You are on page 1of 12

American Economic Association

Understanding Unequal Economic Opportunity


Author(s): Samuel Bowles
Source: The American Economic Review, Vol. 63, No. 2, Papers and Proceedings of the
Eighty-fifth Annual Meeting of the American Economic Association (May, 1973), pp. 346-
356
Published by: American Economic Association
Stable URL: https://www.jstor.org/stable/1817095
Accessed: 25-04-2019 17:40 UTC

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/1817095?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

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 support@jstor.org.

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

American Economic Association is collaborating with JSTOR to digitize, preserve and extend
access to The American Economic Review

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
Understanding Unequal Economic
Opportunity
By SAMUEL BOWLES*

An important objective of liberal social tional opportunity.) While this statistic


reform in the United States has been to tells us little about the extent to which the
render one's chances of economic success probability of getting rich is contingent
independent of one's race, sex, or family upon family background for particular
socioeconomic background. Public concern groups of individuals, we can understand
about economic inequality has focused less the magnitude of inequality of opportunity
upon the degree of inequality itself (as asssociated with such a correlation by hy-
measured by the size distribution of in- pothesizing a simplified version of actual
come) as upon the processes by which mobility processes in which the distribu-
some individuals wind up rich and others tion of the two variables socioeconomic
poor. It is inequality of opportunity not background and income are bivariate
inequality itself that offends the modern normal. Using this assumption, and data
liberal conscience. for 35 to 44-year-olds, the relationship be-
The economic advantage accruing to tween family background and income can
whites and males is widely recognized and be illustrated as in Table 1. While it is
deplored. Yet even among white males, clear that most of the variance of income
the chances of receiving high or low in- is not explained by socioeconomic back-
comes differ widely for individuals whose ground (at least as measured by occupa-
parents had attained differing levels of
education, occupational status, and in-
TABLE 1-PROBABILITY OF BEING IN SELECTED DEC-
come. For "non-Negro" males from non- ILES OF THE INCOME DISTRIBUTION FOR INDIVIDUALS
farm backgrounds, the correlation between OF SELECTED SOCIOECONOmIC BACKGROUND DECILESB
a composite index of socioeconomic back-
ground and income ranges from .35 to .45, Socioeconomic
Background Income (Selected Deciles)
depending on the age of the worker. (I will (Selected Deciles) 1st 3rd 7th 10th
use this correlation as a simple measure of
the degree of inequality of (income- 1st 27.2 13.7 5.2 1.2
3rd 14.1 12.9 8.9 4.0
earning) opportunity. Similarly, I use the 7th 5.5 9.0 11.7 11.2
correlation between socioeconomic back- 10th 1.2 3.8 10.7 27.2
ground and years of educational attain-
a The conditional probabili
ment as a measure of inequality of educa- were not observed directly, but were calculated from a
simple correlation between income and a composite
* Associate professor of economics, Harvard Univer- measure of socioeconomic background. See text. The
sity (on leave), and Visiting Professor, University of correlation on which this table is based is .45 and refers
Massachusetts. The material presented here is the re- to "non-Negro", 35 to 44-year-old males from nonfarm
sult of a collaborative effort with Herbert Gintis. I am backgrounds. The figures in the table indicate the
indebted also to Valerie Nelson and to members of the probability of receiving income in the indicated decile
Harvard Union of Radical Economics Seminar for criti- of the distribution of income for an individual whose
cism and assistance. The research reported here was parents' occupational status, educational level, and
supported financially by a grant from the Ford Founda- income place him at the midpoint of the indicated socio-
tion. economic background decile.

