Handbook of the Sociology of Education in the 21st Century

68 D. B. Downey et al.

number of days children went to school, we regation, the strength of organized labor, tax pol-
would expect that change to benefit low-SES stu- icy, immigrant status, mass incarceration, the real
dents the most and reduce overall inequality in value of the minimum wage, unemployment ben-
math and reading skills.13 efits, and family leave options, all have implica-
tions for the kind of inequality we have outside of
Relatedly, seasonal results have implications schools (Fischer et al. 1996) and they likely shape
for school accountability systems. It is likely that the size and malleability of the achievement gaps
current attempts to isolate teacher or school observed in them (Morsy and Rothstein 2016;
effects based on student growth from one year to Reardon 2011b). It may turn out that broader
the next are producing biased information. If reform is also a less expensive way to reduce
summer learning loss is variable, then differences inequality than is school reform. For example,
in teachers’ “effectiveness” gleaned from value-­ Whitehurst (2016) reports that, for every $1000 in
added models partially reflect the families and public expenditures, programs aimed at provid-
neighborhoods in which their students live, which ing poor families with more money (e.g., Earned
are unlikely to be fully accounted for by statisti- Income Tax Credit) were six to eight times more
cal controls in value-added models.14 We are effective in promoting disadvantaged children’s
unaware of any state that currently employs a cognitive skills than were preschool or Head
value-added accountability method that suffi- Start programs.
ciently accounts for children’s non-school envi-
ronments. The result is predictable—the real To be clear, we do not suggest that scholars
performance of teachers and schools serving dis- discontinue studying school mechanisms that
advantaged children is underestimated. Rather harm the disadvantaged. This research continues
than pressuring the schools that are actually per- to have value and there are indications that some
forming poorly to improve, therefore, the infor- school reforms would reduce achievement gaps.
mation produced by these accountability schemes But when the focus on inequality is overly
is as likely to mislead parents as it is to properly school-centric, which we believe it currently is,
inform them about the best-performing schools we run the risk of misallocating resources toward
(Downey et al. 2008). school reform while the fundamental source of
the problem continues unaddressed. The problem
But our primary message with respect to pol- is that school-based solutions to achievement
icy is this: There is only so much schools can do. gaps run the risk of distracting us from the kind
To make substantial changes to societal-level of broader social reform really needed to reduce
achievement gaps will require reducing the level inequality.
of inequality that exists in the non-school envi-
ronment. We do not share the view of some edu- References
cation scholars that reforms aimed at ameliorating
inequality outside of schools are too politically Alexander, K. L. (1997). Public schools and the public
difficult to confront. Instead, we view these good. Social Forces, 76(1), 1–30.
broader non-school issues as education policies
and we encourage education scholars to start Alexander, K. L., Entwisle, D. R., & Olson, L. S. (2007).
talking about them in this way. For example, Lasting consequences of the summer learning gap.
decisions regarding access to health care, income American Sociological Review, 72(2), 167–180.
inequality, racial and income-based housing seg-
Alexander, K. L., Entwisle, D. R., & Olson, L. S. (2014).
13 We would worry, of course, about whether racial gaps The long shadow: Family background, disadvan-
would increase. taged urban youth, and the transition to adulthood.
New York: Russell Sage Foundation.
14 Growth models constructed with 9-month data remove
summer noise and correlate only around 0.50 with tradi- Atteberry, A. (2011). Defining school value-added: Do
tional growth models using 12-month data, demonstrating schools that appear strong on one measure appear
that summer noise is a nontrivial problem (Atteberry strong on another? Evanston: Society for Research on
2011). Educational Effectiveness.

Borman, G., & Dowling, M. (2010). Schools and inequal-
ity: A multilevel analysis of Coleman’s equality

3  Schools and Inequality: Implications from Seasonal Comparison Research 69

of educational opportunity data. Teachers College Downey, D. B., Workman, J., & von Hippel, P. (2017,

Record, 112(5), 1201–1246. August 15). Socioeconomic, racial, and gender

Bourdieu, P. (1977). Cultural reproduction and social gaps in ­children’s social/behavioral skills: Do they

reproduction. In J.  Karabel & A.  H. Halsey (Eds.), grow faster in school or out? Available at SSRN:

Power and ideology in education (pp.  487–511). https://ssrn.com/abstract=3044923 or https://doi.

New York: Oxford University Press. org/10.2139/ssrn.3044923

Bowles, S., & Gintis, H. (1976). Schooling in capitalist Downey, D. B., Quinn, D., & Alcaraz, M. (2017). The dis-

America: Educational reform and the contradictions tribution of school quality (Working Paper).

of economic life. New York: Basic Books. Duffett, A., Farkas, S., & Loveless, T. (2008). High-­

Burkam, D. T., Ready, D. D., Lee, V. E., & LoGerfo, L. F. achieving students in the era of No Child Left Behind.

(2004). Social-class differences in summer learning Washington, DC. Retrieved http://www.edexcellence.

between kindergarten and first grade: Model specifi- net/detail/news.cfm?news_id=732&id=92

cation and estimation. Sociology of Education, 77(1), Duncan, G.  J., & Magnuson, K. (2011). The nature and

1–31. Retrieved http://www.jstor.org.proxy.lib.ohio- impact of early achievement skills, attention skills,

state.edu/stable/3649401 and behavior problems. In G.  J. Duncan & R.  J.

Chatterji, M. (2006). Reading achievement gaps, corre- Murnane (Eds.), Whither opportunity: Rising inequal-

lates and moderators of early reading achievement: ity, schools, and children’s life chances (pp.  47–69).

Evidence from the Early Childhood Longitudinal New York: The Russell Sage Foundation.

Study (ECLS) kindergarten to first grade sample. Entwisle, D. R., & Alexander, K. L. (1992). Summer set-

Journal of Educational Psychology, 98(3), 489–507. back: Race, poverty, school composition, and math-

Coleman, J.  S., et  al. (1966). Equality of educational ematics achievement in the first two years of school.

opportunity. Washington, DC: Department of Health, American Sociological Review, 57(1), 72–84.

Education and Welfare. Entwisle, D. R., Alexander, K. L., & Olson, L. S. (1994).

Condron, D. J. (2008). An early start: Skill grouping and The gender gap in math: Its possible origins in neigh-

unequal reading gains in the elementary years. The borhood effects. American Sociological Review, 59(6),

Sociological Quarterly, 49, 363–394. 822–838.

Condron, D.  J. (2009). Social class, school and non-­ Fischer, C.  S., et  al. (1996). Inequality by design:

school environments, and Black/White inequalities Cracking the bell curve myth (1st ed.). Princeton

in children’s learning. American Sociological Review, University Press. Retrieved http://www.amazon.com/

74(5), 685–708. Retrieved http://asr.sagepub.com/ dp/0691028982

content/74/5/685 Fryer, R.  G., & Levitt, S.  D. (2004). Understanding the

DiMaggio, P. (1982). Cultural capital and school success: Black–White test score gap in the first two years of

The impact of status culture participation on the grades school. The Review of Economics and Statistics,

of U.S. high school students. American Sociological 86(2), 447–464. Retrieved http://www.jstor.org/

Review, 47(2), 189–201. stable/3211640

Diprete, T.  A., & Buchmann, C. (2013). The rise of Fryer, R. G., & Levitt, S. D. (2006). Testing for racial

women: The growing gender gap in education and differences in the mental ability of young children.

what it means for American schools. CUP Services. National Bureau of Economic Research (Working

Retrieved http://www.amazon.com/The-Rise-Women- Paper). http://www. nber.org/papers/w12066

Education-American/dp/0871540517/ref=sr_1_1?ie= Gamoran, A. (2016). Gamoran comment on Downey and

UTF8&qid=1373561211&sr=8-1&keywords=the+ris Condron. Sociology of Education, 89(3), 231–233.

e+of+women Retrieved December 21, 2016, http://soe.sagepub.

Dobbie, W., & Fryer, R. G. (2011).Are high quality schools com/cgi/doi/10.1177/0038040716651931

enough to close the achievement gap? Evidence from Gamoran, A., & Mare, R.  D. (1989). Secondary school

a social experiment in Harlem. American Economic tracking and educational inequality: Compensation,

Journal: Applied Economics, 3(3), 158–187. reinforcement, or neutrality? American Journal of

Downey, D. B., & Pribesh, S. (2004). When race matters: Sociology, 94(5), 1146–1183.

Teachers’ evaluations of students’ classroom behavior. Gangl, M. (2010). Causal inference in sociological

Sociology of Education, 77(4), 267–282. research. Annual Review of Sociology, 36, 21–47.

Downey, D.  B., & Condron, D.  J. (2016). Fifty years Gibbs, B. (2010). Reversing fortunes or content change?

since the Coleman report: Rethinking the relation- Gender gaps in math-related skill throughout child-

ship between schools and inequality. Sociology of hood. Social Science Research, 39(4), 540–569.

Education, 89(3), 207–220. Han, W.-J. (2008). The academic trajectories of chil-

Downey, D. B., von Hippel, P. T., & Broh, B. A. (2004). dren of immigrants and their school environments.

Are schools the great equalizer? Cognitive inequal- Developmental Psychology, 44(6), 1572–1590.

ity during the summer months and the school year. Hanushek, E. A. (1992). The trade-off between child

American Sociological Review, 69(5), 613–635. quantity and quality. Journal of Political Economy,

Downey, D. B., von Hippel, P. T., & Hughes, M. (2008, 100(1), 84–117. Retrieved January 31, 2013. http://

July). Are “failing” schools really failing? Sociology www.jstor.org/stable/2138807.

of Education, 81, 242–270.

