Minorities and the Process of Stratification

Author(s): Beverly Duncan and Otis Dudley Duncan

Source: American Sociological Review, Vol. 33, No. 3 (Jun., 1968), pp. 356-364
Published by: American Sociological Association
Stable URL: http://www.jstor.org/stable/2091911
Accessed: 11-06-2016 23:28 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/2091911?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact [email protected].

American Sociological Association, Sage Publications, Inc. are collaborating with JSTOR to digitize,
preserve and extend access to American Sociological Review

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms
 MINORITIES AND THE PROCESS OF STRATIFICATION *

BEVERLY DUNCAN AND OTIS DUDLEY DUNCAN

University of Michigan

Data concerning educational and occupational achievement, and the influence thereon of
social and national origin, are presented for a 1962 sample of native American non-Negro
males, ages 25-64, whose family heads had been pursuing a nonfarm occupation when the
respondent was 16. There are substantial differences among national-origin groups with respect
to both educational and occupational achievement. Allowance for inter-group differences in
social origin, by partial regression techniques, reduces the range of difference with respect to
educational achievement, and with respect to occupational achievement, by about one-third.
The national-origin classification is much less important as an explanation of the variance
among respondents with respect to their education and occupation than with respect to the
education and occupation of their family heads. In this sense a "melting-pot" phenomenon
obtains in America. Once equated with respect to starting point in the social structure and
educational attainment, the occupational achievement of one national-origin group differs little
from that of another.
The experience of non-Negro minorities, as revealed by these data, would argue against the
existence of pervasive discrimination on purely ethnic grounds. The notion of equal opportu-
nity irrespective of national origin is a near reality, the outstanding exceptions being the over-
achievement of Russian-Americans and the under-achievement of Latin-Americans. This find-
ing contrasts sharply with the evidence, based on the same mode of analysis, of discrimination
against the American Negro.

Tis commonly held that ethnic and racial lelism between the assumed ranking of groups
minorities differ with respect to the with respect to vertical mobility and their
rapidity of assimilation into the "ma- ranking with respect to the "Achievement
jority" socioeconomic system or in the rate Syndrome" was taken to imply that differ-
of upward social mobility between genera- ences among minorities in the achievement
tions. Indeed, the relative positions of the orientation of their members were reflected
several minorities with respect to vertical in their disparate rates of vertical mobility.
mobility often are assumed to be sufficiently At the same time that this linkage between
well documented that investigations of the achievement orientation and vertical mobil-
correlates or causes of mobility differentials ity was discussed in the literature, Nam
can proceed without re-establishing the mag- (1959) reported a set of findings about dif-
nitude of the differentials themselves. In ex- ferences in socioeconomic status between
ploring the consequences for social mobility immigrant and second-generation Americans
of inter-group differences in "psychological in each of ten national-origin groups. The
and cultural orientations toward achieve- mean of the status scores for the second-
ment," for example, Rosen (1959) took as generation exceeded the mean status score
given the low vertical mobility of Negroes, for the immigrant groups when inter-group
the greater vertical mobility of Southern differences in age and rural-urban residence
Italians and French-Canadians, and the still had been controlled statistically. The na-
greater vertical mobility of the Jewish and tional-origin groups for which the difference
Greek minorities. For him the cause of mo- in score between generations fell closest to
bility differentials rather than the existence the mean difference were the Czechs, the
of such differentials was problematic. Paral- Italians, and the Russians. If the last group
can serve as a surrogate for the Jewish
* The research reported herein was carried out minority and if the experience of Southern
under Contract No. OE-5-85-072 with the U.S. Italians is accurately represented by the ex-
Office of Education, as a part of Project No. 5-0074 perience of all Italians, Nam's findings imply
(EO-191), "Socioeconomic Background and Occu-
that both Jews and Southern Italians were
pational Achievement: Extensions of a Basic Model."
Susan Bittner and Ellen Shantz provided computa- "average" minorities with respect to rate of
tional assistance. vertical mobility. Differences between gen-
356