346

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
VOL. 63 NO. 2 DETERMINANTS OF INDIVIDUAL INCOMES 347

tional status, educational level, and in- tunity as well as of inequality of economic
come), it is equally evident that an indi- opportunity itself. Some writers have lo-
vidual's chances of getting ahead economi- cated the mechanism for the intergenera-
cally are strongly influenced by one's tional perpetuation of inequality in the
family origins. family itself. Thus, Edward Banfield,
Evidence presented in Peter Blau and D. P. Moynihan, Oscar Lewis, Melvin
0. D. Duncan suggests that the extent of Kohn, and others have hypothesized that
intergenerational status immobility in the inequality is reproduced through the inter-
United States has been quite stable over generational replication of differing time
the past half century. Moreover, inter- horizons, attitudes, and other subcultural
national comparisons of intergenerational patterns. While the "culture of poverty"
status transmission as summarized by theorists focus upon the differing aspira-
S. M. Miller, S. M. Lipset, and Lipset and tions, personality characteristics, and other
R. Bendix indicate that among relatively noncognitive traits of individuals from
rich countries, capitalist and state socialist differing family backgrounds, the third
alike, the extent of intergenerational im- interpretation locates the process of inter-
mobility is strikingly similar. The fact that generational status transmission in the
the influence of family background upon inheritance of intellectual skills. According
economic and social advancement is both to this view, most recently put forth by
quantitatively significant and appears to Richard Herrnstein, the poor are poor be-
be relatively stable across a variety of cause they are intellectually incompetent;
historical periods and political systems their incompetence is particularly intrac-
suggests that an explanation of intergener- table because it is rooted in the genetic
ational immobility must focus on processes structure inherited from their poor and
which are common to the advanced coun- also intellectually inferior parents. An
tries in the modern era. explanation of the intergenerational repro-
A number of explanations of inter- duction of the class structure is thus found
generational economic status transmission in the heritability of IQ.
have been proposed. Abstracting from In this paper I will present the results
those which rely primarily upon the inheri- of research on the sources of intergenera-
tance of tangible wealth (which is the tional immobility and outline an interpre-
subject of John Brittain's paper in this tation of my findings. In the first section
panel), three major interpretations may I will develop and estimate a simple model
be distinguished. The first, and probably of the process whereby socioeconomic
most widespread view, is that economic background influences individuals' in-
status is passed from generation to genera- comes. The only data available for the
tion through inequality of educational estimation of the model is from a sample
opportunity. This interpretation, elabo- which excludes blacks, women, people of
rated by T. W. Schultz, Gary Becker, Blau farm backgrounds, and the very old and
and Duncan, and others, provided the very young. Thus, my empirical analysis is
theoretical underpinnings for the assault on limited to a consideration of inequality of
poverty during the decade of the 1960's. opportunity among what is a relatively
The disappointing results of the War on privileged group, most of whom are em-
Poverty have given impetus to a revival ployed in what are coming to be known as
of a second and third interpretation. Both primary labor markets. The paper by R.
the second and third views offer an expla- Edwards, D. Gordon, and M. Reich in this
nation of inequality of educational oppor- volume offers a highly complementary

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
348 AMERICAN ECONOMIC ASSOCIATION MAY 1973

analysis of the stratification of the entire Genotypic


IQ .900 Childhood
labor force through the segmentation of
/1 b I ' . 2 5 .051
labor markets and other mechanisms.
The empirical estimates of the model .219 (a .200 .562 :;Years of.272
imply that the contribution of the genetic 9 Schooling d4 Income
Socio-economic g
inheritance of IQ to the intergenerational Background
status transmission process is trivial. In-
FIGURE 1. RECURSIVE MODEL OF INTERGENERA-
equality of educational opportunity ap- TIONAL EFFECTS ON INCOME
pears to play a significant role. More than
half of the intergenerational status corre-
lation, however, is accounted for by the estimated normalized regression coeffi-
direct effects of family socioeconomic cients which I will consider shortly.) The
background, operating independentiy of double-headed arrow represents a simple
educational attainments and IQ. In the correlation with no posited direction of
second section I offer as an interpretation causation. An extensive discussion of this
of these results a theory of inequality of model, the data and methods used to esti-
opportunity in capitalist society suffi- mate it, and the caveats concerning the in-
ciently general to take account of racial, terpretation of the results is presented in
sexual, and other forms of inequality not Bowles (1972a and 1972b).
explicitly considered in the empirical The normalized regression coefficients
analysis. A conclusion of this analysis is and zero order correlations used in this
that the intergenerational reproduction of study are based on data for 35 to 44-year-
economic inequality operates in large mea- old, non-Negro, male workers from non-
sure through a correspondence between farm backgrounds. The primary source of
the social relations of production on the the data is a 1962 Current Population
one hand and the social relations of educa- Survey, augmented by data from the Na-
tion and of family life on the other. In the tional Opinion Research Center and other
concluding section, I argue that significant sources. The observed correlations have
progress towards equality of economic op- been corrected for errors in measurement,
portunity will depend upon a radical re- using independent information on the reli-
structuring of the social organization of ability of each measure, and for restricted
production. variance in one of the samples. The corre-
lations involving childhood IQ are based
I. A Statistical Model of Intergenerational on a sample collected by the California
Immobility Guidance Study. The correlation of child-
The relative importance of the genetic hood IQ (measured at age 6-8) and adult
inheritance of IQ, inequality of educa- income is estimated using the assumption
tional opportunity, and direct income ef- that the direct effect of childhood IQ upon
fects of family status in determining the income operates entirely through adult
correlation between socioeconomic back- levels of cognitive proficiency, as measured
ground and income may be estimated us- by the Armed Forces Qualification Test.
ing the simple recursive model of income (Childhood IQ influences income indirectly
determination presented in Figure 1. In through its effect on years of educational
that figure, variables on the left are posited attainment.)
as causally prior to those on the right. The The relationship between socioeconomic
arrows indicate directions of causation. background and genotypic IQ on the one
(The numbers next to the arrows are the hand, and childhood IQ on the other are