70 D. B. Downey et al.

Hayes, D. P., & Grether, J. (1983). The school year and Ready, D. D., LoGerfo, L. F., Burkam, D. T., & Lee, V. E.
vacations: When do students learn? Cornell Journal of (2005). Explaining girls’ advantage in kindergarten
Social Relations, 17, 56–71. New York City. literacy learning: Do classroom behaviors make a
difference? The Elementary School Journal, 106(1),
Heyns, B. (1978). Summer learning and the effects of 21–38.
schooling. New York: Academic.
Reardon, S. F. (2011a). The widening academic achieve-
Ho, A. D., & Reardon, S. F. (2011). Estimating achieve-
ment gaps from test scores reported in ordinal “pro- ment gap between the rich and the poor: New evidence
ficiency” categories. Journal of Educational and and possible explanations (pp.  91–116). New  York:
Behavioral Statistics, 37(4), 489–517. Retrieved Russell Sage Foundation.
April 19, 2014, http://jeb.sagepub.com/cgi/ Reardon, S. F. (2011b). The widening socioeconomic
doi/10.3102/1076998611411918 status achievement gap: New evidence and possible
explanations. In Whither opportunity: Rising inequal-
Jencks, C.  S. (1972). The Coleman report and ity, schools, and children’s life chances (pp. 91–115).
the conventional wisdom. In Mosteller, F. & Washington, DC: Brookings Institution.
Moynihan, D.  P. (Eds.), On equality of educa- Reardon, S.  F., & Galindo, C. (2009). The Hispanic–
tional opportunity (pp.  69–115). New  York: White achievement gap in math and reading in the
Vintage. Retrieved https://courses.utexas.edu/bbc- elementary grades. American Educational Research
swebdav/pid-2031893-dt-content-rid-2384509_1/ Journal, 46(3), 853–891.
xid-2384509_1 Reardon, S. F., Cheadle, J. E., & Robinson, J. P. (2009).
The effect of Catholic schooling on math and read-
Jennings, J.  L., Deming, D., Jencks, C., Lopuch, M., & ing development in kindergarten through fifth grade.
Schueler, B. E. (2015). Do differences in school qual- Journal of Research on Educational Effectiveness,
ity matter more than we thought? New evidence on 2(1), 45–87. Retrieved http://www.tandfonline.com/
educational opportunity in the twenty-first century. doi/abs/10.1080/19345740802539267
Sociology of Education, 88(1), 56–82. Rothstein, R. (2004). Class and schools: Using social,

Kozol, J.  (1991). Savage inequalities: Children in economic, and educational reform to close the Black–­
America’s schools (1st ptg). New  York: Harper White achievement gap. Washington, DC/New York:
Perennial. Economic Policy Institute/Teachers College.
Thernstrom, A. M., & Thernstrom, S. (2003). No excuses:
LoGerfo, L.  F., Nichols, A., & Reardon, S.  F. (2006). Closing the racial gap in learning. New York: Simon
Achievement gains in elementary and high school. & Schuster.
Washington, DC: Urban Institute. Von Drehle, D. (2010). The case against summer vaca-
tion. Time.
Meyer, J.  W. (2016). Meyer comment on Downey and von Hippel, P.  T., & Hamrock, C. (2016). Do test score
Condron. Sociology of Education, 89(3), 227–228. gaps grow before, during, or between the school years?
Retrieved December 21, 2016, http://soe.sagepub. Measurement artifacts and what we can know in spite
com/cgi/doi/10.1177/0038040716651679 of them. Educational Researcher, 45(8), 443–453.
von Hippel, P. T., Powell, B., Downey, D. B., & Rowland,
Morsy, L., & Rothstein, R. (2016). Mass incarceration N.  J. (2007). The effect of school on overweight
and children’s outcomes. Washington, DC: Economic in childhood: Gain in body mass index during the
Policy Institute. school year and during summer vacation. American
Journal of Public Health, 97(4), 696–702. Retrieved
Murnane, R. J. (1975). The impact of school resources on November 30, 2014, http://www.pubmedcentral.nih.
the learning of inner city children. Cambridge, MA: gov/articlerender.fcgi?artid=1829359&tool=pmcentre
Ballinger Publishing Company. z&rendertype=abstract
Whitehurst, G. J. (2016). Family support of school readi-
Oakes, J. (1985). Keeping track: How schools structure
inequality (1st ed.). New Haven: Yale University ness? Contrasting models of public spending on
Press. children’s early care and learning. Evidence Speaks
Reports, Vol 1.
Quinn, D. (2015). Kindergarten Black–White test score Yoon, A., & Merry, J. J. (2015). Understanding the role of
gaps: Re-examining the roles of socioeconomic sta- schools in the Asian–White gap: A seasonal compari-
tus and school quality with new data. Sociology of son approach. In American Sociological Association,
Education, 88(2), 120–139. Chicago, IL.

Quinn, D. M., Cooc, N., McIntyre, J., & Gomez, C. J.
(2016). Seasonal dynamics of academic achievement
inequality by socioeconomic status and race/ethnicity.
Educational Researcher, 45(8), 443–453.

Raudenbush, S.  W., & Eschmann, R.  D. (2015). Does
schooling increase or reduce social inequality? Annual
Review of Sociology, 41, 443–470.

Part II
The Changing Demographics of Social

Inequality

Race, Class, and Theories 4
of Inequality in the Sociology
of Education

Samuel R. Lucas and Véronique Irwin

Abstract 4.1 I ntroduction
After explaining a focus on race and class
inequality, we briefly sketch contemporary Multiple analysts have documented a relation
racial and socioeconomic inequality in educa- between educational outcomes and students’
tion. Then, we convey key criteria used to socioeconomic (e.g., Blau and Duncan 1967;
select which of the many theories to consider. Featherman and Hauser 1978; Sewell and Hauser
We then describe ten theories of racial/ethnic- 1980) and racial/ethnic (e.g., Featherman and
and class-linked inequality in education. After Hauser 1978; Jaynes and Williams 1989; Jencks
the last theory has been described, we identify and Phillips 1998) origins. Such works have doc-
selected points of contact across the theories. umented the changing power of class and race/
We then discuss three examples of existing ethnicity, but none have documented the eradica-
research to demonstrate how research may be tion of either effect. Additional research indicates
used to assess the theories. We conclude by powerful education associations with and effects
offering suggestions for next steps. on multiple individually and societally conse-
quential outcomes, from matters as material as
We thank Jan Jacobs, Susan Schacht (posthumously), health (e.g., Kimbro et  al. 2008) and mortality
H.  Sorayya Carr, Aimée Dechter, and Olivia Garcia for (e.g., Kitagawa and Hauser 1968) to matters as
many helpful conversations. This research has been sup- ideological as political efficacy (e.g., Paulsen
ported by funding from the NSF-GRFP (Grant No. DGE 1991) and prejudice attitudes on grounds of sex
1106400). All errors and omissions are the fault of the (e.g., Cherlin and Walters 1981), race (e.g., Bobo
authors. Please direct correspondence to Samuel R. Lucas / and Licari 1989), and anti-semitism (in liberal
Sociology Department / University of California-B­ erkeley democracies) (Weil 1985). Because effects of
/ 410 Barrows Hall #1980 / Berkeley, CA 94720-1­ 980 or education are wide-ranging, class and racial/eth-
by e-mail to [email protected] nic inequalities in education ramify far beyond
S. R. Lucas (*) · V. Irwin the realm of schooling. Perhaps owing to the
Department of Sociology, University of California-­ importance of education in individuals’ well-­
Berkeley, Berkeley, CA, USA being and thus society’s capacities, the intransi-
e-mail: [email protected]; [email protected] gence of class and race effects on educational
outcomes has motivated many analysts to attempt
explanations. In the pages below we attend to
some of the most widely-researched and/or
promising explanations at present.

© Springer International Publishing AG, part of Springer Nature 2018 73
B. Schneider (ed.), Handbook of the Sociology of Education in the 21st Century, Handbooks
of Sociology and Social Research, https://doi.org/10.1007/978-3-319-76694-2_4

74 S. R. Lucas and V. Irwin

One could take one of two vantage points for (GATE), special education assignments), treat-
considering the relation between class and educa- ment (e.g., suspensions, expulsions), academic
tion. One approach considers how the socioeco- performance (e.g., grades, test scores) and attain-
nomic position of children’s, adolescents’, and ments (e.g., years of school completed, college
young adults’ families of origin affect children’s, degree attainment, advanced degree attainment).
adolescents’ or young adults’ educational trajecto- Inequalities can exist along lines of class, race,
ries and outcomes. A second approach studies how gender, sexual orientation, disability status, and
young adults’ education matters for their own more. This chapter focuses specifically on the
placement in the labor force, occupational distribu- inequalities between students from different socio-
tion, and earnings distribution. Both approaches economic and racial/ethnic backgrounds. In this
are important, but we will focus on the former section, we first explain our focus on race/ethnic-
because the research claiming racial fluidity (e.g., ity and class; afterwards, we convey a snapshot of
Saperstein and Penner 2010, 2012) is seriously class and racial/ethnic inequality in education.
flawed in the U.S. context (Lucas and Beresford
2010, pp. 32–37; Defina and Hannon 2016; Kramer 4.2.1 W hy Race and Class?
et al. 2016), making it more correct to consider a
persons’ race as a factor in their educational trajec- The decision to focus on race and class necessar-
tories, not as a result thereof. To make our focus ily omits many other factors of great importance.
consistent, we will address race and class effects on One could justify the decision by noting that it
education, not education effects on class or race. reflects a widespread emphasis on these ascribed
characteristics as bases of stratification beyond
Even so, some theories explain race and/or the school. For social reproduction in education,
class effects on education by considering how edu- however, the interest in race and class is more
cation affects later class position. Thus, our stark than a historical artifact of the discipline.
division, while empirically possible, is not neces- Particularly in the United States, where public
sarily always recognized in the literature. Where schools are funded through property taxes and
necessary, we will follow the theoretical claims, students are generally allocated to schools based
and not enforce an arbitrary narrowing of focus. on the neighborhood in which they live,
­generations-long patterns of the geographic con-
We begin by justifying our joint focus on race centration of disadvantage are amplified in educa-
and class inequality and by providing a brief sketch tion. Because neighborhoods are segregated along
of contemporary racial and socioeconomic inequal- race and class lines rather than along other very
ity in education. Afterwards, we introduce key cri- important axes of stratification, such as gender,
teria used in selecting which of the many theories and because construction of school catchment
to consider. Then, ten theories are conveyed. After areas can result and has resulted in even more
the last theory has been described, we identify racial/ethnic and class segregation than neighbor-
selected points of contact across the theories. In our hoods would actually have (Saporito and Sohoni
next-to-final section, we draw on empirical research 2006, 2007), it is especially important to under-
to show how the theories might be assessed in an stand how education is implicated in these
effort to trim the list of viable theories. We con- inequalities.
clude by offering suggestions for next steps.
Race and class, for better or worse, are also key
4.2 R ace and Socioeconomic sites of struggle in educational policy reform in
Status: Processes the United States. This is especially apparent in
and Inequalities postsecondary education, likely because bache-
lor’s degrees long ago replaced high school diplo-
Across developed nations, inequalities exist mas as the prerequisite for good jobs (Jencks et al.
between more and less advantaged students in 1988) while access to the institutions that award
opportunities (e.g., gifted and talented education

