely that the individual will retire
(e.g., Adams & Rau, 2004; Kim & Feldman, 2000; Wang et al., 2008). This supports the life
course perspectives emphasis that older workers are facing the decline of both physical
energy and cognitive abilities (Jex et al., 2007) and their engagements in further employment
become increasingly limited (Adams & Rau, 2004). Education has also been demonstrated
to be related to retirement decisions (e.g., von Bonsdorff, Shultz, Leskinen, & Tansky, 2009).
Supporting continuity theory, highly educated individuals have more capacity and options in
maintaining their life patterns because of their professional knowledge and/or skills. Thus,
they may have more opportunities to continue to work in their career field by engaging in
consulting or other entrepreneurial roles (Ekerdt, Kosloski, & DeViney, 2000).
Health is another major factor that influences retirement decisions (Jex et al., 2007;
Mutchler, Burr, Pienta, & Massagli, 1997; Shultz & Wang, 2007). Health problems might
lead to constraints on an individuals ability to perform effectively or achieve continuity of
life structure through further participation in the workforce. Consequently, employees who
are healthy are likely to continue to stay employed whereas those employees with health
problems will be more likely to retire (Barnes-Farrell, 2003). Financial status consistently
predicts retirement decisions as well (Gruber & Wise, 1999; Quinn et al., 1990). However,
the relationship is rather complicated. For example, Wang et al. (2008) found that retirees
total wealth was not able to predict the odds for retirees to engage in career bridge employment against full retirement. This suggests that financial motivation may not be a primary
driving force for people to keep working. In other words, retirees may continue working
because they are satisfied with and attached to their career jobs (e.g., Shultz, 2003), and
committed to their organizations (e.g., Adams & Beehr, 1998), but not because they are in
relatively bad financial shape.
The mesolevel work-related predictors of retirement decisions include characteristics of
the work role and a persons thoughts and attitudes about work. For example, workers in jobs
with higher substantive complexity were less likely to retire, whereas workers in jobs with
greater physical and psychological demands (Elovainio et al., 2005; Gobeski & Beehr, 2008;
Hayward, Grady, Hardy, & Sommers, 1989; Lin & Hsieh, 2001; Wang 2007) or those dissatisfied with their job (Shultz et al., 1998; Wang et al., 2008) are more likely to choose
retirement. Other studies found, however, that job satisfaction was not related to retirement
(e.g., Beehr, Glazer, Nielson, & Farmer, 2000; Gobeski & Beehr, 2008; Schmitt & McCune,
1981), but related to taking a bridge job in a noncareer field (Gobeski & Beehr, 2008).
Several studies have reported that those who report simply being tired of work are
likely to decide to retire (Beehr et al., 2000; Bidwell, Griffin, & Hesketh, 2006). In contrast,
researchers (e.g., Adams, 1999; Adams & Beehr, 1998; Adams, et al., 2002; Luchak, Pohler,
& Gellatly, 2008) have shown that attachment or commitment to ones career is negatively
related to the decision to retire. A recent study by Schmidt and Lee (2008) demonstrated that
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
work centrality and commitment to leisure activities were unique predictors of retirement
intentions, whereas affective organizational commitment only predicted turnover intentions.
These results support both role theory and continuity theory.
According to the life course perspective, family is an important life domain that may influence retirement and employment status (Szinovacz, 2003). Specifically, spouses working
status, spousal support, and marital and dependent care status have been shown to be related
to retirement decisions (Henkens, 1999; Henkens & Tazelaar, 1997; Henkens & van Solinge,
2002; Szinovacz, DeViney, & Davey, 2001). Nevertheless, there are studies (e.g., Wang et al.,
2008) that have shown other family-related variables, such as marital status and quality, were
not related to retirement decisions. Because there have been relatively fewer studies investigating this type of antecedent, it is still too early to draw conclusions.
Finally, macrolevel environmental factors have also been examined in the retirement
decision-making literature. These factors can include organizational level policies and occupational norms with regard to retirement as well as perceived biases and negative stereotypes
with regard to older workers in the workplace or society more broadly. For example,
Settersten and Hagestad (1996) reported that workplace and societal norms regarding appropriate retirement ages produce pressures on older workers with regard to their retirement
preferences and plans. Specifically, those individuals who are behind schedule with regard
to their career advancement or have plateaued in their career or job, are more likely to feel
pressure from the organization and society to retire.