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
VOL. 63 NO. 2 DETERMINANTS OF INDIVIDUAL INCOMES 349

based on the following assumptions, se- (d) an effect operating via the correla-
lected deliberately to yield an upper bound tion of socioeconomic background
estimate of the role of genetic inheritance with genotypic IQ and, thence,
of IQ in the intergenerational transmission through the effect of genotypic IQ
of economic status. First, I have assumed, upon childhood (phenotypic, i.e.,
following Arthur Jensen, that the herita- measured) IQ, and, thence, both
bility of IQ is .81, implying a normalized directly upon income and indirectly
regression coefficient of .9 for a genotypic via the years of educational attain-
IQ. Second, I have assumed that the im- ment (fba plus heba).
pact of socioeconomic background consti-
T he empirical decomposition of the
tutes one-half of a minimal estimate of the
correlation between socioeconomic back-
environmental effects on IQ. Thus the
ground and income is presented in Table
normalized equation for childhood IQ
2. The genetic inheritance of IQ (even as-
reads:
suming a relatively high degree of heri-
Childhood IQ = .9 Genotypic IQ tability) accounts for only a miniscule
portion of intergenerational immobility.
+ .2 Socioeconomic Background
This may seem particularly surprising in
The remaining estimated normalized re- view of the substantial impact of childhood
gression coefficients presented in Figure 1 IQ upon years of schooling.
are all significant at the .95 level. The contribution of inequality of edu-
The correlation between socioeconomic cational opportunity to the correlation
background and income may be decom- between socioeconomic background and
posed into the following additive compo- income is significant. (An analogous calcu-
nents :' lation using occupational status as the
(a) a direct effect of family socioeco- dependent variable shows inequality of
nomic background (indicated in educational opportunity to be of consider-
Figure 1 by the letter g); ably greater importance (Bowles, 1972b).)
(b) an indirect effect of family back- By far the largest contribution, however,
ground operating through the influ- is the direct effect of socioeconomic back-
ence of socioeconomic background
upon years of educational attain- TABLE 2-DECOMPOSITION OF THE SIMPLE CORRELA-
TION (r) BETWEEN SOCIOECONOmIC BACK-
ment (hd plus hec);
GROUND (SEB) AND INcOMEa
(c) an indirect effect of socioeconomic
background operating through en- Intergenerational Contribution to r
vironmental effects on childhood Income Effectsb Absolute Percentage

IQ, and, thence, directly upon in-


Direct Family SEB effect (g) .25 55
come, (c); and Indirect SEB effect via School-
ing (hd+hec) .17 39
1 The derivation of this decomposition is based on the Indirect SEB effect via Child-
normal equations of the linear regression model. Thus, if hood IQ(fc) .01 2
Genotypic IQ effect (fba+heba) .02 4
xo = Pioxi + * ' * + pnOxn + it

the correlation ro,i may be expressed Total (r) .45 100

toi pjorji forj= 1, ,n; a Calculation based


b Letters in parentheses refer to the normalized re-
gression coefficients
where rij is the simple correlation between variables andxi
simple correlation coefficient in
Figure 1 and
and xj and pjo is the normalized regression indicate the method
coefficient of of calculation of each
variable j in an equation predicting variable
effect. xO.