4  Race, Class, and Theories of Inequality in the Sociology of Education 75

those degrees remains more a privilege than a Table 4.1  Average scores of U.S. 15-year-old students
right. Most visibly, race-based affirmative action on 2012 PISA assessmentsa
Reading Math Science
remains a hotly contested issue. At the same time,
reproduction of stratification at these institutions Avg. s.e. Avg. s.e. Avg. s.e.
through legacy admissions policies (Howell and
Turner 2004), which function as affirmative action OECD 496† 0.5 494† 0.5 501† 0.5
for wealthy Whites, occurs almost completely Average
without protest. Therefore, among other reasons,
U.S. Averageb 498 3.7 481* 3.6 497 3.8

Percent of students in school receiving free or reduced
price lunchc

understanding how inequalities along race and Less than 559† 8.6 540† 7.8 556† 7
class lines play out in education, both before and 10%
after matriculation to college, is essential to better 524* 5.3 513* 5.7 528* 6.5
inform policy decisions. 10–24% 519 6.7 506 6.4 523 5.6
479* 4.7 464* 4.6 483* 5.0
25–49.9%

50–74.9%

75% or more 452* 8.5 432* 7.2 442* 8.1

4.2.2 Inequalities in Education Student race/ethnicityd
by Race and Socioeconomic
Class: A Snapshot White 519† 4.1 506† 3.7 528† 3.7

Black 443* 8.3 421* 6.2 439* 6.8

Hispanic 478* 4.5 455* 4.8 462* 4.7

Asian 550* 8.1 549* 9.0 546* 8.6

Every 3 years, the Program for International Multiracial 517 7.6 492* 7.4 511 7.8

Student Assessment (PISA) tests the reading, † reference group, * p < 0.05
math, and science literacy of 15-year-old stu- aSource: National Center for Education Statistics,
dents in the 34 nations from the Organization for Archived International Data Table Library
Economic Cooperation and Development bSignificance stars are relative to OECD average
cIncludes only students in public schools. Significance

(OECD), along with 31 partner nations/econo- stars in this portion of the table refer to the difference rela-
mies. Students’ report of their parents’ education, tive to the FRL group in the immediately preceding row
occupation, and “classical” cultural material in dSignificance stars in this portion of the table are relative
the home are used to construct an index of eco- to White students

nomic, social, and cultural status (ESCS).

National Center for Education Statistics (NCES) larly find that students from higher socioeco-

data allow comparison of PISA scores by stu- nomic backgrounds perform better on

dents’ national quartile rank on the ESCS index. international assessments in all OECD countries.

With only one exception (students in the second Thus, while the remainder of the chapter focuses

ESCS quartile in Liechtenstein outperform their heavily on evidence from the United States, we

third quartile peers by a statistically non-­ treat socioeconomic inequalities in education as a

significant margin), students from higher-ESCS universal dilemma.

quartiles perform better in math and reading than Table 4.1 demonstrates strong socioeconomic

their (adjacent quartile) lower-ESCS compatriots and racial patterns in test performance in the

in every participating country. Over 90% of United States. Across all subjects, scores decline

country-quartile differences were statistically steadily as one moves from students who attend

significant.1 Carnoy and Rothstein (2013) simi- schools with the fewest socioeconomically disad-

vantaged peers to those who attend schools with

1 Three comparisons were made in each of 65 countries the most socioeconomically disadvantaged peers.
(2nd-1st quartile, 3rd-2nd, and 4th-3rd), for a possible Moreover, because socioeconomic disadvantage

195 significant within-country quartile gaps in each sub-

ject. Non-significant differences were found in only 17

countries for math and 21 countries for reading and gener- average. Data from the National Center for Education

ally only in 1 of the 3 comparisons. In all other instances, Statistics International Data Table Library: Table B.1.119

students in higher quartiles performed statistically signifi- (PISA 2012 Results Table M8) and Table B.1.95 (PISA

cantly better than their adjacent lower-quartile peers on 2012 Results Table R8).

76 S. R. Lucas and V. Irwin

is measured at the school level, rather than the In comparison, nearly 75% of students attending
student level, these figures may underestimate schools with no more than 1 in 10 students in
the achievement gap between the most advan- poverty exceed performance level 2. In contrast
taged (wealthy students attending wealthy in hyperpoverty schools, schools with three-­
schools) and most disadvantaged (poor students quarters or more students in poverty, barely 25%
attending poor schools) students. Black and of students exceed level 2. For race/ethnicity,
Hispanic students also underperform relative to shown in Fig. 4.2, similar disparities are evident.
their White and Asian peers. Given the relative
concentration of Black and Hispanic students in It is difficult to see how a nation can maintain
the most socioeconomically disadvantaged a productive economy if large numbers of its ado-
schools, these achievement gaps reflect com- lescents do not have the mathematics literacy to
pound disadvantages. execute sequential procedures with basic reason-
ing. It is difficult to see how future citizens will
The test scores summarize socioeconomic and make well-informed decisions in a democracy if
racial/ethnic differences in performance, but may substantial proportions of its adolescents cannot
not make it clear what differences in test scores integrate assumptions and connect them to real-­
mean for differences in students’ capabilities. world arguments. Thus, failure to reach noted
PISA reports also indicate students of different benchmarks, and the race- and class-linked nature
socioeconomic contexts and racial/ethnic back- of the shortfall, is consequential not only for indi-
grounds’ distribution along benchmarks of math- viduals, but also (perhaps) for society.
ematics literacy. Abstracting from the NCES
report on PISA (NCES 2013, p. 3), one can sum- Educational stratification occurs not only in
marize the levels as in Table 4.2. performance at a given grade or level of school-
ing, but in the highest level of education that
Considering these capability thresholds, ­individuals pursue and complete. While the
Fig. 4.1 sketches the distribution of U.S. 15-year- expansion of the community college in the United
old students by the proportion of schoolmates States has opened the door to postsecondary edu-
eligible for free or reduced price lunch. In Fig. 4.1 cation for many low-SES and underrepresented
(and Fig.  4.2, below), the marks are connected minority students, both enrollment and persis-
with lines to facilitate recognition of the patterns. tence in college continue to lag for these groups.
Considering the patterns, slightly less than 59% The first panel of Table 4.3 presents the college
of the students attending schools with one-­quarter enrollment rates of recent high school completers
over three decades, with the most recent year
Table 4.2 Proficiency levels in mathematics, PISA chosen to align with the PISA assessments from
15-year-olds Table  4.1.2 The second panel presents degree
attainment after 6 years for students who enrolled
Level Students are able to full-time for the first time in a bachelor’s degree
1 “answer clearly defined questions with routine program in the 2003–2004 school year. These
data, taken from the Current Population Survey
procedures” (CPS) and Beginning Postsecondary Study
2 “make direct inferences and provide literal (BPS), respectively, show that Black, Hispanic,
and lower-i­ncome students are not only less
interpretations” likely to enroll in college than their White and
3 “execute sequential procedures with basic higher-SES peers, they are less likely to complete
a degree if they do.3 As with their performance on
reasoning”
4 “integrate assumptions and connect to 2 Recent high school completers are 16- to 24-year-olds
who completed high school during the calendar year.
real-world arguments”
5 “compare and select strategies to develop 3 By reporting enrollment and persistence only for recent
high school completers (CPS) these figures overlook the
complex models”
6 “develop and communicate complex models for

novel contexts”

to one-half of students qualifying for free or
reduced price lunch exceed performance level 2.

4  Race, Class, and Theories of Inequality in the Sociology of Education 77

Fig. 4.1 Math
distribution by school
poverty, U.S. 15-year-
olds, 2012

Fig. 4.2 Math
distribution by race/

ethnicity, U.S. 15-year-

olds, 2012

important increase in “non-traditional” college students the PISA assessments, Asian American students
(CITE). Thus, enrollment rates are likely understated outperform White students, both attending and
because of the omission of older students, while persis- completing college at higher rates.4
tence rates are likely overstated because of the omission
of students who begin postsecondary education The tables above report the connection
part-time. between socioeconomic position and racial/eth-
nic category on the one hand, and achievement or
Because percentages have a ceiling of 100% and a floor of attainment outcomes on the other. Yet, these out-
0%, assessing change through percentages is often mis- comes are produced by opportunity and t­reatment
leading. Odds ratios provide a better indicator. Odds ratios
between High/Mid SES are 2.75, 2.76, and 2.22 across 4 Degree completion rates may not differ significantly.
cohorts respectively. Mid/Low SES odds ratios are 1.74, NCES QuickStats does not provide standard errors for
1.35, and 1.83, and High/Low SES odds ratios are 4.81, BPS.
3.74, and 4.05 across the cohorts, respectively. The advan-
tage of High SES students compared to Mid and Low SES
students is extremely large.