After reviewing the literature on retirement decision making, several gaps in the literature
are evident. First, the decision to retire has dramatically evolved and now includes not only
when to retire, but also what to do in retirement. There has been some recent theorizing on
this issue (e.g., Wang et al., 2009); however, most empirical research is narrowly focused
on the antecedents or decision process around just a single outcome, such as full retirement
or bridge employment. As a result, future research will need to examine a wider swath of
options at the same time in order to better understand the key choice factors in deciding
among the various retirement options.
Second, the theoretical mechanisms that have been applied to studying retirement decisions all emphasized retirement decision as a rational process. In other words, they assumed
that retirees had comprehensive and accurate information about themselves and the situations
when making the decision. However, in reality, retirement decisions are often made in the
face of incomplete and imperfect information, which renders a sense of uncertainty in the
decision-making process (e.g., Ekerdt, Hackney, Kosloski, & DeViney, 2001). So far, no
theoretical framework has incorporated this notion of uncertainty, which could have important implications for understanding the psychological aspect of retirement decision.
A third gap in the literature on retirement decision making is with regard to the multilevel
nature of retirement decision making. Beehr and Bennett (2007) discuss multiple levels (e.g.,
individual, job, organization, and society) of predictors in the retirement decision. However,
most empirical research focuses on only a single level of analysis at one time (see Hardy &
Hazelrigg, 1999, as a rare exception). Therefore, more studies are needed to examine how
factors from different levels interact in influencing retirement decision making. Specifically,
in our later discussion of new theoretical directions in retirement research, we suggest
incorporating a personenvironment fit perspective may help achieving this goal.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Early Retirement
Research on early retirement has typically taken one of several avenues. The first has
focused on the predictors of who will accept Early Retirement Incentives (ERIs). The second
has been primarily group comparisons of those older workers who retire early versus those
who either stayed at work or those who ultimately retired on time, whereas the third line
of research has focused more on the outcomes associated with early retirement. As a result,
much of the ERI research has taken either a retirement as decision-making approach (i.e.,
how do older workers weigh the perceived advantages and disadvantages with regard to ERIs
and come to a decision; Feldman, 1994) or a human resource management approach (i.e.,
how can the organization control who and how many eligible older workers accept ERIs at
a reasonable cost to the organization). In contrast, the second type of early retirement
research has used a wider variety of theoretical frameworks (e.g., continuity theory and role
theory). Similarly, the third line of research that focuses on the outcomes of early retirement
has also used a continuity framework for examining the early retirement process.
According to Feldman (2003), although organizations have near complete control over
the financial incentives offered, they have much less control over the valence such incentives
will have for eligible workers. In addition, the projections made regarding the impact of ERI
specifics on acceptance rates can be wildly inaccurate if nonfinancial considerations are not
taken into account. Kim and Feldman (1998), for example, examined the acceptance rates of
early retirement incentive packages over three points in time for university faculty members.
Their results showed that poorer health, lower salary, and higher pension benefits led to early
retirement decisions. In addition, faculty members who planned on working part time after
retirement and who expected there would be no future incentive programs were more likely
to accept early retirement offers, whereas those with higher productivity and with spouses
still in the workforce were less likely to do so. Thus, it is clear that it is a combination of
organizational factors and individual characteristics that persuade early retirement decisions.
Doerpinghaus and Feldman (2001) tried to predict early retirement decisions in defined
benefit pension plans. They found that unlike defined contribution plans, workers who
earned more were more likely to retire early. In addition, gender, age, martial and family
status, and income all interact in unique ways for those covered under defined benefit pensions versus those under defined contribution plans. Also, the major leverage point for organizations with defined benefit pension plans is to lower the eligibility age for full retirement,
whereas in organizations with defined contribution plans lump sum bonuses or direct contributions to the employees retirement plans have the largest influence. Thus, whether the
employee is covered under a defined contribution or defined benefit pension plan can make
a dramatic difference on the value of any specific incentive used in an ERI and how the
various influences come together to inform the ERI decision (Feldman, 2003).