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
350 AMERICAN ECONOMIC ASSOCIATION MAY 1973

ground. Estimates for three additional age authority structure and property relations
groups-both younger and older are con- of the enterprise. Pursuit of these objec-
sistent with the results presented here. The tives serves the employers' dual interest in
evidence suggests that an understanding making profits and perpetuating or en-
of the intergenerational effects upon in- hancing their social class position.
come requires a more detailed description From these employer objectives, we
of the way in which family background is may derive some insight into the demand
translated into higher income opportuni- for particular worker attributes. Five im-
ties, both directly and through the medium portant sets of worker characteristics will
of schooling. be considered, the first being cognitive
attributes, such as scholastic achievement,
II. Towards A Theory of Intergenerational
as well as concrete technical and opera-
Immobility
tional skills (e.g., knowing how to do typ-
In this section I seek to explain how ing, accounting, chemical engineering, or
socioeconomic background interacts with carpentry). Second, there are personality
schooling to influence an individual's traits (such as motivation, perseverance,
chances of economic success and in so do- docility, dominance, flexibility, or tact)
ing to reproduce a family's level of income which enable the individual to operate
and economic status.2 I begin with an ob- effectively in a work role. Third, there are
vious point: in capitalist society the in- traits which I will call modes of self-
come of the vast majority of individuals presentation, such as manner of speech
derives predominately from the sale of and dress, patterns of peer identification
their labor services to employers. An ade- and perceived "social distance" from indi-
quate explanation of the intergenerational viduals and groups of different social posi-
transfer of economic status thus requires tion. These traits do not necessarily con-
understanding (a) the criteria used by tribute to the worker's execution of tasks,
employers in hiring, promotion, and pay; but may be valuable to employers in their
and (b) the way in which family structure effort to stabilize, validate, and legitimize
and schooling function to allocate those the particular structure of work roles in
attributes relevant to employers' criteria the organization as a whole. Similar in
among individuals of differing levels of function is a fourth set of traits: ascrip-
socioeconomic background. Thus, an in- tive characteristics such as race, sex, and
quiry into the intergenerational immobil- age. Finally, I will consider the role of
ity process must focus on the supply, de- credentials, such as level and prestige of
mand, and production of those personal education, which, like modes of self-
attributes and ascriptive traits which are presentation and the ascriptive traits, are
relevant to getting ahead in the world of a resource used by employers to add to the
work. I will begin with the demand for overall legitimacy of the social organiza-
personal attributes by employers. tion of production.
Three immediate objectives of em- The analytical problem, of course, is to
ployers seem relevant to the formulation of determine the precise content of these five
job-adequacy criteria: the technical effi- factors, and how each affects the income
ciency of the productive process; the main- determination process.
tenance of secure top-down control over The first type of worker attributes
production; and the legitimation of the cognitive abilities appears to be of lim-
ited
2 This section represents a brief summary of material
importance in access to prestigious
described in greater detail in Bowles and Gintis. jobs and high income. Gintis and I have