78 S. R. Lucas and V. Irwin

Table 4.3  College enrollment and persistence (%)

Recent high school completers enrolled in 2- or 4-year collegea Attainment by 08–09 for students
(standard errors in parentheses) starting bachelor’s in 03–04b

1992 2002 2012 BA AA Neither

Total 63.2 (0.92) 63.7 (0.78) 66.8 (0.94) 63.2 2.9 33.9

Socioeconomic statusc

Low 43.6 (2.60) 50.9 (2.14) 50.3 (2.63) 51.7 2.7 45.6
64.3 3.6 32.1
Middle 57.4 (1.26) 58.4 (1.08) 64.9 (1.26) 77.7 1.7 20.6

High 78.8 (1.38) 79.5 (1.20) 80.4 (1.59)

Race/ethnicity

White 64.2 (1.06) 66.5 (0.97) 67.6 (1.12) 67.4 3.3 29.3
47.6 2.2 50.2
Black 50.0 (2.98) 57.3 (2.33) 60.5 (2.64) 47.5 2.5 49.9
73.0 0.4 26.6
Hispanic 58.2 (5.04) 54.8 (2.75) 65.9 (1.99) 56.6 2.6 40.8

Asian – – 82.3 (3.59)

Other – – –

aSource: NCES tabulations from Current Population Survey (CPS)
bSource: BPS:2009 Beginning Postsecondary Students, NCES QuickStats
cSES for enrollment rates is provided by the CPS simply as “low,” “middle,” and “high.” From BPS these groups are
based on dependent students’ parental income in 2003–2004 (lowest 25%, middle 50%, highest 25%)

processes within education. If there are class and/ schools also had higher police presence than did
or racial/ethnic inequalities in in-school opportu- wealthier schools, suggesting students in poorer
nity and treatment, then observed class- and schools engage their learning under the watchful,
racial/ethnic-linked differences in outcomes are possibly intimidating, and potentially anxiety-­
at least somewhat to be expected. Are there inducing gaze of state surveillance officers. These
opportunity and treatment differences by race differences in students’ experience of schooling
and class? certainly contextualize achievement and attain-
ment differences analysts have documented.
Table 4.4 addresses opportunity, and indicates Taken together, the information provided in
that White and Asian students are two to three Tables 4.1, 4.3, 4.4, and 4.5 indicate that both
times as likely to enter gifted and talented educa- processes and outcomes are unequal, and connect
tion (GATE) than are Black students. At the same in multifaceted and intertwining ways.
time, Black students are more likely than White
students, and four times more likely than Asian Many theories have been advanced to explain
students, to be assigned to special education. the race and class achievement gaps described
And, while in 2009 nearly two-thirds of Asian above. The remainder of the chapter focuses on
students enrolled in Advanced Placement courses, ten key theories of racial/ethnic and class inequal-
less than a quarter of Black students enrolled in ity. We select these theories based on criteria we
Advanced Placement courses. Advanced establish in the next section.
Placement also tracked with school poverty, as
the poorer the school, the less likely students 4.3 Theories of Inequality
were to enroll in Advanced Placement courses.
We focus on theories because they are the tools
Table 4.5 continues the documentation of dif- by which we can interpret the changing facts of
ference. In 2007, Black students were over 2.5 inequality. We first convey criteria that all theo-
times more likely to be suspended than were ries of inequality must meet. Then, we describe
Whites, and over 9 times more likely to be the characteristics of expansive and narrow theo-
expelled than were Whites, even though research ries of inequality.
shows Blacks have infraction rates comparable to
(e.g., McNulty and Bellair 2003) or lower than
(e.g., Bachman et  al. 1991) Whites. Poorer

4  Race, Class, and Theories of Inequality in the Sociology of Education 79

Table 4.4  Inequalities in opportunity: special education, GATE, and College prep.

Percent in SPEDa Percent in GATE Percent of graduates who earned dual credit or
programb AP creditc

Dual credit AP courses

2007 2004 2006 2005 2009 2005 2009

Total 4.55 6.70 6.70 8.9 9.3 28.8 36.3

(0.05) (0.04) (0.60) (0.76) (0.68) (0.94)

Race/ethnicity

White 4.03 7.90 8.00 10.0 9.7 29.8 37.3

(0.07) (0.07) (0.73) (1.00) (0.86) (0.95)

Black 6.59 3.50 3.60 4.7 6.4 18.3 22.2

(0.05) (0.05) (0.80) (0.99) (0.97) (1.00)

Hispanic 4.95 4.30 4.20 7.7 10.8 28.5 33.8

(0.05) (0.04) (1.10) (1.18) (1.29) (1.30)

Asian 1.78 11.90 13.10 9.2 9.2 47.2 66.3

(0.20) (0.29) (1.25) (1.46) (2.25) (2.56)

Percent of students in school eligible for free or reduced-price lunch

Less than 25% – – – 9.8 9.3 32.9 44.9

(1.32) (1.56) (1.27) (1.72)

25–49.9% – – – 9.6 9.2 24.9 31.3

(1.31) (1.25) (1.16) (1.40)

More than 50% – – – 5.9 9.1 24.5 28.6

(1.32) (1.33) (1.46) (1.64)

aFigures refer to students of all ages receiving Special Education due to a “specific learning disability” or being “emo-
tionally disturbed” (these subgroups were chosen because they are likely more discretionary than physical disabilities,
autism, or “mental retardation”). Source: U.S. Department of Education, Office of Special Education Programs (OSEP),
2007 [NCES Table 8.1b]
bFigures refer to elementary and high school public school students in Gifted and Talented Education programs. Source:
U.S. Department of Education, National Center for Education Statistics, High School and Beyond Longitudinal Study
of 1980 Sophomores (HS&B-So:80/82), “High School Transcript Study”; and 1990, 1994, 1998, 2000, 2005, and 2009
High School Transcript Study (HSTS) [NCES Table 225.30]
cNumber and percentage of public high school graduates taking dual credit (courses that earn both high school and
college-level credit), Advanced Placement (AP), and International Baccalaureate (IB) courses in high school. Source:
U.S. Department of Education, National Center for Education Statistics, 2000, 2005, and 2009 High School Transcript
Study (HSTS) [Table 225.60]

4.3.1 Theoretical Criteria ena. Were this not possible evaluation of the
theory would also be impossible. Indeed, if one
We agree with Silberberg (1990, p.  10) that “A cannot map conceptual entities to observed enti-
theory, in an empirical science, is a set of explana- ties, doubt arises as to whether the statements are
tions or predictions about various objects in the relevant for the real social world.
real world.” For claims to coalesce into a theory
five criteria must be met. First, the claims must ref- Third, the claims, once mapped onto real enti-
erence conceptual entities (e.g., classes, ethnic ties, must imply some observable patterns, events,
groups). These entities are conceptual in that no outcomes that may or may not pertain. That is, there
pure example of the entity may exist. For example, must be multiple possible states of affairs, and the
essentialists notwithstanding, no member of an claims and the mapping must imply at least one
ethnic group is only a member of an ethnic group. fewer state of affairs than is otherwise possible. In
Consequently, one can never attain the pure form other words, the implications must be falsifiable.
of the conceptual entity. Even so, to be a theory one
or more claims must reference conceptual entities. Fourth, the postulates cannot be internally
contradictory. One cannot claim, for example,
Second, it must be possible to map the con- that A = B, B = C, and C ≠ A. If a set of claims
ceptual entities to observable entities or phenom- are internally contradictory it is impossible to
assess the veracity of the claims.

80 S. R. Lucas and V. Irwin

Table 4.5  Inequalities in treatment: discipline and indicators of potential disciplinea

Totalb Suspended 2007 Expelled 2007
White 2003 24.5 2003 3.2
Black 20.4 17.7 3.9 1.1
Hispanic ! 18.1 49.0 3.2 10.3
Asian/Pacific 30.2 26.5 8.5 4.1
Islander !! 21.9 12.8 3.6
11.6
Total (public
schools)c 2011–2012 Daily presence of police or security
Less than 25% Random metal detector checks (%) se
26–50% (%) se 28.1 (0.51)
51–75% 5.0 (0.32)
76 or More 26.3 (1.39)
1.9 (0.45) 24.1 (0.99)
2.2 (0.40) 25.8 (1.21)
5.3 (0.65) 36.2 (1.52)
9.5 (0.88)

! Interpret “expelled” data with caution. The coefficient of variation (CV) for this estimate is 30% or greater
!! Interpret “suspended” and “expelled” data with caution. The coefficient of variation (CV) for this estimate is 30% or
greater
aTables included both discipline and potential indicators because statistics (from public-use data) were available only
broken down by either race or class for each
bTotal includes other racial/ethnic groups not shown separately. Source: U.S. Department of Education, National Center
for Education Statistics, Parent and Family Involvement in Education Survey of the National Household Education
Surveys Program (PFI-NHES), 2003 and 2007
cSource: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS),
“Public School Principal Data File” and “Private School Principal Data File,” 2011–2012

Fifth, the postulates cannot be tautological. narrowest theories of inequality are specific,
One cannot claim, for example, that A = B, and static, and merely correlational.
B = A. If a set of claims are tautological, nothing
is gained by assessing the claims. Generality  What we call specific theories apply
to only one outcome and/or apply to only one cat-
Sociological theories are usually conveyed egorical system. In contrast, general theories of
informally, in words alone. Formalization of the- inequality apply to multiple outcomes and multi-
ories—often their translation from words to ple categorical systems. So, for example, a spe-
mathematical relations—can make it easier to see cific theory might explain only class inequality in
and root out tautologies and contradictions. The test scores, which is less general than a theory
dearth of formalization means that it is possible that explains inequality with respect to both class
that some claims offered as a theory may some- and race in test scores and college entrance.
day be shown to fail to satisfy one or more of the Parsimony is a valued criterion for theories to sat-
criteria above. However, without formalizing the isfy and, all else equal, a general theory that
theories, we use these criteria to select theories explains multiple outcomes for multiple social
for attention. divisions is more parsimonious than is the sum of
specific theories needed to explain each single
4.3.2 C haracteristics of Expansive outcome for each social division.
and Narrow Theories

The most expansive theories of inequality are Dynamics  All theories of inequality focus on
general, dynamic, and identify mechanisms. The some form of the XY relation in Fig.  4.3. The

4  Race, Class, and Theories of Inequality in the Sociology of Education 81

Fig. 4.3  Re-labeling positions in a less than fully enlight- C: X = Amount of financial resources, 1 = None, 2 = A
ening way little, 3 = Some, 4 = A lot
A: X  =  Class categories, 1  =  underclass, 2  =  working D: X = Enjoys school, 1 = None, 2 = A little, 3 = Some,
class, 3 = small proprietor, 4 = capitalist 4 = A lot
B: X = Racial/ethnic groups, 1 = Blacks, 2 = Latino/as, E: X = Number of teachers certified, 1 = None, 2 = A few,
3 = Whites, 4 = Asians 3 = About half, 4 = Almost all

relation may be linear or curvilinear; positive or makes the very plausible claim that financial
negative; and reflected in a bar graph as in Fig. 4.3 resources are associated with education out-
(for categorical X variables), in a line-graph (for comes. However, as an explanation of the XY
continuous X-variables), or in other ways. Given relation, the claim in note C simply replaces 1, 2,
our focus, in Fig. 4.3 X might indicate parents’ 3, and 4 class categories with labels for financial
class category, and Y might be measured achieve- resources: None, A little, Some, and A lot.
ment (e.g., test scores). Note, before we proceed,
that the bars summarize the relationship. Surely, The explanation that children who attend schools
some persons in category 1 on X obtain higher Y that match their culture do better may be offered to
than the bar indicates. Some persons in category explain racial differences in achievement. But,
1 on X obtain lower Y than the bar indicates, too. again, this threatens to simply substitute note D for
The claim is not that every person is right at the note B. A similar substitution—for notes A and/or
level of the bar; the claim is that the bars sum- B—is offered by note E.
marize differences in the averages for persons
located in different positions on X. If there were True though the claims expressed in notes C,
no average differences, all the bars would be the D, and E may be, the simple re-labeling does not
same height, and Y would be mean independent take us very far or, rather, it takes us in one pos-
of X (Goldberger 1991, pp.  61–63), suggesting sibly helpful direction, but not in another one. A
no causal effect of X on Y. simplistic example may make the point. The re-­
labeling may take us to an assessment of what an
The differences in the heights of the bars individual student with a given value of X might
reflect the relationship between X and Y, and that do to perhaps change their prospects on Y. If stu-
relationship is the fundamental matter to be dents in category two average lower achievement
explained. Many claims focus so much on the than their category three peers, the re-labeling by
specific relationship in the data that the explana- note D suggests that category two students might
tions threaten to provide mere substitute labels deepen their familiarity and understanding of the
for the observed relation. So, for example, notes culture of the school, and then their performance
A and B in Fig. 4.3 reflect two variables known to on Y might improve. Or, if one is uncomfortable
be associated with education outcomes. Note C with a blaming the victim approach, one could
use the re-labeling of note D to claim that schools