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms
 MINORITIES AND STRATIFICATION 357

erations for the other minority groups age 16; and the socioeconomic status score
studied by Rosen were not assessed by Nam, of the job then held by the family head. Na-
but the possibility that Rosen's assumed tional origin appears as a 13-fold classifica-
ranking of minorities with respect to vertical tion based on the country of birth of the
mobility was erroneous must be entertained. fathers of native white males, the region of
If so, the consequences of ethnic variation birth of the white males of native parentage,
in achievement orientation for social mobil- and a residual category for the native males
ity become ambiguous. of the "minor" nonwhite races. Achievement
Recently Taeuber and Taeuber (1968) is indexed by the school years completed by
have pointed out some pitfalls in inferring the respondent and by the socioeconomic
rates of vertical mobility for national-origin status score of the job he held, or had most
groups from comparisons of the socioeco- recently held, at the time of the survey.
nomic status of immigrant and second-gen- The quality of the survey data is thought
eration Americans. Specifically, they call at- to be generally good, insofar as independent
tention to the implicit and demonstrably checks can be made (Duncan, 1965; Blau
false assumption that the foreign-born and and Duncan, 1967). The numbers of sample
second-generation Americans surveyed on a cases in several national-origin strata are
given date approximate "immigrants and small, however, and even a rough check on
their children" or "second-generation Amer- the differences among national-origin groups
icans and their parents." It is, however, on can only be carried out for one social
precisely such comparisons, bolstered by an characteristic, the educational attainment of
occasional intensive study for a select sub- the respondents. A mean education score was
population, that generalizations about the calculated from published tabulations of the
social promotion of American minorities rest. 1950 Census (Vol. IV, Pt. 3, Ch. A, Tables
A body of data now exists which permits 10 and 20) for native white males aged 25
somewhat tighter measurement of the educa- to 44, of foreign or mixed parentage, and re-
tional and occupational achievements of na- siding in urban or rural-nonfarm territory.
tive Americans and the influence thereon of The coefficient of correlation between these
social origin and national origin. scores and the scores reported in Table 1
was found to be 0.98 over nine national-ori-
THE DATA ON ACHIEVEMENT AND ORIGIN gin groups (the six nations separately identi-
fied in the stub of Table 1, England-Wales,
Each civilian noninstitutional male be-
Norway, Sweden, and Austria equated with
tween the ages of 20 and 64 who was in-
Northwest Europe except Ireland and Ger-
cluded in the March, 1962, Current Popula-
many, Mexico equated with America except
tion Survey conducted by the United States
Canada, and a residual "all other"). The ed-
Bureau of the Census was asked to com-
ucation score for nonwhites other than Ne-
plete a supplementary questionnaire, "Oc-
groes is somewhat suspect, however. The
cupational Changes in a Generation," which
score reported in Table 1 resembles closely
dealt with his social background. A report on
a Census-based score for Japanese and
the survey has been published (Blau and
Chinese, who are estimated to make up
Duncan, 1967), and the details are not re-
roughly half the group. It appears unduly
capitulated here.
high when allowance is made for the fact
The subset of data presently under scru-
that the other group members are primarily
tiny is restricted to native non-Negro males
American Indians, whose educational at-
who were between the ages of 25 and 64
in 1962 and whose family heads had been tainment is substantially less. (Although we
have no specific indication that this is the
pursuing a nonfarm occupation when the
respondent was 16 years old. The seven case, it may be that the respondents to the
items of information which enter the anal- survey and supplementary questionnaire do
ysis are identified in Table 1. Most crucial not proportionately represent American In-
are the indicators of origin and achievement. dians.)
Social origin is indexed by two variables: The survey results reveal fairly substan-
school years completed by the head of the tial differences among national-origin groups
family in which the respondent was living at with respect to both educational and occupa-

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms
 358 AMERICAN SOCIOLOGICAL REVIEW

TABLE 1. MEAN SCORES AND STANDARD DEVIATIONS ON SIX SOCIAL CHARACTERISTICS FOR NATIVE
CIVILIAN NON-NEGRO MALES OF NONFARM BACKGROUND, AGED 25 TO 64 IN 1962, AND
DEVIATIONS FROM THE RESPECTIVE GRAND MEANS FOR NATIONAL-ORIGIN
SUBGROUPS: UNITED STATES, 1962

Family Head's
First Current
Edu- Occu- Siblings, Edu- Occu- Occu- Number a
National Origin cation pation number cation pation pation (000's)

All Non-Negroes .... ..... .... 22,712

Mean Score 8.63 34.06 3.85 11.69 .30.35 43.45
Standard Deviation 3.66 22.72 2.94 3.30 22.10 24.58

White, Native Father,

Respondent Born-
South -0.28 -0.37 0.43 -0.65 -2.32 -2.47 4,549
North or West 0.85 3.23 -0.42 0.35 0.92 1.31 11,349