Although much of the research on early retirement has looked retrospectively at who
accepts ERIs and for what reasons, there is another line of research that looks at early retirement intentions, an important precursor to the act of retirement. For example, both Lin and
Hsieh (2001) and Elovainio et al. (2005) reported that those who perceived their jobs as
being stressful and having higher workloads intended to retire early. Although intentions do
not always lead to the desired behaviors, they do serve as one of the strongest predictors of
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
the actual behavior (Prothero & Beach, 1984). Thus, research that examines the precursors
to older workers preferences for early or on-time retirement is important to examine.
The majority of research on early retirement has been conducted with workers in whitecollar professional occupations. However, recent research from blue-collar workers in
Poland (Szubert & Sobala, 2005) indicated that work conditions (e.g., heavy lifting at work),
health conditions, alcohol abuse, and absence of leisure time were related to increased risks
for leaving the workforce before the nationally mandated age of retirement (e.g., early retirement). Thus, it appears that the mechanisms behind early retirement for different classes of
occupations may be somewhat different. That is, whereas professional workers may focus on
factors such as autonomy and psychological stress, blue-collar workers may be more likely
to focus on the physical demands of the job and the potentially collective negative physical
and psychological health effects that may have accumulated over the course of their careers.
There is also some research examining the potentially unique outcomes associated with
early retirement. Although not all on-time retirement decisions are likely to be voluntary,
it is clear that early retirement decisions are much more likely to be perceived as involuntary
and negative by older workers (Shultz et al., 1998). Gowan (1998), for example, noted that
older workers who did not desire to retire, had lower self-esteem, fewer financial resources,
and plans to continue working, were much more likely to view an offer of an early retirement
package as harmful. In another study, Mein and Ellison (2006) found that older workers
viewed early retirement incentive offers as representing the downsizing effort of their organization, which led them to have negative perceptions about their work environment.
On reviewing the literature on early retirement, several gaps become evident. First, as
early retirement may often be initiated by the organization via an early retirement incentive
package, the need for non-self-report data on the details of the package offered are needed.
Much of the past research on ERIs has focused on the self-report perceptions of the older
worker with regard to the ERI being offered, rather than an explicit comparison of the actual
offer details to the employees perceptions. Although perceptions are clearly important in
how the older employee interprets the ERI offer, the accuracy of those perceptions may well
be important in how the retiree ultimately adjusts to the decision of whether to retire early
or not.
Second, older workers perceptions of the fairness of an ERI have rarely been explored.
Barnes-Farrell (2003) suggests that using an organizational justice framework might be particularly helpful when examining early retirement decision making. That is, how do the older
workers view the informational and interpersonal justice of an ERI in terms of how fair they
perceive they were treated during and subsequent to the ERI process? The changing nature
of the psychological contract at work may also relate to retirees perception of fairness,
because the traditional paternalist stance of the organization is being replaced with one that
requires the workers themselves to bare the burden of forging their career and accepting the
risk of funding their retirement via defined contribution pension systems (Sullivan, 1999).
Finally, early retirement has broader implications for organizations and society as it represents an off time event (i.e., the timing of the event is inconsistent with the societal
norms) that are often unexpected. As a result, from the human resource management perspective, organizations and society, more broadly, must account for the influence that early
retirements have on other members of the organization or society. Although policy makers
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
have proposed initiatives to increase the average retirement age (e.g., elimination of mandatory retirement at a given age, increase in the age to receive full social security benefits), it
is unclear how much influence these factors have. Thus, more cross-national research is
needed to determine the influence such policy initiative has on the culture with regard to
perceptions of the transition to retirement as well as actual retirement patterns. Furthermore,
questions such as why these policies have been successful and what would constitute an
efficient policy design remain unanswered (Henkens & Van Dalen, 2003).