82 S. R. Lucas and V. Irwin

attended by mostly category two students should along which inequality is a concern; for example,

become more culturally matched to that specific in our work the X-dimension could be socioeco-

population of students. Note that both counsels nomic status/class. The Y-variable, therefore,

leave the relation intact; both simply change the would be the outcome that is distributed

score on “cultural match” for some students in unequally—in our case it may be measures of

some schools. educational attainment (years of schooling, pro-

The direction the re-labeling does not go is portion obtaining a bachelor’s degree), cognitive

toward telling us why the heights of the bars are achievement, or some other education treatment

sloped as they are, and not more equal (flatter or outcome. In Panel 1 entities at point A on X

sloped) or less equal (steeper sloped). To deter- have certain values on Y; moving an entity from

mine what makes slopes steepen or flatten is a point A to point B will give them higher

complex matter, but one essential part of the task (expected) values on Y. This is the most common

requires embedding any single claim in a coher- kind of change inequality analysts address. We

ent web of claims. Together such a web would term this kind of change cross-sectional change,

provide resources to aid us in understanding the which should signify that difference between per-

dynamics of inequality, not simply offer a possi- sons at points A and B, not change (i.e., not

bly tautological, often highly individualistic re-­ movement from point A to point B), has actually

labeling of observed patterns. been studied.

To clarify, there are, of course, multiple kinds In Panel 2 entities at point A move to point A′,
of change. Claims about inequality necessarily while entities at point B move to point B′. Both
address at least one. Panel 1 of Fig. 4.4 traces the moves in Panel 2 constitute change, but obvi-

most common kind of change claim-sets refer- ously the order of the entities on Y remains

ence. The variable X represents the variable unchanged, and, indeed, the amount of inequality

Fig. 4.4  Types of
change

4  Race, Class, and Theories of Inequality in the Sociology of Education 83

is also unchanged. Essentially, what changes in ­magical about zero; it only appears to be the
Panel 2 is the marginal distributions of X and magic number for three chained reasons. First,
Y.  Both X and Y are higher after the change. few social theories calibrate their claims pre-
However, the relation between them is unchanged. cisely. Second, this means that most theories can-
We term this kind of change marginal change not attach numeric values that will signal
because all that has changed is the marginal (i.e., important thresholds of change. Third, because
univariate) distributions of X and Y. of this, most theories are stated in terms or trans-
lated into terms of whether statistical relations
An example of marginal change might be are positive or negative, thus institutionalizing
helpful. If all prices, including the price of labor zero as the key criterion for extracting conclu-
(i.e., wages) and capital, doubled, everyone sions concerning a theory. This is clear in that if
would receive 100% more for any sale and every- there were a theory of the nation-state which,
one would have to pay 100% more for anything once traced precisely, implied that the simple
they buy. Everyone would have twice as much regression coefficient summarizing the XY rela-
money as now, but no one would be richer or tion will fall between 1 and 1.5 in “true” welfare
poorer, as the relation between all prices (as well state economies, but be higher in laissez-faire
as everyone’s ability to pay) would be unchanged. economies, observing the coefficient shift over a
decade from 1.2 to 1.8 would signify a regime
Panels 1 and 2 do not contain the kind of change, from welfare state to laissez-faire.
change we mean when we indicate that a theory Consequently, just as dynamic theories address
will be dynamic. A dynamic theory is one that can changes within a regime, more fully dynamic the-
account for possible shifts in the structure of ories also address regime change—they identify
inequality. Panels 3 and 4 more accurately reflect thresholds of regime change, and they identify the
the criterion. In Panel 3, the slope of line AB mechanisms that cause or prevent the crossing of
shifts, which is reflected in line A″B″. We term those thresholds. Thus, both Panels 3 and 4 indi-
this kind of change effect magnitude change. And, cate that expansive theories will address the
in Panel 4, the slope of the line shifts so much as causes of the direction and size of the slope and
to reverse the relationship between X and Y, from its change over time, and, given the tenets of the
positive to negative. Such shifts are rare and theory and their precision, some more fully
momentous. For example, the Russian revolu- dynamic theories can signify regime change.
tions of 1917 altered the relationship between
support for the czar and attainment of cushy occu- Microfoundational Mechanisms  Relatedly, expan-
pational positions, taking it from positive to nega- sive theories will identify the specific microfoun-
tive. In this sense, such shifts often reflect regime dational mechanisms underlying the XY relation.
changes; thus, we term this kind of change regime Inequality is produced and/or maintained by
change. We present both Panels 3 and 4 to convey humans acting consciously or unconsciously.
that deciding whether a regime has changed is not Expansive theories are not satisfied with simply
always straightforward, for it raises the ques- observing a correlation between X and Y, nor
tion—how much change in quantity can occur with simply substituting other terms for the value
before a change in quality pertains? labels of X. Expansive theories seek to explicitly
state the desires, beliefs, opportunities, and
The answer to that question must be specific actions (Hedström 2005) that coalesce to consti-
to the issue in question and the theories under tute the microfoundations upon and through
consideration. For example, a Marxist could which all social entities—institutions, norms,
claim that a regime change has occurred if the extraindividual structures—are ground, the
relationship between capitalist class origins and mechanisms through which they activate their
outcomes moves from above zero (positive) to complex, often nonlinear effects. The task is
below zero (negative).5 But, there is nothing tricky, because the theory must attend to the real

5 The Marxist might also say that the relationship will be
below zero for some specified time, then return to zero.

84 S. R. Lucas and V. Irwin

­motivations of real persons even as the theory determined by one’s parents (e.g., Jensen 1969;
itself constitutes an abstracted model of the pro- Herrnstein and Murray 1994). To complete the
cesses at issue. The difficulty of this task may circle, assortative mating, the tendency of mating
partly explain why the number of expansive theo- pairs to contain people of similar levels of educa-
ries is dwarfed by the number of narrow theories. tion (Kalmijn 2001; Schwartz and Mare 2005),
occupation (Kalmijn 1994), and earnings
4.3.3 Theories Expansive (Sweeney and Cancian 2004), reinforce genetics-­
and Narrow based ability differences by race and class
(Herrnstein and Murray 1994).
An expansive theory of inequality will explain
multiple outcomes, will explain those outcomes Such old-school views have not been informed
for multiple categorical systems, will explain stasis by more recent genetic research. Geneticists have
and change in the XY relation, and will identify the long seen DNA as the basic building block of life.
microfoundational mechanisms underlying both However, for DNA (a genotype-level phenome-
static and dynamic relations of interest. The fewer non) to matter in a living organism (a phenotype) it
of those features a theory has, the narrower it is. must be expressed. How DNA is expressed and
what determines its expression is a cutting edge
Certainly, narrow theories have their value. area of early twenty-first century research. Notably,
First, a narrow theory is more finely focused, eas- epigeneticists have found that determinants of gene
ing empirical assessment. Second, being more expression are directly affected by the environ-
focused, a narrow theory is likely to more closely ment. An important, crucial finding of this research
match empirical observation than will an expan- is that organisms pass not just the DNA, but the
sive theory. Third, narrow theories can be used as proclivity for expression to the next generation. Far
building blocks for more expansive theories. from deepening the determinism of DNA, this new
evidence explains the crucial importance of envi-
However, the focus of narrow theories means ronment while providing a more precise specifica-
that one requires many such theories to explain tion of the mechanisms underlying evolution.
broad phenomena such as inequality in educa-
tion. As education involves many outcomes, there What is meant by gene expression?
is insufficient space to survey the set of narrow Analogically, imagine one has one blueprint for a
theories applicable to important outcomes, much 3-bedroom house. One builds two houses in dif-
less do so for both race- and class-based inequal- ferent environments. One house is built on flat
ity. Consequently, our review attends only to terrain in an earthquake zone, while the other is
major expansive theories of inequality. We treat built on sloped terrain in a seismically stable
genetics/epigenetics, human capital theory, the zone. To express the 3-bedroom house blueprint
Wisconsin social-psychological model, creden- in the former environment one will have to bolt
tialism, structural Marxism, cultural capital the- the house to the foundation, while in the latter
ory, (what we label) incorporation theory, terrain one may have to sink stilts into the hill on
oppositional culture theory, relative risk aversion, which part of the house may rest. The blueprint,
and effectively maintained inequality. We begin by itself, is insufficient to determine the actual
with genetics/epigenetics. realization of the house in any environment. But
the differing elements of each realized house—
4.4 F rom Incoherent Genetics bolted foundation or stilts—are intrinsic elements
to Epigenetics without which the house would not be viable for
the length of its otherwise designed life. Similarly,
Old-style biogenetic theorists see educational DNA, by itself, does not fully determine the
attainment and achievement as driven by ability, actual realization of the living being in any envi-
see ability as driven by genes, and see genes as ronment. The blueprint analogy is clarifying in
that it shows that DNA is insufficient to describe
a particular living organism. Yet, the blueprint

4  Race, Class, and Theories of Inequality in the Sociology of Education 85

analogy is incomplete in that it misses an impor- can occur through those mechanisms. And,
tant implication—epigeneticists are finding that because epigenetically-informed genetic theories
humans, other mammals, and insects experience of education potentially address all outcomes, the
certain environments that, through identifiable theory promises to be general. But, to date, the
hormonal pathways, affect DNA expression, such research steps needed to realize the theory’s
that the resulting phenotypes are visible in mul- promise has not commenced for education.
tiple later generations even after the environment
changes (e.g., Lumey 1992). 4.5 Human Capital Theory