White, Foreign Father,

Father born-
USSR -1.51 2.39 -0.46 1.43 9.55 8.77 575
NW Europe, exc.
IrelandandGermany 0.55 -0.12 -0.31 0.26 4.52 5.55 1,040
Ireland -0.48 -5.67 0.03 0.32 4.24 1.03 413
Canada 0.38 1.85 0.29 0.16 -0.53 0.89 538
Germany 0.18 -1.56 0.37 -0.55 -2.37 0.05 643
Europe, exc. NW, Italy,
Poland, and USSR -2.09 -8.09 0.62 -0.04 -1.41 -0.69 1,165
Poland -2.91 -11.77 0.87 -0.65 -4.00 -4.80 620
Italy -2.94 -10.94 0.91 -0.64 -2.28 -6.03 1,282
America, exc. Canada -3.23 -11.30 1.62 -2.77 -6.67 -14.85 244
All Otherb -1.99 -3.54 0.86 -0.32 -2.95 -5.30 137

Nonwhite, exc. Negrob 0.24 -1.20 0.60 -0.12 -4.79 3.12 157

a Includes men not reporting specific items; approximate sample frequencies may be obtained by
dividing population frequencies, in thousands, by 2.17.
bBase is fewer than 100 sample cases.
SOURCE: Unpublished tabulations from March, 1962 Current Population Survey and supplement
thereto, Occupational Changes in a Generation, conducted by the U.S. Bureau of the Census.

tional achievement. Especially distinguished would appear to be low achievers in the edu-
by high achievement are the Russian-Ameri- cational sphere.
cans, who outrank not only the other mi- Perhaps a more fruitful approach is to
norities, but also the native-of-native ma- view membership in a given national-origin
jority. The lowest achievement is recorded group as a predetermined variable in a
on the part of native Americans whose model of the process of stratification (Blau
fathers were born in Latin America, most and Duncan, 1967, Ch. 5)-one which may
often in Mexico. It is neither of these groups influence achievement both through its link-
which most closely resembles the third-gener- ages with other ascribed characteristics and
ation in achieved status, however, but rather by fostering an "achievement syndrome" or
the Irish, the Canadians, the Germans, and circumscribing opportunities for achieve-
the "other Europeans," such as Czechs. ment.
Moreover, were a measure of vertical mobil-
ity to be constructed by subtracting from the ORIGINS AND EDUCATIONAL ACHIEVEMENT
mean achievement score of respondents the
corresponding mean score for their family From a life-cycle perspective, the earliest
heads, the group of "other European" origin measure of achievement available in this
would outrank Russian-Americans with re- subset of data is the number of school years
spect to occupational mobility; and German- or grades completed by the respondent. Mea-
Americans rather than Latin-Americans sures on the family head's educational and

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms
 MINORITIES AND STRATIFICATION 359

occupational attainment are temporally prior background characteristics. Only to the ex-
to the achievement measure, and typically all tent that the achievement of a national-origin
siblings will have been born prior to the group exceeds or falls short of the achieve-
completion of the respondent's schooling. ment "expected" on the basis of the group's
These three characteristics along with mem- social origin could an effect of national ori-
bership in a national-origin group are taken gin per se on achievement be entertained as
as antecedent to the respondent's educa- a possibility.
tional achievement. Values of the partial regression coefficients
Let us suppose that the process of stratifi- summarizing the relations of the number of
cation operates in an identical fashion for school years completed by the respondent to
men in the so-called majority and in the family head's education and occupation and
several minorities, i.e., that the respective the number of siblings are reported in the
net effects of head's education and occupa- first column of Table 2. Gross or observed
tion and the number of siblings on ed- differences among national-origin groups
ucational achievement are constant over with respect to school years completed are
national-origin groups. Differences in edu- shown in the second column. Appearing in
cational achievement would nonetheless be the third column are the differences "ex-
observed, because the several national-origin pected" among these groups on the basis of
groups differ with respect to social origin or their mean scores on the three family-back-
their mean scores on the respective family- ground characteristics and the partial re-

TABLE 2. SUMMARY OF ANALYSES OF THE EFFECTS OF SOCIAL AND NATIONAL ORIGINS ON

EDUCATIONAL ATTAINMENT: AMERICAN NATIVE CIVILIAN NON-NEGRO MALES OF
NONFARM BACKGROUND, AGED 25 TO 64 IN 1962

Mean Education Score

Partial Partial
Regression Obs. Regression
Origin Characteristic Coefficients Observed Expected Less Exp. Coefficients
Social Origin