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
VOL. 63 NO. 2 DETERMINANTS OF INDIVIDUAL INCOMES 351

shown that adult cognitive abilities exert follow directly from differences in the
a relatively minor influence on success scope of independent decision making,
(measured by a composite index of income which increases with hierarchical status.
and occupational status) independent of Thus the lowest level of worker must sim-
level of schooling and socioeconomic back- ply refrain from breaking rules. On the
ground. These results are consistent with highest level, it becomes crucial that the
the findings of a Norwegian study de- worker internalize the values of the
scribed in P. deWolff and A. van Slijpe as organization, act out of personal initiative,
well as work by Zvi Griliches and William and know when not to go by the book. In
Mason, and others. Moreover, Gintis between, workers must be methodical,
(1971), drawing on a number of different predictable, and persevering and, at a
data sets, showed that the quite sub- somewhat higher level, must respond
stantial simple correlation between years flexibly to their superiors.
of schooling and income cannot be ex- Much of this description of functional
plained by the role of schools in develop- personal attributes of job performance is
ing or identifying cognitive abilities. based on the work of Richard Edwards and
All four of the remaining types of per- has been supported by his empirical re-
sonal attributes personality traits rele- search. Edwards compared supervisor rat-
vant to the work task, modes of self- ings of workers with a set of thirty-two
presentation, ascriptive traits, and cre- personality ratings by the workers' peers.
dentials appear to be integral to the In a large sample of Boston workers, he
process of intergenerational immobility. finds that a cluster of three personality
The personality traits required of "effi- traits- which he summarizes as respect for
cient" workers must correspond by and rules, dependability, and internalization of
large to the requirements of harmonious the norms of the firm predict strongly
integration into the bureaucratic order of supervisor ratings of workers in the same
the enterprise. The social relations of pro- office or work group. In addition, Ed-
duction under corporate capitalism char- wards noted that respect for rules was
acterized by a fragmentation of tasks, most important at the lower occupational
absence of worker control or ownership, levels, dependability appearing strongly
and a bureaucratically structured hier- for middle levels, and internalization of
archical division of labor determine the the norms of the firm predicting super-
personality traits required of workers. visor ratings best at the higher levels.
Some of these are general traits valuable When we pass to the literature docu-
to the employer at all levels of hierarchy menting the importance of self-presenta-
and status. All workers must be depend- tion as attributes relevant to the alloca-
able (i.e., follow rules) and be properly tion of individuals to status positions, we
subordinate to authority. Further, all are faced with a difficult problem of
workers must respond adequately to the assessment. Numerous studies, summa-
external incentives of the organization- rized by Erving Goff man, have shown
the crudest being threat of dismissal, and these personal attributes to be important
the more subtle including the possibility (albeit often covert) criteria for hiring and
of promotion to higher status, authority, promotion. Gloria Shaw Hamilton and
or pay. J. David Roesner, for example, found that
While these requirements hold for all among employers of "disadvantaged"
workers, there are important qualitative workers, "personal appearance" was be-
differences among levels. These tend to tween three and two times as important a

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
352 AMERICAN ECONOMIC ASSOCIATION MAY 1973

selection criterion as any of the following: traits work-related personality character-


work experience, specific job training, high istics, modes of self-presentation, ascrip-
school diploma, or test scores. tive characteristics, and credentials are
We may now consider the importance of central to the process of income determi-
our last two sets of employability traits: nation, an understanding of the transmis-
ascriptive characteristics (race, age, sex) sion of economic status from generation to
and acquired credentials (e.g., educa- generation requires that we discover how
tional degrees). T he legitimation of the these worker characteristics are acquired.
hierarchical division of labor, as well as T he ascriptive traits are, of course, ac-
the smooth day-to-day control over the quired at birth, or, in the case of age, in-
work process, requires that the authority escapably as life progresses, so little need
structure of the enterprise with its cor- be said of them. The acquisition of
responding structure of pay and privilege educational credentials requires survival
-respect the wider society's ascriptive in the school system and is an arduous but
and symbolic distinctions. In particular, not particularly complex process. The way
socially acceptable relations of domination in which workers come to have a particular
and subordination must be respected: set of work-relevant personality character-
white over black; male over female; old istics or modes of self-presentation requires
(but not aged) over young; and schooled a more searching analysis.
over unschooled. Gintis (1970) and I (1971) have argued
I make no claim that these social prej- at some length that the social relations of
udices originated as a capitalist contriv- schooling are a reflection of the social rela-
ance, though a strong case could probably tions of production. The school is a
be made that the form and strength of bureaucratic order with hierarchical au-
both sexism and racism in the United thority, rule-orientation, stratification by
States derive in large measure from the "ability" (tracking) as well as by age,
particular historical development of cap- role differentiation by sex (physical educa-
italist institutions in the United States and tion, home economics, shop), and a system
Europe. Save credentialist distinctions, of external incentives (marks, promise of
all predate the modern capitalist era. promotion, and threat of failure) much
"Rational business practice" has simply like pay and status in the sphere of work.
reinforced and extended them. The cre- Thus schools are likely to develop in
dentialist mentality, as Gintis and I have students traits corresponding to those re-
argued, was indeed contrived in the early quired on the job. Gintis (1971) has shown
decades of the century to perpetuate the that higher grades in high school courses
concept of social rank in a society increas- are awarded to students exhibiting per-
ingly eschewing distinctions of birth. sonality traits associated with subor-
The individual employer, acting singly, dinacy, discipline, and rule-following even
normally takes societal values and beliefs when the level of scholastic achievement is
as data, violating as few as possible, and controlled for. The economic importance of
using those which will add to the technical these aspects of school structure is strongly
efficiency, top-down control, or legitimacy suggested by Edwards' analysis of data on
of the enterprise. In this way the pursuit high school records and work super-
of profits and security of class position vision ratings collected by Brenner. The
reinforces the racist, sexist, and credential- Brenner data indicate that variables
ist mentality. measuring teacher's evaluation of student
If these four sets of noncognitive worker conduct are far more important than the