This epigenetics research means that the Human capital theory makes sense of race and
nature–nurture dichotomy at the center of the class inequality in education, the role of class in
effort to emphasize biological rather than social inequality in education, and the intergenerational
factors is even more unsustainable than critics transmission of inequality. The theory posits the
have usually maintained. Analysts have already following relations. First, adults’ ability and prior
established that the statistical separation of out- investment drive adults’ productivity (e.g., output
comes into that owing to genes and that owing to per unit of time, quality of product per unit of
environment is impossible because genes and inputs). Investment thus generates a later income
environment intertwine to produce observed out- stream. Although some versions of the theory
comes (e.g., Daniels et  al. 1997). New findings focus solely on education and material earnings,
from epigenetics go farther, suggesting that the the broader version Becker (1962) offers consid-
very expression of an organism’s DNA is affected ers multiple kinds of human capital investment
by environment, and thus the environment funda- (e.g., migration, health care) as well as both
mentally produces the way in which the very material and psychic income. The broader Becker
genetic code of the organism is translated into definition is the one we consider here.
material existence and, in this way, produces the
biological endowment of the progeny of that Human capital exists along a continuum
organism (e.g., Meaney 2010). Such research anchored at one point by general human capital
implies that the claim that genes set a limit on the and at the other by specific human capital. In the
power of social factors will finally be revealed to extreme general human capital raises persons’
have been as fundamentally mistaken as oppo- productivity in all firms, while at the other extreme
nents (e.g., Fischer et al. 1996) of that view have specific human capital raises persons’ productiv-
oft maintained. Indeed, it appears that social fac- ity in one firm, only. Reading provides an example
tors, including education, not only may nurture of a skill closer to the general human capital pole,
native ability, but they may cause the very while the Byzantine procedures for requesting a
“native” ability they later nurture. blackboard for a classroom at the University of
California-Berkeley provide an example of a skill
The old genetics literature made many asser- closer to the specific human capital pole, i.e., of
tions about education, often calling for the sad arguably absolutely no value outside the specific
but sober acceptance that nothing could be done campus. Firms are unlikely to pay for general
in the face of the alleged overwhelming power of human capital acquisition (e.g., literacy) because
genetics. The literature on epigenetics has yet to if the person so-aided quits the job, some other
address inequality in education. But the evidence firm would recoup the returns to the first firm’s
on other issues suggests a much more hopeful investment. But, the closer the training is to the
posture is warranted. Indeed, such evidence sug- specific (i.e., firm-specific) pole, the fewer firms
gests that a society’s level of cognitive perfor- can gain from the investment, and thus the more
mance, as well as inequality in that performance, likely a firm will pay at least some part of the cost
is a direct function of the society’s tolerance for of the human capital investment. Thus, in the face
substandard and unequal environments. The the- of temporary downturns in firm performance,
ory identifies a key mechanism, hormonal path-
ways involving gene expression, and how change

86 S. R. Lucas and V. Irwin

firms are less likely to temporarily lay-off those thereby reduce their later productivity. This chal-
with specific human capital, because once the lenge becomes an intergenerational one in that
downturn ends the firm might be unable to rehire children’s credit constraint or lack thereof is a
the laid-off workers, for many may have found downstream implication of the resource limita-
other employment, thereby forcing the firm to pay tions or non-limitations of their parents (Tomes
to assess and hire new employees and then bring 1981; Becker and Tomes 1986). Becker and
new hires up to the same level of specific human Tomes (1986) show that only children of wealthy
capital attainment the laid-off workers had for- parents do not face credit constraints; children of
merly reached. Instead, firms are likely to lay-off middle-income and poor parents do face credit
those with general human capital. One way that constraints that hinder their ability to make opti-
these relations explain the positive association mal human capital investments. In this way
between education and employment is that spe- human capital theory suggests and explains a
cific human capital typically builds on general high association between parent and child educa-
human capital, such that those with specific human tional attainment. Indeed, as ability is a realized
capital typically have higher overall education. phenomenon partly produced by early childhood
socialization, part of the inequality generated by
Human capital resembles other investments in differences in ability are also arguably produced
that the longer persons have to accrue income through family differences in human capital,
from the investment, the more likely they are to such that even the ability pathway is partly a
make the investment. To make an investment the function of inequality in human capital.
investor must have resources sufficient to pay the
costs of the investment. The costs are both direct Human capital theory offers many ways to
(e.g., tuition) and indirect (e.g., time). The latter explain racial/ethnic inequality in education.
is interesting in reference to human capital First, if racial/ethnic groups differ in wealth,
because in order to make the investment the credit constraints may produce lower investment
investor must spend the time in the activities that for members of poorer racial/ethnic groups inde-
embody the investment, and thus must forego any pendent of their ability. Second, if members of a
gains that would accrue to spending time in some racial/ethnic group are more likely to doubt
other activity. The theory phrases this claim in access to the occupational positions that would
terms of foregone income; the classic example is allow them to reap the returns of additional
that in order to attend school full-time a college investment, perhaps owing to current or historic
student must forego the earnings they would have discrimination (Loury 1992), then the average
obtained had they taken a paying full-time job. human capital investment of members of that
The foregone earnings are added to the cost of racial/ethnic group would be expected to be
tuition and fees to produce the total cost of col- lower than that for others. Third, if different
lege attendance. Notably, the above explains why racial/ethnic groups have different health profiles
younger persons are more likely to invest in edu- and life expectancies, members of groups with
cation, for older workers have average higher worse health and/or shorter life expectancies
earnings than younger workers and thus foregone should be expected to invest less in education
earnings costs are lower for younger persons. because they will have less time to accrue the
benefits of that education.
Human capital theory contends that if persons
lack money or credit (i.e., loans) to enable them This third pathway may seem odd to some
to pay the direct and indirect (i.e., opportunity) who doubt that children look into the future, see
costs of an investment, they may fail to make dim life expectancy prospects, and then reduce
investments they otherwise might make. In this their investment in education. But such a criti-
way human capital theory has direct implications cism caricatures the human capital logic while
for class inequality. First, and most notably, per- ignoring the literature on children’s decision-­
sons with insufficient resources face financial (or making. Recall that human capital investment
credit) constraints that prevent investment and imposes opportunity costs in the form of other

4  Race, Class, and Theories of Inequality in the Sociology of Education 87

activities in which one cannot engage while mak- 4.6 W isconsin Social-­
ing the investment. Those opportunity costs could Psychological Model
entail foregone leisure. Seen in this way, a key
reason to forego a benefit in the short term is to The Wisconsin Social-Psychological Model of
obtain a larger benefit in the long-term. Given Status Attainment (aka the Wisconsin model)
that some communities may have higher than addresses race and class inequality in educational
average doubt there will be sufficient time to attainment, placing a social-psychological factor
obtain later long-term benefits (owing, perhaps, at the center of the process of educational attain-
to long-running poor access to or experience with ment, occupational success, and earnings (e.g.,
the health care system (e.g., Jones 1981; McBean Sewell and Hauser 1980; Hauser et al. 1983). The
and Gornick 1994)), the theory suggests that peo- key factor in the Wisconsin model is significant
ple in those communities will invest less in others’ influence, for the theory asserts that a pri-
human capital, on average. mary conduit of social background factors’ (e.g.,
parents’ earnings) causal effect on later outcomes
Intriguingly, the empirical evidence is consis- works through this chokepoint.
tent with this third pathway. Research indicates
that not only are adolescents who doubt they will Figure 4.5 reveals the structure of the claims
live to age 35 more likely to begin selling drugs, at the conceptual level. Both academic perfor-
but also, the higher the proportion of schoolmates mance and family socioeconomic position—
who doubt reaching age 35, the more likely the measured by parents’ education, father’s
adolescent is to begin selling drugs (Harris et al. occupation, and family income—cause signifi-
2002). These findings are consistent with the cant others’ influence, which is measured via stu-
third pathway above. dents’ report of their parents’ and teachers’
encouragement for college and peers’ plans for
The clear generality of human capital theory college. Significant others provide the main con-
does not imply only as grim conclusions as the duit through which social background has its
above empirical relations may suggest, for the effects on adult outcomes, and the effect runs
theory contains the possibility of change. If through children’s educational aspiration, occu-
investment returns and/or financial constraints pational aspiration, and educational attainment.
change, inequality will likely change, too. With
respect to the role of race and class inequality in Class inequality in producing educational
education outcomes, changing the financial con- attainment is referenced in the models’ relating
straints to investment can alter the role of race parent status characteristics to the encouragement
and class in educational attainment and achieve- of parents, teachers, and peers. But the relation
ment. And, with respect to the role of education can be explained in one of two ways. One view
in producing class inequality, changing the claims the theory asserts that socioeconomically
returns to education can, by definition, alter the advantaged parents socialize their children to suc-
role of education in class inequality. However, ceed in school and this leads teachers and peers to
the direction of any change in either case depends encourage those children to seek higher levels of
on implementation and other factors beyond (but education and occupational success (Kerckhoff
perhaps related to) human capital theory. For 1976). An alternative view claims that teachers
example, whether reducing financial constraints respond more positively to socioeconomically
on early childhood education will raise or lower advantaged students and that parents select socio-
race and/or class inequality may depend on the economically advantaged contexts (e.g., neigh-
means by which the financial constraints are borhoods) such that their children’s peers will
reduced, how widespread the reduction is, and also be encouraging in a matter-of-fact manner. In
how childcare and education providers respond such neighborhoods it is as obvious that college
to the reduction. entry follows high school completion as it is that

88 S. R. Lucas and V. Irwin

Fig. 4.5  Wisconsin model, trimmed structural version. (Adapted from Table 1, Hauser et al. 1983)