Family Head's-

Education .1680 .... .... .... .1898

Occupation .0376 .... .... .... .0374
Siblings, Number -.2802 ... .... .... -.2708

National Origin

White, Native Father,

Respondent Born-
South .... -0.66 -0.18 -0.48 -0.51
North or West .... 0.36 0.38 -0.02 -0.04

White, Foreign Father,

Father Born-
USSR .... 1.44 -0.04 1.48 1.51
NW Europe, exc. Ireland
and Germany .... 0. 26 0.17 0.09 0.08
Ireland .... 0.32 -0.30 0.62 0.63
Canada .... 0.16 0.05 0.11 0.09
Germany .... -0.54 -0.13 -0.41 -0.42
Europe, exc. NW, Italy,
Poland, and USSR .... -0.04 -0.83 0. 79 0.83
Poland .... -0.66 -1.18 0.52 0.58
Italy .... -0.64 -1.16 0.52 0.57
America, exc. Canada .... -2.78 -1.42 -1.36 -1.30
All Other a -0.32 -0. 71 0.39 0.42

Nonwhite, exc. Negro a .... -0.12 -0.16 0.04 0.04

a Base is fewer than 100 samples cases.
SOURCE: See Table 1.

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms
360 AMERICAN SOCIOLOGICAL REVIEW

gression coefficients summarizing the rela- sion of national origin as well as social ori-
tion of the education score to the respective gin as an explanatory factor results in an
background scores for this non-Negro, non- increment of only 1.8 percent to the "ex-
farm, native population as a whole. Both the plained" variance in educational attainment.
observed and expected scores for each group Of -omewhat special interest may be the
are expressed as deviations from the mean finding that membership in a non-Negro
score for all respondents. Finally, the fourth minority typically has a positive effect on
column contains a set of estimates of the educational achievement. In fact, achieve-
national-origin effects per se, the excess or ment for no minority save the Latin-Ameri-
deficit of the observed score with respect can is damped vis-a-vis the performance
to the expected score. of third-generation Southerners, when all
An alternative set of estimates of the na- groups are equalized with respect to starting
tional-origin effects, net of social origin, is point. Only one other minority, the German-
reported in the fifth column of Table 2. The American, could be identified as "under-
respondent's education score has been re- achieving" even by the more stringent stand-
gressed not only on family head's education ard of performance set by third-generation
and occupation and the number of siblings, Americans born in the North or West.
but also on the national-origin classifica- Whatever mechanism one wishes to ad-
tion. The classification enters the analysis duce as causative, there can be little doubt
as a set of dummy variables: the respondent that a "melting-pot" phenomenon obtains in
is assigned a score of unity on the variable America, in at least one sense of that term.
representing the national-origin class in The national-origin classification is found to
which he holds membership, and a score of account for only three percent of the var-
zero on the other variables representing iance among survey respondents with re-
national-origin classes. (Values of the co- spect to grades of school completed. On the
efficients measuring the net effect of nation- other hand, the national-origin classification
ality-group membership are so scaled that accounts for some 11 percent of the variance
their weighted sum equals zero.) These esti- among the respondents with respect to the
mates of the effects on educational achieve- formal schooling of their family heads. The
ment of national origin, net of social origin, rather sharp differentials in formal schooling
resemble closely in magnitude the estimates by nationality that obtained in the parental
of national-origin effects obtained in the generation did not persist among their native
analysis reported above. sons, who assimilated the American norms
Although allowance for inter-group dif- of school attendance.
ferences in social origin reduces the range
of difference with respect to educational ORIGINS AND OCCUPATIONAL ACHIEVEMENT
achievement by about a third, some 2.8
grades separate the mean scores of the high- In the basic model of the process of strati-
est-achieving and lowest-achieving national- fication suggested by Blau and Duncan
origin groups, after group starting points (1967, Chap. 5), the effects of social origin
have been equalized. At the lower limit of on occupational achievement are represented
the range, after as well as before adjustment as in large measure indirect. The effect of so-
for social origin, are the Latin Americans; cial origin is transmitted to occupational
the performance of Russian-Americans con- achievement primarily by way of the in-
sistently sets the upper limit on the range. fluence of origin on educational attainment,
Membership in a particular national-origin which, in turn, influences occupational
group can rather clearly constitute a sub- achievement. One may entertain the possibil-
stantial "handicap" or "bonus" in the strati- ity that members of some minorities may
fication process, although the relative num- not secure occupational status consonant
bers of individuals holding such membership with their educational attainment, however,
may be too small for the nationality factor whether because of high achievement moti-
to make a major contribution to variation in vation not fully realized in the educational
the total population. In the subset of data sphere or because of discrimination in the
under analysis here, for example, the inclu- job market.