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
(e.g., Choi, 2001; Dorfman, 1992; Gall et al., 1997; Hardy & Quadagno, 1995; Pinquart &
Schindler, 2007; Quick & Moen; 1998; Reitzes & Mutran, 2004) have been repeatedly found
to be positively related to the quality of retirement transition and adjustment. Furthermore,
health declines during the retirement transition were negatively related to the quality of
retirement, supporting continuity theory (e.g., Kim & Moen, 2002; van Solinge & Henkens,
2008; Wang, 2007). The findings regarding retirees financial declines have been mixed. The
negative impact of financial decline was found for a recent Dutch sample (van Solinge &
Henkens, 2008), but not American samples who retired in the 1990s (Kim & Moen, 2002;
Wang, 2007). American retirees who entered retirement in 1990s have been recognized to
accumulate enough financial resources to support their retirement (Engen, Gale, & Uccello,
1999; Gustman & Steinmeier, 1998), which may not be the case for the Dutch.
Among preretirement job-related variables, supporting role theory, retirees work role identity (e.g., Matthews & Brown, 1987; Quick & Moen, 1998; Reitzes & Mutran, 2004) is negatively related to retirement transition and adjustment outcomes, whereas work stress (e.g., Wang,
2007), psychological and physical job demands (e.g., Quick & Moen, 1998; Wang, 2007), job
challenges (e.g., van Solinge & Henkens, 2008), job dissatisfaction (e.g., Wang, 2007), and
unemployment before retirement (e.g., Marshall, Clarke, & Ballantyne, 2001; Pinquart &
Schindler, 2007) is positively related to retirement transition and adjustment outcomes.
Among family-related variables, supporting the life course perspective, married retirees
usually adjust better than single or widowed retirees (e.g., Pinquart & Schindler, 2007), but this
beneficial effect disappears when their spouses are still working (e.g., Moen, Kim, &
Hofmeister, 2001; Wang, 2007). Retirees with happier marriages (e.g., Rosenkoeter & Garris,
1998; Szinovacz & Davey, 2004; Wang, 2007) and fewer numbers of dependent children (e.g.,
Kim & Feldman, 2000; Marshall et al., 2001) are more likely to achieve better retirement
adjustment outcomes. Finally, van Solinge and Henkens (2008) showed that losing a partner
during the retirement transition had negative impact on retirement satisfaction.
Among retirement transition-related variables, supporting continuity theory, the voluntariness of the retirement (e.g., Gall et al., 1997; Reitzes & Mutran, 2004; Shultz et al., 1998;
van Solinge & Henkens, 2007, 2008) and retirement planning (e.g., Elder & Rudolph, 1999;
Reitzes & Mutran, 2004; Rosenkoeter & Garris, 2001; Wang, 2007) are positively related to
retirees retirement satisfaction, life satisfaction, and well-being. Retiring earlier than
expected is negatively related to retirement transition and adjustment outcomes (e.g., Quick
& Moen, 1998; Wang, 2007). Quick and Moen (1998) also showed that different retirement
motivations led to different levels of retirement satisfaction. Specifically, retiring for health
care reasons negatively related to retirement satisfaction, whereas retiring to do other things
and to receive financial incentives positively related to retirement satisfaction.
Among postretirement activities, bridge employment (e.g., Kim & Feldman, 2000; Wang,
2007; Zhan et al., in press), volunteer work (e.g., Dorfman & Douglas, 2005; Kim &
Feldman, 2000; Smith & Moen, 2004), and leisure activities (e.g., Dorfman & Douglas,
2005; Dorfman & Heckert, 1988; Kim & Feldman, 2000; Reeves & Darville, 1994) are
positively related to retirement transition and adjustment outcomes. Furthermore, retirees
anxiety associated with maintaining their social status and social contacts via social activities
was negatively related to retirement satisfaction (e.g., van Solinge & Henkens, 2007, 2008).
Several gaps are readily noted in reviewing the empirical findings in the research of retirement transition and adjustment. First, despite the fact that personality variables have shown
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
good prediction on transition and adjustment quality in other types of life transition and
adjustment, such as relocation transition (e.g., Kling, Ryff, Love, & Essex, 2003), cross-
cultural adjustment (e.g., Wang & Takeuchi, 2007), and transition to parenthood (e.g.,
Levy-Shiff, 1994), very few empirical studies have linked them as individual attributes to
retirement transition and adjustment. Given that personality variables have important impact
on individuals coping style and social behaviors, more research should focus on explicating
how these variables may influence retirement transition and adjustment.