February follows January—for children with 4.7 C redentialism
such peers, both “truths” are so true that comment
on their truth is almost non-­existent. The theory, Credential theory comes in two variants. One
thus, identifies social-p­ sychological connections perspective, which we term the non-linear effects
that link parental sociodemographic characteris- version, simply highlights the empirical evidence
tics to children’s educational and occupational that the earning gains are boosted for obtaining a
expectations and outcomes. But the explanatory credential over and above the gain persons accrue
basis of the linkage remains under study. owing to the completion of an additional year of
schooling. At major credential-completion years,
With respect to race, a key question the theory such as college graduation (e.g., Goodman 1979;
poses is whether the process works the same for Grubb 1992, 2002), analysts have observed such
different racial groups—where to work the non-linearities.
“same” is reasonably interpreted as structural
coefficients being equal across groups. The evi- Collins (1974, 1977, 1979) offers what we
dence of whether the process works the same term a monopolization process version, which is
across races is unclear, however. Some research a more complex version of the theory that sub-
finds similarity (e.g., Wolfle 1985); some does sumes the possible non-linear effects of creden-
not (e.g., Kerckhoff and Campbell 1977); and tials into a wider discussion of the genesis of
some claims the highly variable statistical meth- specific credentials as markers of earnings-­
ods, sample designs, and populations studied enhancement. Collins (1979) argues that creden-
undermine any general answer to the question tials are the result of and resource for a joint,
(e.g., Gottfredson 1981), a conclusion that unfor- complex process of ethnic status competition and
tunately has not changed in the intervening occupational professionalization.
decades (e.g., Morgan 2004). What can be noted
is that the Wisconsin model provides an encom- It is well-known that members of a field that
passing perspective within which one may assess successfully secures the designation “profes-
racial inequality, socioeconomic inequality, and sional” obtain earnings and other advantages
other sociodemographic grounds for inequality (Klegon 1978). One mechanism that can increase
(e.g., gender). earnings is professionals’ control of certification
to practice the profession, as professions

4  Race, Class, and Theories of Inequality in the Sociology of Education 89

­generally obtain largely independent control of tional niches, it focuses on both race/class effects
certification (Greenwood 1957) on the argument on education and later education effects on class.
that only they, guided by a code of ethics, have Because monopolizers can extract rents (Sørensen
sufficient expertise to evaluate competence and 2000)—payment over and above the level of pro-
recognize appropriate conduct of the discipline ductivity—and non-monopolizers cannot, cre-
(Mitchell and Kerchner 1983). dential theory implies an increase in inequality
along lines of race and class. Notably, by linking
In a context of ethnic competition, in which processes assigning earnings to occupations (e.g.,
ethnic groups attempt to dominate particular firms’ reward structures), prerequisites (e.g., edu-
occupational niches, the resources of profession- cation credentials) to positions (e.g., jobs), and
alization are quite useful. The ability of profes- racial/ethnic closure, this more complex version
sions to certify practitioners facilitates reducing of credentialism theory becomes potentially rel-
competition between co-ethnic peers, just as the evant for the intergenerational transmission of
same resource facilitates reducing competition inequality.
between professional colleagues. Notably, con-
trolling the certification process facilitates main- 4.8 Structural Marxism
taining scarcity as well as barring persons whose
sociodemographic category will lower the status In Schooling in Capitalist America, Bowles and
of the profession. Maintaining scarcity and the Gintis (1976) investigate the function of educa-
social status of practitioners can help erect a floor tion in social reproduction. They argue that, rather
beneath earnings for the profession. than developing cognitive skills that foster meri-
tocratic social mobility, the primary function of
Schools enter this process as a cite for certifi- the school is to prepare students for work in (their
cation, but schools are not independent because ascribed status in) the capitalist labor market.
for a field designated as a profession the faculty They support this argument in three ways. First,
involved in teaching the material will themselves although cognitive skills are important in the
tend to be certified practitioners. Consequently, labor market, they show that this only partly
professions and would-be-professions turn to the explains the advantage attributed to more years of
school—first the high school, then the colleges, education, with personality traits signaling con-
and later (perhaps) post-graduate institutions—to formity having notable additional effects (Bowles
certify at least some stages of the training deemed and Gintis 1976, pp. 137–139). Second, children
necessary. This position becomes clearer upon reproduce their parents’ socioeconomic status at
noting that the placement of occupational train- rates that could not be fully explained by either
ing inside schools is a historically recent phe- their inherited cognitive advantage or by the elite
nomenon (Benavot 1983, p. 64; Jacoby 1991). educational opportunities they are afforded.
Finally, the authors argue that historically in the
This variant of credentialing theory identifies United States, periods of school reform have
the role of signaling amongst firms as key to tracked periods of change in the structure of labor.
explaining why firms make college (for example)
a prerequisite even for jobs whose tasks (e.g., fil- Based on these patterns, Bowles and Gintis
ing, keyboarding, simple mathematics) do not argue that education prepares students for the
require college training. Basically, firms signal stratified labor market through what they call the
their quality to important others (e.g., clients, correspondence principle. The correspondence
regulators) by requiring high levels of education principle refers to the parallel between the social
for even many rudimentary jobs. relations of labor and the social relations of edu-
cation. In the capitalist context the correspon-
The stark nonlinear effects version of creden- dence principle implies that schools inure
tialism theory is more directly focused on how students to the types of hierarchical relationships
education affects class (e.g., earnings, wealth).
But, because the broader monopolization process
variant highlights class- and ethnic-based efforts
to erect barriers to entry and monopolize occupa-

90 S. R. Lucas and V. Irwin

that are characteristic of corporations. Rather ideal of education persists, the failure of members
than cooperation, students are encouraged to of disadvantaged groups to achieve social mobility
compete—or, more accurately, made to believe is understood to result from their own failures.
they are engaged in meritocratic competition—
for the few spots at the top, and only those who The structural Marxist theory of class inequal-
secure these school positions are given the tools ity in education, particularly as exemplified by
for autonomy and advanced critical thinking Bowles and Gintis, differs importantly from some
reserved for the capitalist elite. Rather than fos- theories in that the reproduction mechanism it
tering an actual meritocracy, schools reinforce proposes is institutional rather than individual. It
students’ place in the educational hierarchy is not the students’ resources or aspirations that
beginning at a very young age and, by “corre- primarily drive inequality, but rather how the
spondence,” cultivate the impression that workers stratified school system shapes and realizes them.
arrive in the only position in the hierarchy of pro- Yet, while structural Marxism is generally inter-
duction for which they are inherently qualified. preted as one of rigid reproduction, with schools
populated by passive, non-­agentic students (e.g.,
Melvin Kohn and colleagues (e.g., Kohn and Giroux 1981; McNeil 1981), the theory actually
Schooler 1969) highlight a similar correspondence relies on individual variation and student action.
between men’s occupation and the values they It is the few working-class kids who succeed in
hold for their children, such that upper-class men attaining middle-class positions, after working
value self-direction, a useful orientation in jobs hard in school of course, who are truly indispens-
that, within circumscribed limits, require creativ- able to the perception of a meritocratic competi-
ity. In contrast, working-class men value confor- tion, a perception that is necessary to maintain
mity and rule-following, an essential orientation capitalism. However, because the mechanism is
given the much more constraining coercion of the at the institutional level, altering this mechanism
shop floor. Kohn implicates education in the for- (the correspondence between the social relations
mation and maintenance of these values insofar as of education and the social relations of labor)
it provides the space for intellectual flexibility for could potentially change not only the distribution
some students and fails to provide it for others, of outcomes and thus inequality, but also the rela-
foreshadowing Bowles and Gintis’ correspon- tionship between origin and destination class.
dence principle. Put together, these theories sug- The theory is therefore dynamic. Finally, the the-
gest that working-class students are not only less ory is general because, as we see with Kohn, the
likely to be given the opportunity in school to concept of “correspondence” can be applied to
engage and enhance their critical and creative institutions beyond the school.
thinking skills, but they are also less likely to have
parents who emphasize the fostering of critical and 4.9 C ultural Capital Theory
creative orientations as the purpose of education.
In Reproduction in Education, Society, and
The correspondence principle offers a grim Culture, Pierre Bourdieu and Jean-Claude
perspective on the role of education in the poten- Passeron (1977) explain inequality, among other
tial for social mobility of lower-income and minor- phenomena, by contending that schools reward
ity students. By beginning from disadvantaged behavior that complies with the norms and stan-
positions, these students are nearly guaranteed to dards of the dominant group in a society.
be placed low in the initial educational hierarchy Inequality follows because, try as they might,
and, if the correspondence principle holds, are outsiders cannot fully adopt the norms and stan-
unlikely to be given the tools to struggle their way dards of the dominant group because one’s core,
out of this position. Moreover, once in the labor one’s habitus, develops in the family, is impossi-
force, Kohn argues that the stratification of job- ble to change, and directly affects one’s behavior
relevant skills and behaviors cements the corre- despite one’s efforts. Consequently, one’s
spondence between education and class-s­pecific
values. Not only this, but because the meritocratic

4  Race, Class, and Theories of Inequality in the Sociology of Education 91

­likelihood of educational success is constrained ment is capital, in both material and symbolic
by one’s earliest formative experiences, sedi- forms. The theory is dynamic, but its conclusion
mented into one’s habitus. is that, alas, plus ça change, plus c’est la même
chose.
Bourdieu (1986) describes cultural capital—
of which habitus is one type—as a resource one 4.10 Incorporation Theory
may use to navigate various fields. Success in the
schooling process and the many labor markets Ogbu (1987) articulates a theory of immigrant
depends on one’s deployment of cultural capital incorporation. He maintains that the posture
in such fields. One does not deploy cultural capi- native-born minority students strike with respect
tal in a neutral arena because there are no neutral to school depends upon the predominant histori-
arenas, for all arenas have differing mixtures of cal pattern of incorporation of their racial/ethnic
material and symbolic criteria for success and group. Ogbu conceives of minority incorporation
any criterion inescapably advantages some and as either voluntary or involuntary. Voluntary
disadvantages others. Yet, Bourdieu highlights minorities are those who have entered the U.S.
gatekeeper exclusion on the basis of arbitrarily primarily through immigration. The theory sug-
selected criteria of evaluation that advantage the gests that voluntary minorities continue to view
previously advantaged. their opportunity structure in relation to that of
peers in their ancestral country. Further, volun-
Some readings of Bourdieu assert that mark- tary minorities can explain difficulties, inequali-
ers and mechanisms of success are selected ties, and poor treatment by their lack of knowledge
because of their ability to legitimate social clo- of their newfound land. Thus, they view the
sure for the advantaged (e.g., Lareau and returns to education favorably even though they
Weininger 2003). In this view, much that schools’ may be lower than for natives, because voluntary
value has no intrinsic utility, but rather serves to immigrants anticipate better returns for later gen-
distinguish (upper-) middle-class children from erations. With this posture, voluntary minorities
their lower-class peers. Others see exclusion via a engage school in ways that can facilitate success-
symbolic as opposed to material dimension as the ful performance.
key theoretical contribution of the concept of cul-
tural capital (Lamont and Lareau 1988), regard- In contrast, involuntary minority groups are
less of how the symbols are selected. those who “were originally brought into United
States society involuntarily through slavery, con-
If the content and character of childhood quest, or colonization” (Ogbu 1987, p.  321,
socialization depend on parents’ cultural reper- emphasis in original). Native Americans, Native
toire, and cultural repertoires are associated with Hawaiians, and African Americans are primary
class location and race/ethnicity, then childrens’ examples in the United States. The phenomenon
developing habitus will differ by class and race. is not confined to the United States, as many
Consequently, cultural capital theory implies that examples exist, including the Burakumin in
intergenerational transmission of socioeconomic Japan, the Maori in New Zealand (Ogbu 1987,
and racial inequality occurs partly through the p.  321), travelers in Eastern Europe, and more
intergenerational transmission of culturally dis- (Fischer et al. 1996, p. 192, Table 8.1). Involuntary
tinct repertoires along lines of race and class that minorities and their children cannot explain dif-
do not match socially-constructed definitions of ficulties, inequalities, and poor treatment by lack
merit. Further, intragenerational inequality—the of knowledge of their homeland. Historical
association between early and later placements of enslavement, conquest, or colonization echoes in
a person in various educational and/or occupa- contemporary poor treatment, creating a clanging
tional positions—is explained by virtue of inconsistency with any expectation of fair returns
habitus. now or better returns for later generations. This