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms
 MINORITIES AND STRATIFICATION 361

Let us begin by regressing the socioeco- study. Again, minority-group membership

nomic status score of the respondent's cur- is conducive to high achievement vis-a-vis
rent occupation on the three family-back- the achievement of third-generation Ameri-
ground characteristics and then regress the cans when levels of origin in the social
score on the national-origin classification structure are equalized; the noteworthy ex-
along with the social-origin indicators. The ception remains the Latin-American mi-
results, summarized in columns (1) and (2) nority. Once more distinguished by high
of Table 3, are, in broad outline, similar to achievement, given their social origin, are
the results observed when educational the Russian-Americans.
achievement was regressed on social and na- In the next set of analyses, summarized in
tional origins. Allowance for inter-group dif- columns (3) and (4) of Table 3, a measure
ferences in social origin reduces the range of of prior achieved status, the educational at-
difference with respect to occupational tainment of the respondent, is introduced as
achievement by a third, but the extremes an explanatory factor along with the ori-
remain separated by some 15 score points gins indicators. Differences in occupational
on a scale whose standard deviation is 25 achievement among the national-origin
score points in the whole population under groups are reduced substantially when allow-

TABLE 3. SUMMARY OF ANALYSES OF THE EFFECTS OF SOCIAL AND NATIONAL ORIGINS AND PRIOR
ACHIEVED STATUS ON CURRENT OCCUPATIONAL ACHIEVEMENT: AMERICAN NATIVE CIVILIAN
NON-NEGRO MALES OF NONFARM BACKGROUND, AGED 25 TO 64 mN 1962

Partial Regression Coefficients

Characteristic (1) (2) (3) (4) (5) (6)

Social Orgin

Family head's-
Education 0.81 0.92 0.15 0.17 0.05 0.03
Occupation 0.29 0.29 0.14 0.14 0.10 0.10
Siblings, Number -1.53 -1.50 -0.42 -0.43 -0.30 -0.30

National Origin

White, Native Father,

Respondent Born-
South .... -1.66 .... 0.34 .... 0.38
North or West .... -1.03 .... -0. 87 .... -0. 53

White, Foreign Father,

Father Born-
USSR .... 8.78 .... 2.83 .... 1.33
NW Europe, exc. Ireland
and Germany .... 4.62 .... 4.32 .... 3.40
Ireland .... 3.15 .... 0.66 .... -0. 54
Canada .... 0.44 .... 0.07 .... 0.43
Germany .... 0.89 .... 2.56 .... 2.68
Europe, exc. NW, Italy,
Poland,andUSSR .... 4.50 .... 1.22 .... 0.89
Poland .... 2.56 .... 0.28 .... -0.13
Italy .... . 1.18 .... .-1.07 .... .-1.92
America, exc. Canada .... -6.20 .... -1.09 .... -2.61
All Othera .... -1.18 .... -2.85 .... -2.82

Nonwhite, exc. Negro' .... 4.14 .... 3.99 .... 5.11

Prior Achieved Status

Education .... .... 3.96 3.94 3.13 3.13
FirstOccupation .... .... .... .... 0.28 0.27
Coefficient of
Determination .182 .193 .392 .395 .432 .434

a Base is fewer than 100 sample cases.

SOURCE: See Table 1.

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms
362 AMERICAN SOCIOLOGICAL REVIEW