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
VOL. 63 NO. 2 DETERMINANTS OF INDIVIDUAL INCOMES 353

student's scholastic record in predicting standards of direction for behavior; con-


the individual's work adequacy as per- formity focuses on externally imposed
ceived by the supervisor. rules." Thus working-class parents value
But recall that the work-related per- obedience, neatness, and honesty in their
sonality traits required of employees differ children, while higher status parents em-
according to the work role in question, phasize curiosity, self-control, and happi-
those at the base of the hierarchy requir- ness. Kohn concludes his statistical analy-
ing a heavy emphasis on obedience and sis: "In this exceptionally diverse society-
rules and those at the top, where the dis- deeply marked by racial and religious divi-
cretionary scope is considerable, requiring sion, highly varied in economy, geography,
a greater ability to make decisions on the and even degree of urbanization- social
basis of well-internalized norms. This pat- class stands out as more important for
tern is closely replicated in the social rela- men's values than does any other line of
tions of schooling. Note the wide range of demarcation, unaffected by all the rest of
choice over curriculum, life style, and them, and apparently more important
allocation of time afforded to college than all of them together."
students, compared with the obedience To refine the relation between social
and respect for authority expected in high class, values, and child-rearing, Kohn
school. Differentiation occurs also within classifies his respondents according to the
each level of schooling. One need only to amount of "occupational self-direction"
compare the social relations of a com- inherent in their jobs. His analysis indi-
munity college with those of an elite four- cates that the "relationship of social class
year college, or those of a working class to parents' valuation of self-direction or
high school with those of a wealthy sub- conformity for children is largely attrib-
urban high school, for illustration of this utable to class-correlated variation in
point. men's exercise of self-direction in work."
The correspondence between the social The relationship of these findings to
relations of production and the social rela- the study of intergenerational immobility
tions of childhood socialization itself is not, is clear enough. Again, quoting Kohn:
however, confined to schooling. There is Whether consciously or not, parents
strong evidence for a similar correspon- tend to impart to their children lessons
dence in the structure of family life. The derived from the conditions of life of
their own social class-and thus help
male dominated family, with its structure
prepare their children for a similar class
of power and privilege further articulated
position . .. Class differences in paren-
according to age, replicates many aspects tal values and child-rearing practices in-
of the hierarchy of production in the firm. fluence the development of the capaci-
Yet more relevant for our immediate con- ties that children will someday need ...
The family, then, functions as a mecha-
cerns here is the evidence, summarized by
nism for perpetuating inequality.
Melvin Kohn, on social class, parental
values, and child-rearing practices. Kohn's Such differential patterns of child-rear-
major findings are that "middle class ing affect more than the worker's per-
parents . . are more likely to emphasize sonality and aspiration level. They also
children's self-direction, and working class influence his or her style of self-presenta-
parents to emphasize their conformity to tion: patterns of class loyalties and modes
external authority.... The essential dif- of speech, dress and interpersonal be-
ference between the terms, as we use them, havior.
is that self-direction focuses on internal The argument may now be summa-