Second, although it may be fruitful to systematically examine different retirement motivations on the retirement transition and adjustment outcomes (Quick & Moen, 1998), few
studies have done so. As a result, despite that motivation research has made significant progress in both empirical and theoretical fronts in the past 20 years (for reviews, see Latham &
Pinder, 2005; Mitchell & Daniels, 2003), this progress has not been applied to advance our
understanding about retirement transition and adjustment. Motivational theories, such as
self-regulation, resource perspective, and the model of selective optimization with compensation, may be able to link the examinations of retirement motivations to postretirement
activities and inform the processes that underlie retirement adjustment.
Third, few studies have examined socioeconomic contextrelated variables on retirement
transition and adjustment outcomes. These variables may include labor market needs, macroeconomic trends, government policies related to retirement, and public opinions (see
Figure 1 for other examples). To the extent that these variables may influence retirees postretirement life style (as predicted by continuity theory) or the specific nature of the transition
(as predicted by the life course perspective), these variables could have important impact on
retirees transition and adjustment outcomes.
Finally, the empirical research on retirement transition and adjustment have mostly
focused on detecting predictive effects instead of exploring boundary conditions of these
predictive effects. In fact, few studies have examined moderators other than gender. For
example, Quick and Moen (1998) showed that preretirement work hours was positively
related to retirement satisfaction, whereas van Solinge and Henkens (2007, 2008) showed
that work hours was negatively related to retirement adjustment outcomes. Given that both
findings had solid theoretical foundations, it may be important to consider job satisfaction as
a potential moderator. Specifically, when preretirement job satisfaction was low, following
role theory, working longer would make retirement a transition that brings relief, thus leading to positive transition outcomes. However, when preretirement job satisfaction was high,
following continuity theory, retirees who had worked longer would have a harder time
adjusting to life without fewer work activities, thus leading to negative transition outcomes.
Future studies may also investigate cultural values and government practice in retirement
polices as potential moderators to understand inconsistent findings across samples from different countries (Henkens & Van Dalen, 2003).
Bridge Employment
Bridge employment is defined as the pattern of labor force participation exhibited by
older workers as they leave their career jobs and move toward complete labor force
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
ithdrawal (Shultz, 2003). It is a relatively new area of study with most of the empirical
w
studies emerging after 1990. The empirical examination started with describing the increased
heterogeneity in labor force participation among retirees and showed that retirement was not
synonymous with the absence of paid employment any more (e.g., Doeringer, 1990; Herz,
1995; Mutchler, Burr, Pienta, & Massagli, 1997; Quinn, 1997, 1999, 2000; Ruhm, 1990,
1994). These studies were often driven by the need to recognize and document the prevalence of the phenomenon (e.g., Ruhm, 1990). As such, they often conceptualize bridge
employment as a part of the retirement transition and adjustment process, which provides
functional accounts for bridge employment. Furthermore, during this research stage, the
predictors of bridge employment were often studied in an exploratory manner, mainly focusing on the demographics (e.g., age, race, gender, education, marital status; Ruhm, 1994) and
socioeconomic variables (e.g., economy trend and unemployment level; Quinn, 2000).
Starting at the end of the 1990s, empirical studies began to systematically examine the
antecedents of bridge employment decisions. The majority of them have conceptualized
bridge employment as a motivated choice behavior that may facilitate retirees transition and
adjustment to full retirement (Kim & Feldman, 2000). Accordingly, researchers have largely
relied on the informed decision-making approach to conduct their investigations. For example, Weckerle and Shultz (1999) found that older workers who wanted to continue with
bridge employment were more satisfied with their financial situations, had flexible jobs, and
felt that the decision to retire would be voluntary. Kim and Feldman (2000) found that health
status, job tenure, having a working spouse, and having dependent children were all positively related to the extent of bridge employment for retirees. They also found that age, salary at retirement, and the declination of early retirement opportunity were negatively related
to the extent of bridge employment for retirees. Kim and DeVaney (2005) found that retirees
who were self-employed and had college degrees were more likely to hold bridge jobs.