ance has been made not only for their dif- ket disadvantage; the Russian-Americans,
ferences in social origin, but also their dif- the Irish-Americans, and the group whose
ferences with respect to the intervening origin can be traced to other nations of
factor, educational attainment. The range northwestern Europe, all of whom enter
of differences over groups falls from 15 to 7 near the top of the occupation structure, fail
score points and is bounded no longer by the to maintain fully their initial job-market
Latin-Americans and Russian-Americans. advantage. Selective promotion or employer
Such influence as nationality has on achieve- discrimination on the basis of nationality
ment appears to operate differently in the cannot be inferred with confidence, however,
job market than in the schools. for the typical pattern of job mobility may
The nationality factor per se does not ap- be distinctive for men entering the occupa-
pear, then, to loom large in the process of tion structure near its extremes irrespective
matching men and jobs. Once equated with of their nationality.
respect to starting point in the social struc- Occupations were somewhat more closely
ture and educational attainment, the occupa- related to national origin in the parental
tional achievement of one national-origin generation than among the men currently
group differs little from that of another. in the work force. The national-origin classi-
Membership in one of three minorities- fication accounts for some 4 percent of the
Italian-American, Latin-American, or the variance in the socioeconomic status scores
residual "all other"-may depress occupa- for the family heads' occupations as com-
tional success by comparison with the norm pared with I percent of the variance in the
for the non-Southern third-generation. Eval- socioeconomic status scores for the respond-
uated against the performance of the South- ents' occupations on either the first or cur-
ern third-generation, not only these groups rent job. Unlike the intergenerational com-
but also the men whose origins trace to parison of the effect of nationality and
Canada or Poland suffer a modest handicap. educational attainment, in which the school-
Achieving increasingly greater occupational ing of many members of the parental genera-
success given their social origin and formal tion occurred in foreign countries, the inter-
schooling are the Irish, the "other Euro- generational comparison with respect to oc-
peans," the Germans, the Russians, the cupational status reflects the placement of
members of the minor nonwhite races, and the respective generations in the American
the men whose origins trace to nations of occupational structure. This comparison
northwest Europe other than Ireland and takes no account of generational differences
Germany. in job qualifications for the several national-
When the national-origin groups are origin groups, but it indicates an unambigu-
equated not only with respect to starting ous narrowing of differences among group
point in the social structure and educational positions in the occupation structure between
attainment, but also with respect to starting generations.
point in the occupational structure as in-
dexed by the socioeconomic status score
EDUCATIONAL OR OCCUPATIONAL
of the first job, the description of the na-
ACHIEVEMENT
tionality factor and occupational success
remains essentially unchanged. The relevant The influence of origin, social or national,
analyses are summarized in columns (5) and on occupational achievement can operate
(6) of Table 3. Nationality does not appear either (a) by way of an effect on educational
to operate as a common basis for selective achievement which, in turn, influences oc-
promotion once entry into the job market cupational success or (b) directly, i.e., with-
has been effected, although a close reading out mediation by schooling. The impact of
of the data reveals a few instances in which social origin on occupational achievement
a minority loses ground in the competition occurs primarily through social differentials
for jobs as work experience lengthens. The in schooling and educational differentials in
Italian-Americans and Latin-Americans, who occupational achievement. No such blanket
enter near the bottom of the occupational generalization can be made, however, with
structure, appear to accumulate a job-mar- respect to the transmission of influence

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms
 MINORITIES AND STRATIFICATION 363