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
354 AMERICAN ECONOMIC ASSOCIATION MAY 1973

rized. To get a job at any particular level social relations of schooling and of family
in the hierarchy of production one has to life on the other. At the same time, the
meet two tests: first, one must be able and corresponding relations of family life,
willing to do the work; and second, one schooling, and work tend to reproduce
must be of appropriate age, race, sex, edu- economic status differences among families
cation, and demeanor so that one's assign- from generation to generation.
ment to the job will contribute to the
III. Conclusion
sense that the social order of the firm is
just. Thus criteria of worker adequacy What may we conclude from this analy-
reflect more than the employer's desire sis about the power of social reform to
that workers be hardworking and capable. reduce significantly the degree of inequal-
They reflect as well the need for acquies- ity of opportunity in capitalist society?
cence to the employer's monopolization of Three main conclusions seem warranted.
power. The smooth exercise of control First, the trivial role of the genetic in-
from the top of the enterprise rests on the heritance of IQ in the intergenerational
daily reconfirmation of the employee's transmission of economic status bids us
sense of the just claim of his or her su- reject the pessimism of those who, like
periors, co-workers, and subordinates to Herrnstein, see a "virtually hereditary
their particular jobs. meritocracy" as the fruit of successful
The ability to operate well at a par- liberal reform in an advanced industrial
ticular level in the hierarchy and the legiti- society.
mate claim to one's place in the authority Second, the fact that inequality of edu-
structure and to the rewards associated cational opportunity accounts for con-
with it depend to a large extent on experi- siderably less than half of the correlation
ences in the home and at school. The pro- between socioeconomic background and
ductive enterprise is by no means a fully income cautions us to question the op-
articulated socialization agency capable of timism of those who assert that significant
shaping worker consciousness and be- progress towards equality of opportunity
havior to its needs: its control over re- can be achieved through educational re-
cruitment and internal organization can forms alone. Of course, progress towards
but reinforce patterns of consciousness de- equalization of education may help to
veloped in the larger society. That is, the undermine the legitimacy of income in-
particular structure of authority within equality and thus hasten the development
the firm which will be seen as legitimate- of stronger egalitarian movements which
whether based on distinctions of race, sex, in turn may have some impact on the inter-
educational credentials, age, manners of generational transmission process (Bowles
speech, or whatever-is an expression of 1973). But this is hardly the argument put
broader social values and prejudices. And forth by most proponents of egalitarian
these, too, are both reflected in and de- school reforms.
pendent upon the structure of family life Third, inequality of economic opportu-
and schooling. Specifically, I argue that a nity is deeply rooted in the social relations
work force which is both competent to do of production and their reflection in the
the job and consistent with the perpetua- structure of family life and schooling. The
tion of the hierarchical division of labor is differing patterns of child-rearing and
generated in large measure through a cor- other aspects of socialization faced by
respondence between the social relations children of differing social backgrounds
of production on the one hand and the represent an adaptation- sometimes con-