Incorporating the conceptualization that retirement is a career development stage, Davis
(2003), von Bonsdorff et al. (2009), and Wang et al. (2008) recategorized bridge employment
decisions into three types: career bridge employment (i.e., individuals accept bridge employment in the same industry/field as their career jobs), bridge employment in a different field,
and full retirement. Davis (2003) found that male retirees and retirees who had higher entrepreneurial orientations were more likely to engage in career bridge employment than full
retirement, whereas married retirees and retirees who had longer organizational tenure were
less likely to engage in career bridge employment than full retirement. Davis (2003) further
found that younger retirees and retirees who had more desire to pursue a new career were
more likely to engage in bridge employment in a different field than full retirement. von
Bonsdorff et al. (2009) further found that employees, who perceived the job market to be
good, wanted to have better use of their skills, and had less concerns about changes in their
benefits, were more likely to want to hold bridge job in a different field than retire.
Meanwhile, employees who had fewer nonwork interests, wanted to have better use of their
skills, and had more monetary desire, were more inclined to want to hold career bridge job
than retire. Finally, younger female employees who had fewer nonwork interests were more
likely to want to hold career bridge job instead of bridge job in a different field.
Building on Kim and Feldman (2000) and Davis (2003), Wang et al. (2008) found that
(a) retirees who were younger, received more years of education, had better health, experienced
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
less work stress and higher job satisfaction at preretirement jobs, and thought less about
retirement, were more likely to engage in career bridge employment than full retirement;
(b) retirees who were younger, received more years of education, had better health and financial conditions, experienced less work stress at preretirement jobs, and thought less about
retirement, were more likely to engage in bridge employment in a different field than full
retirement; and (c) retirees who had better financial conditions and experienced less work
stress and higher job satisfaction at preretirement jobs were more likely to engage in career
bridge employment than bridge employment in a different field.
Conceptualizing retirement as a career development stage, other researchers have
applied a job seeking model to understanding bridge employment. For example, Adams
and Rau (2004) found that retirees age, traditional job search constraints, and work
ethic were negatively related to job seeking, whereas their negative attitude toward
retirement and job seeking social support were positively related to job seeking as
expected. In another study, Loi and Shultz (2007) found that younger retirees were more
motivated to seek bridge employment opportunities than older retirees for financial reasons, whereas older retirees were more motivated than younger retirees to seek jobs with
more flexible schedules.
Furthermore, empirical studies on bridge employment that conceptualizes retirement/
bridge employment as a part of human resource management have begun to emerge. These
types of studies focus on addressing the research question: what are the desirable jobs or
organizational characteristics of bridge employment that attract retirees? For example, Rau and
Adams (2005) found that scheduling flexibility and a targeted equal employment opportunity
statement positively influenced older workers attraction to the organization. Research along
this line will be important for organizations to develop better recruitment strategies and
practices to attract retirees to the bridge employment jobs they offer.
Relatively fewer studies, however, have examined outcomes of bridge employment.
These studies have mainly conceptualized bridge employment as a part of the transition and
adjustment process of retirement and accordingly, role theory and continuity theory have
been most often used as the theoretical frameworks for this type of investigations. For
example, Kim and Feldman (2000) found that bridge employment was positively related to
both retirees retirement satisfaction and life satisfaction. Wang (2007) found that bridge
employment helped retirees to maintain their psychological well-being during the retirement
transition process. In a recent study, Zhan et al. (in press) applied both role theory and continuity theory and found that retirees who engaged in bridge employment had fewer major
diseases and declines of daily functions than those who engaged in full retirement. They also
found that compared with retirees who engaged in full retirement, career bridge employment
was beneficial to retirees mental health, whereas bridge employment in a different field did
not show a beneficial effect to retirees mental health. Another study by Dendinger, Adams,
and Jacobson (2005) showed that the generative reason for working (i.e., working for teaching and sharing knowledge with the younger generation) was positively related to retirees
bridge employment satisfaction and attitude toward retirement, whereas the social reason for
working was negatively related to retirees attitude toward retirement.