from the nationality factor to occupational achievement, but also encounter discrimina-
achievement. tion subsequent to school leaving.
Estimates of the importance of the indi- Irish-Americans (0.66; 2.48) and Polish-
rect and direct routes of transmission, re- Americans (0.28; 2.29) enjoy a somewhat
spectively, are derived from partial regres- preferential position with respect to the dis-
sion coefficients reported in Tables 2 and 3. tribution of schooling by comparison with
The strength of the direct route is reflected other Americans of similar social origin.
in the magnitude of the partial regression Their occupational success is, however, in
coefficent for current occupation on the line with the American "average" for men of
given origin variable when all other origin similar social origin and educational qualifi-
variables and educational attainment have cation. In contrast, the men whose origin
been held constant statistically, as in regres- can be traced to Russia (2.83; 5.94) or
sion model (4) of Table 3. The strength of "other Europe" (1.22; 3.27), Czechoslova-
the indirect route is reflected in the magni- kia, for example, occupy a distinctly advan-
tude of the product of two partial regression taged position with respect to the distribution
coefficients: current occupation on educa- of schooling and a detectably advantaged
tional attainment when all original variables position with respect to the distribution of
have been held constant statistically (3.94, jobs. Still another distributive pattern is
as given in column (4) of Table 3); and represented by the experience of Italian-
educational attainment on the given origin Americans (-1.07; 2.25). Italians, like the
variable when all other origin variables have Irish and the Polish, enjoy a somewhat pref-
been held constant statistically (coefficients erential position with respect to the distribu-
in the fifth column of Table 2). (Numeri- tion of schooling, but their occupational
cal values of the two components are given success falls short of the American "average"
in parentheses following the names of na- for men of similar social origin and educa-
tionality groups discussed in succeeding tional qualification.
paragraphs.) Among the national-origin groups sepa-
Loosely speaking, the indirect effect of the rately identified here, there occurs but one
nationality factor on occupational success re- instance of distinctly preferential position
flects inequalities in the distribution of edu- with respect to the competition for jobs. The
cation above and beyond those based on group is a rough approximation to the Anglo-
social origin. The direct effect reflects in- Saxon Protestants, stereotypically held to
equalities in the distribution of job oppor- enjoy a favored position in American society.
tunities above and beyond those based on The men whose origin can be traced to na-
formal educational qualifications and social tions of northwestern Europe other than
origin. A national-origin group can suffer Ireland and Germany (4.32; 0.32) are pre-
discrimination or enjoy preferential treat- ponderantly sons of migrants from England
ment in either distributive process. and Wales, the Scandinavian countries, and
Third-generation Southerners (direct, Austria. There is no evidence that they are
0.34; indirect, -2.01) and German-Ameri- over-achievers in the American school sys-
cans (2.56; -1.65) are detectably handi- tem, but their occupational success is
capped in the competition for jobs by virtue greater than that of other Americans with
of their educational under-achievement. similar social origin and educational quali-
Given their formal schooling as well as their fication.
starting point in the social structure, the
Southerners experience neither discrimina-
DISCRIMINATION AGAINST MINORITIES
tion nor preferential treatment in the job
market, however, and the German-Ameri- The experience of non-Negro minorities
cans encounter somewhat preferential treat- in America, as revealed by these observations
ment in the competition for jobs. Their cases on their educational and occupational
are rather different from that of the Latin- achievements would argue against the ex-
Americans (-1.09; -5.12), who not only istence of pervasive discrimination on purely
are severely handicapped in the competition ethnic grounds. The notion of equal oppor-
for jobs by virtue of educational, under- tunity irrespective of national origin is a

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms
364 AMERICAN SOCIOLOGICAL REVIEW

near reality, challenged most severely by the terns of differentials associated with mi-
cumulative over-achievement of Russian- nority-group status and disclose something
Americans and the cumulative under-achieve- of their mechanism. Whether the causes of
ment of Latin-Americans. the residual factors in differential social
The experience of the Negro minority in promotion can be taken to be cultural,
America, as revealed by comparable obser- structural, or social-psychological in nature
vations on their educational and occupa- cannot be determined with the type of data
tional achievements, makes one less sanguine available for this analysis.
about the equality with which distributive
processes operate in this country. The cur- REFERENCES
rent occupations of Latin-Americans are re- Blau, Peter M. and Otis Dudley Duncan.
flected in a mean socioeconomic status score 1967 The American Occupational Structure.
which falls short of the mean score for all New York: John Wiley and Sons.
non-Negro, nonfarm, native males by 15 Duncan, Beverly.
1965 Family Factors and School Dropout: 1920-
points. The current occupations of Negro-
1960, Final report on Cooperative Research
Americans are reflected in a mean score Project No. 2258, Office of Education. Ann
which falls short of the same (non-Negro) Arbor: The University of Michigan.
mean by 24 points. After adjustment for Duncan, Otis Dudley.

starting point in the social structure and 1968 "Inheritance of poverty or inheritance of
race?" in Daniel P. Moynihan (ed.),
formal educational qualifications, member-
forthcoming volume.
ship in the Latin-American minority results Nam, Charles B.
in a handicap of one point on the socio-eco- 1959 "Nationality groups and social stratifica-
nomic status scale; but membership in the tion in America." Social Forces 37 (May):
328-33.
Negro minority results in a handicap of
Rosen, Bernard C.
twelve points on this scale (Duncan, 1968).
1959 "Race, ethnicity, and the achievement
The evidence of discrimination against the syndrome." American Sociological Review
American Negro in the competition for jobs 24 (February):47-60.
is difficult to discount. Taeuber, Alma F. and Karl E. Taeuber.
1968 "Recent immigration and studies of eth-
Models of the type displayed in this re-
nic assimilation." Demography (forthcom-
port can bring to light the extent and pat- ing).

This content downloaded from 140.127.23.2 on Sat, 11 Jun 2016 23:28:16 UTC
All use subject to http://about.jstor.org/terms