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
VOL. 63 NO. 2 DETERMINANTS OF INDIVIDUAL INCOMES 355

scious, sometimes not- by parents, school from the goal of equality is illegitimate.
personnel, and others to the actual condi- You can't have one without the other. As
tions of work at the various levels of the long as the hierarchical division of labor
hierarchical division of labor. The role of exists, you are not likely to have either.
families and schools in the production of
the labor force is inescapably bound up REFERENCES
with a second role of these institutions:
Edward Banfield, The Unheavenly City, Ne
the reproduction of economic inequality
York 1968.
from generation to generation. Gary Becker, Humtan Capital and the Personal
Yet the perpetuation of economic in- Distribution of Income, Ann Arbor 1967.
equality does not originate with the family Peter Blau and Otis Dudley Duncan, The
or the school, for the structure of home life American Occupational Structure, New York
and of education is but a reflection of the 1967.
broader social structure. Racist, sexist, Samuel Bowles, "Unequal Education and the
and credentialist distinctions, and class Reproduction of the Social Division of
differences in values and modes of self- Labor," Rev. Rad. Polit. Econ., Fall-Winter
1971, 3.
presentation, are systematically used and
, "Schooling and Inequality from Gen-
thus reinforced by employers in their ef-
eration to Generation," J. Polit. Econ., May-
forts to reproduce the hierarchical division
June 1972a.
of labor within the enterprise. The im-
, "The Genetic Inheritance of IQ and
portance of these traits in the determina- the Intergenerational Reproduction of Eco-
tion of individual incomes is thus not an nomic Inequality," Harvard Instit. for Econ.
expression of irrational or uninformed Res., Sept. 1972b.
employment practices subject to correc- , "The Integration of Higher Education
tion by "enlightened" employment policies into the Wage Labor System," in Michael B.
and social legislation. The link between the Katz, ed., Education in American Social His-
social relations of production and the in- tory, New York 1973.
and Herbert Gintis, "IQ in the U.S.
come determination process is so intimate
Class Structure," Social Policy, Jan. 1973.
that any substantial change in the latter is
Marshall H. Brenner, "Use of High School
contingent upon the transformation of
Data to Predict Work Performance," J. of
the hierarchical division of labor as the
Applied Psych., Jan. 1968, 52.
archetype of productive activity. Thus, P. deWolff and A.R.D. van Slijpe, "The Re-
while the inheritance of tangible wealth lation Between Income, Intelligence, Educa-
unquestionably contributes to the inter- tion, and Social Background," unpublished
generational transmission of economic paper presented at the European meetings
status, the above analysis, supplemented of the Econometric Society, Budapest Sept.
by evidence of significant immobility in 1972.
those socialist countries which have not Richard C. Edwards, "Alienation and Inequal-
ity: Capitalist Relations of Production in a
overturned the hierarchical division of
Bureaucratic Enterprise," unpublished
labor, suggests that the elimination of
Ph.D. thesis, Harvard University, July 1972.
wealth inheritance is insufficient to achieve
, D. Gordon, and M. Reich, "A Theory
equality of economic opportunity. of Labor Market Stratification," Amer.
If the above analysis is correct we are Econ. Rev., Proc., May 1973.
led to one further conclusion: inequality of Herbert Gintis, "New Working Class and
opportunity and inequality itself are part Revolutionary Youth," Socialist Revolution,
of the same process. To consider the goal May 1970.
of equality of opportunity as separable , "Education, Technology, and the

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms
356 AMERICAN ECONOMIC ASSOCIATION MAY 1973

Characteristics of Worker Productivity," Study in Values, Homewood, Ill. 1969.


Amer. Econ. Rev. May 1971, 61, 266-279. Oscar Lewis, "The Culture of Poverty," Sci.
Erving Goffman, The Presentation of Self in Amer., Oct. 1966, 215, 16-25.
Everyday Life, New York 1959. S. M. Lipset, "Social Mobility and Equal Op-
Zvi Griliches and William M. Mason, "Edu- portunity," Public Interest, Fall 1972, 29,
cation, Income and Ability," J. Polit. Econ., 90-108.
May-June 1972, 80.
and R. Bendix, Social Mobility in In-
Gloria Shaw Hamilton and J. David Roesner,
dustrial Societies, Berkeley 1959.
"How Employers Screen Disadvantaged
S. M. Miller, "Comparative Social Mobility,"
Workers," Monthly Labor Rev., Sept. 1972,
14-21. Current Soc., 1960, 9, 1-81.
Daniel P. Moynihan, The Negro Family: The
Richard Herrnstein, "IQ," Atlantic Monthly,
Sept. 1971, 43-64. Case for National Action, Cambridge 1967.
Arthur R. Jensen, "How Much Can We Boost Theodore W. Schultz, "Investment in Poor
IQ and Scholastic Achievement?" Harvard People," seminar on Manpower Policy and
Ed. Rev., Winter 1969, 39, 1-123. Programs, Office of Manpower Policy Evalu-
Melvin L. Kohn, Class and Conformity: A ation Research, DOL, Washington 1966.

This content downloaded from 177.220.86.171 on Thu, 25 Apr 2019 17:40:03 UTC
All use subject to https://about.jstor.org/terms

You might also like