Reviewing the empirical findings in the research on bridge employment, several gaps
should be noted. First, there is a lack of research that examines bridge employment as a
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
longitudinal phenomenon. For example, we typically assume that during bridge employment, retirees would gradually decrease their labor force participation and finally fully
transition into retirement. This is the basis for believing that bridge employment may
facilitate retirees transition and adjustment to full retirement (Kim & Feldman, 2000; Wang
et al., 2009). However, this may not be the case in reality. It is plausible that bridge employment serves as a good adjustment mechanism and follows a smooth workforce withdrawal
trajectory for some retirees, but it could also be a bumpy process for other retirees, driven
by motivations that are not related to adjustment to workforce withdrawal. Future research
needs to explore more about the actual bridge employment process and provide a more
comprehensive evaluation on its form and functions.
Second, inconsistency exists regarding the effects of family-related variables in predicting
bridge employment decisions. For example, Kim and DeVaney (2005), Kim and Feldman
(2000), and Wang et al. (2008) all failed to find a predictive effect of marital status. This
inconsistency may be due to the fact that there are other aspects of family-related life that
may moderate the effects of these particular predictors. For example, Wang (2007) has
shown that ones spouses work status moderated the predictive effect of marital status on
retirement transition and adjustment patterns. An alternative explanation may be that the
family-related variables are overshadowed by other more proximal predictors (e.g., financial
pressure or health considerations; Barnes-Farrell, 2003). Future studies may consider testing
more complex models for effects of family-related variables.
Third, one important gap in the previous literature is that no study has systematically
examined the financial outcomes of bridge employment. Specifically, although we know that
some retirees engage in bridge employment for financial reasons (e.g., Kim & Feldman,
2000; Wang et al., 2008), we are still unclear about whether retirees with bridge employment
would necessarily do better financially. Thus, empirical studies need to explicate how bridge
employment may influence ones financial well-being in both the short and long terms.
Finally, bridge employments relationships with a lot of other variables (both as antecedents and outcomes) are still unknown. Here, we highlight several concrete theoretical
approaches that would help guide the search for potential relevant variables. First, a newcomer socialization model (Bauer, Bodner, Erdogan, Truxillo, & Tucker, 2007) could be
applied to study job outcomes (e.g., job satisfaction, organizational commitment, job performance, and turnover) of bridge employment. Following this model, these job outcomes
are proximally determined by retirees adjustment to bridge employment jobs (e.g., role clarity, social acceptance at work, self-efficacy in conducting the job) and distally determined by
both retirees information seeking and organizational socialization tactics. Second, a
resource allocation model (e.g., Edwards & Rothbard, 2000) may be applied to conceptualizing linkages between bridge employment and family-related variables. Specifically, bridge
employment jobs may change retirees resource allocation in time, attention, and energy that
are invested on both work and nonwork activities. As such, a broad range of family-related
behaviors (e.g., leisure activities, intimacy behaviors, child care, and elder care) may be
influenced by the shift in resource allocations. Third, the demand-control model (e.g.,
Karasek & Theorell, 1990) could be applied to recognizing the effects of job embedded work
stressors (e.g., quantitative and qualitative workload, organizational constraints, time pressure) on retirees safety behaviors and physical and psychological well-being. Finally,
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
c ombining the models of selective optimization with compensation (e.g., Baltes, 1997) and
job seeking (e.g., Kanfer, Wanberg, & Kantrowitz, 2001) may better explain job search and
job choice related behaviors associated with bridge employment.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
needs, people experience misfit. For example, some older workers need to fulfill family care
obligations. However, such need may not be supplied by a job that does not allow for a flexible schedule. As such, when having to decide whether to retire, older workers with family
care needs will be more likely to retire from such jobs. Second, the demandsabilities fit
mechanism suggests that when individual abilities are not sufficient to meet environmental
demands, people experience misfit. For example, for retirees who do not have enough financial resource to meet the financial demands in an inflated economy, they are more likely to
take bridge employment jobs. Finally, the supplementary fit mechanism concerns the comparison between the person and his or her environment. If the similarity (e.g., demographic
and value congruence) is low, people will perceive less of affiliation, belonging, closure, and
clarity, which leads to the experience of misfit (e.g., Kristof-Brown et al., 2005). As such,
older workers who have high generative motivation may be more likely to retire from an
organization that does not value experience and knowledge sharing.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
conditions), and individual level (e.g., health behaviors and personal development). As such,
adopting the resource perspective may lead to more comprehensive and fruitful examination
of different retirement transition and adjustment outcomes. For example, applying the
resource perspective, certain individual differences (e.g., openness to change, goal orientation in retirement, and need for structure), which may impact retirees motivational resources,
and certain environmental factors (e.g., family support, community cohesiveness, and unemployment rate in the local labor market), which may impact retirees financial and social
resources, can be hypothesized to predict how fast the turning points will be reached for
retirees who experience negative change first but positive change in their well-being later.
This is because the more resources the retirees accumulate, the more likely they will switch
from the downward trend to the upward trend.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
life course perspective is that the variability in peoples developmental history and context
will influence peoples retirement transition and bridge employment status. However, tests
of this premise often yield inconsistent findings in empirical studies. As Barnes-Farrell
(2003) pointed out, although a number of individual and contextual factors such as individual dispositions (e.g., locus of control and self-image orientation), psychosocial variables
(e.g., social support and organizational climate), and family-related variables (e.g., marital
status and marital quality) may have the potential to influence retirees preferences and
intentions for bridge employment, these variables do not always play out in terms of actual
decision making at one specific time point. Instead, their predictive effects may be overshadowed by other more proximal predictors at that time point (e.g., financial pressure or health
considerations; Barnes-Farrell, 2003; Wang et al., 2008). Therefore, although these individual
and contextual variables may be theoretically relevant to bridge employment decision at each
specific time point, their effects may be better tested in a model that predicts the overall
dynamic characteristics of bridge employment (e.g., frequency or average level of involvement in engaging in bridge employment during a long period of retirement life). In other
words, these distal variables can be viewed as background variables that shape the decisionmaking contingencies, thus predicting the longitudinal bridge employment change patterns
better than the one snap-shot measure of bridge employment status.
Similar logic applies to examining outcomes of bridge employment as well. We should
recognize that it is not the onetime bridge employment decision that shapes retirees adjustment patterns, retirement life quality, or health, but the overall length and quality of bridge
employment. Therefore, the current operationalization of bridge employment in the literature (i.e., one snap-shot status measure) tends to provide unreliable estimates for beneficial
effects of bridge employment on retirement adjustment outcomes and health outcomes (e.g.,
Wang et al., 2009). Furthermore, applying the dynamic conceptualization of bridge employment may also help us to clarify the beneficial effects of bridge employment. It is conceivable that at the beginning of the retirement transition, bridge employment is beneficial in
maintaining the continuity of life patterns and styles, thus, leading to salient positive psychological outcomes. With the retirement adjustment approaching a desirable state, the beneficial effect of bridge employment may become more physical and activity based, thus leading
to more salient physical and functional benefits on retirees health.
References
AARP. 2005. The business case of workers age 50+: Planning for tomorrows talent needs for todays competitive
environment. Washington, DC: Author.
Adams, G. A. 1999. Career-related variables and planned retirement age: An extension of Beehrs model. Journal
of Vocational Behavior, 55: 221-235.
Adams, G. A., & Beehr, T. A. 1998. Turnover and retirement: A comparison of their similarities and differences.
Personnel Psychology, 51: 643-665.
Adams, G. A., Prescher, J., Beehr, T. A., & Lepisto, L. 2002. Applying work-role attachment theory to retirement
decision-making. International Journal of Aging and Human Development, 54: 125-137.
Adams, G., & Rau, B. 2004. Job seeking among retirees seeking bridge employment. Personnel Psychology, 57:
719-744.
Ajzen, I. 1991. The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50:
179-211.
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016
Downloaded from jom.sagepub.com at Universiti Teknologi MARA (UiTM) on June 15, 2016