Cultural capital theory attempts to be nothing
short of a complete theory of attainment, and thus
is extremely general. The mechanism of attain-

92 S. R. Lucas and V. Irwin

history of unfairness makes education a poor tion, it is at least as important to keep the ear’oles
investment. in mind as we consider race and class inequalities
in education. Their existence raises important
Some analysts point to an “immigrant para- questions about whether peer subcultures offer an
dox,” in which children of some immigrant adequate means of explaining variation in the
groups attain higher levels of education than their correspondence between school and work.
native-born peers on average, an advantage that
tends to dissipate or even reverse by the third Although Willis’s theory is based on class—
generation (Rumbaut 1999; Perreira et al. 2006). and the White male subculture he describes is
The “immigrant paradox” basically compares propped up by rampant racism and sexism—the
better than expected performance of the first and most famous school subculture theory, opposi-
second generation with worse than expected per- tional culture, aims instead to explain racial
formance for later generations. Evidence sug- inequality in education. From this theory, the
gests the “paradox” may be explained by “burden of acting White” hypothesis (Fordham
considering the educational context of immigrant-­ and Ogbu 1986; Ogbu 2003) states that Black
sending countries (e.g., Feliciano and Lanuza students view academic achievement as a
2017). But even if the paradox were to hold, it “White” enterprise and therefore resist this path
suggests that incorporation into a society where so as not to be labeled a traitor to their race.
racial stereotypes and White advantage are perva- According to this theory, minority students per-
sive may produce sustained disadvantage relative ceive that their efforts and achievement in school
to native-born Whites, unravelling initial volun- will result in fewer career opportunities than that
tary immigrant optimism and fostering disen- same effort or achievement would produce for
gagement among some immigrant groups. White students. As a result, involuntary minority
students, particularly Blacks, demonstrate resis-
According to incorporation theory minorities’ tance to school and negatively sanction their
initial reception is critical, as history cannot be high-performing co-ethnic peers. Ogbu hypothe-
re-run. Thus, incorporation theory implies strong sizes that it is this racialized rejection of educa-
inertia in the inequality between groups. By tion that best accounts for the persistence of the
explicitly theorizing stasis even as conditions achievement gap between Black and White
may change, their theory satisfies our criteria for students.
dynamic theories of inequality.
However, Fordham and Ogbu’s (1986) origi-
4.11 Oppositional Culture nal research that proposed the theory used a poor
sample design (Lucas 2016) that prohibited the
In Learning to Labour, Willis (1977) studies “the drawing of any conclusions beyond the specific
lads,” a White, male working-class peer group at students studied, while at the same time conflat-
a single school in England. Resigned to their fate ing labels such as “brainiac” with Whiteness.
as manual laborers, in a town where there are vir- Similarly, the premise that involuntary minority
tually no available alternatives, these young men students (Ogbu 1987) reject education or view
develop a hypermasculine counter-school ethos achievement as White has been largely discred-
that values common sense over book knowledge ited (e.g., Ainsworth-Darnell and Downey 1998;
and measures worth through physical and sexual Downey et  al. 2009; Harris 2006). Other work,
prowess. Yet, Willis also studies the “ear’oles” including Willis’s, also clearly demonstrates that
who, despite sharing job prospects similar to the disengagement from schooling is not exclusively
lads, uphold the meritocratic ideal of education. a minority phenomenon (Willis 1977; MacLeod
Although it is the “lads” who are typically con- 1987; Tyson et al. 2005). Yet, the legacy of under-
sidered the noteworthy case because they reject standing some students’ underperformance in
school authorities’ orientation towards educa- terms of a conflict between their racial/ethnic
identity and dominant cultural values endures.

4  Race, Class, and Theories of Inequality in the Sociology of Education 93

Notably, Prudence Carter (2005) finds that stu- The theory is also dynamic because if schools

dents do not interpret academic success as a were to change their reward structure to value

White trait, but identifies the importance of “kee- students’ adaptability (an arguably important life

pin’ it real,” or being authentic, to students’ eval- skill), then Carter’s typology could accommodate

uations of their peers (Carter 2003, 2005, 2006). a different pattern of inequality (e.g., where only

Carter does not suggest that students are never the ability to “straddle,” not dominant competen-

negatively sanctioned by their peers for “acting cies alone, would predict greater school

White,” but rather that this epithet was used on success).

students regarded as snobs, not on students

regarded as pursuing academic excellence. Thus,

the epithet’s use is distinct from students’ opin- 4.12 R elative Risk Aversion

ions about the institution of education, which she

finds to be uniformly positive among her sample Relative Risk Aversion (RRA) is offered by

of Black and Latino/a adolescents in Yonkers, Breen and Goldthorpe (1997) to contest cultural

New  York. Rather, educational achievement is theories of inequality while explaining stable

associated with their ability or willingness to class differentials across cohorts, declining class

enact the behaviors and competencies valued by effects across education transitions, and rapidly

the school. Students who straddled school (i.e., changing gender effects. RRA accepts that edu-

dominant) and nonschool (i.e., non-dominant) cational opportunities require both financial and

competencies were the most socially successful cognitive resources. Conditional on those con-

and also performed well academically. Flores-­ straints, RRA posits that students (and families)

Gonzàlez (2002) similarly finds that the ability to make decisions based on students’ understanding

maintain and meld diverse identities is also key to of their likelihood of success were they to follow

persistence in high school in her sample of Puerto specific educational paths and their estimation of

Rican adolescents. the probability of attaining sought occupational

While Carter does identify a group of students positions via those paths. The core of the theory

who behave in a manner that echoes Ogbu’s rests on three key theorems: (1) Adolescents seek

“opposition”—using “Black English Vernacular,” to avoid downward socioeconomic mobility, (2)

putting forth minimal effort in school, and dem- each educational path entails some risk that stu-

onstrating high ethnic-centrality—and the hege- dents will seek to avoid if possible, and (3) cul-

monic masculinity of “the lads,” she finds that tural differences are not necessary to explain

these students regard education as important and inequality (Breen and Goldthorpe 1997, p. 238).

do not view achievement as White. Rather, the With respect to the first theorem, assume the

seemingly oppositional cultural codes employed socioeconomic distribution is divided into

by many minority youths were simply intended thirds—top, middle, and underclass. Those hail-

“to create a coherent, positive self-image (or set ing from the middle can avoid downward mobil-

of images) in the face of hardship or subjugation” ity by obtaining middle or top occupations, but

(Carter 2005, p. 57). Thus, although student sub- those at the top can only avoid downward mobil-

cultures arguably exist, evidence does not sup- ity by reaching a top occupational destination.

port the notion that noncompliance is synonymous The theory states that this difference produces

with rejection of education. Carter identifies stu- different incentives for the level and kind of edu-

dents’ ability to negotiate competing sets of val- cational attainment pursued.

ues as the operative mechanism in social and With respect to the second theorem, the theory

academic school success. Understood this way, posits that paths that entail demanding educa-

the theory is general—not only can it be applied tional opportunities are great for those who suc-

to different minority groups, but the reward struc- ceed, but those who follow that path yet fail will

ture of the school has also been shown to conflict encounter worse outcomes than they would have

with class-identity expression (e.g., Willis 1977). encountered had they succeeded in a less demand-

94 S. R. Lucas and V. Irwin

ing curriculum path. This assumption is the ering only one dimension may lead analysts to pre-

source of the theory’s name, relative risk aver- sume a decline in inequality when, in actuality, for

sion; specifying costs to failure makes it possible example, all that has happened is that the locus of

for some students to expect to do better by taking consequential inequality shifted from the quantita-

less than the most demanding curriculum avail- tive to the qualitative dimension.

able. Thus, such students will engage as if risk EMI has been applied to education almost

averse. exclusively (e.g., Esping-Anderson and Wagner

With respect to the third theorem, their rejec- 2012). Further, most applications focus on only

tion of the subcultural thesis, Breen and one aspect of the theory, its assertion that all

Goldthorpe (1997) posit a society-wide consen- goods have both qualitative and quantitative

sus that certain educational pathways are more dimensions, to highlight inequality in qualitative

likely to lead to occupational success. Although dimensions of education.

students’ assessment of their likelihood of educa- Applying this general theory of inequality to

tional success will depend in part on what they education, EMI explained socioeconomic effects

see as their ability, it will not depend on sub-­ on education in one of at least two ways. When

cultural values, norms, or behaviors. some attain a particular level of schooling

The theory, thus, explains class and race whereas many others do not (e.g., high school

inequality in education with the same mecha- completion throughout the first half of the twen-

nism—socioeconomically disadvantaged students tieth century in the United States), the socioeco-

and students from racially and/or ethnically dis- nomically advantaged use their advantages to

empowered communities are likely to have par- secure that level of schooling. However, if that

ents with lower occupational attainments. Children level of schooling becomes widely or perhaps

whose parents have lower occupational attain- even universally attained, the socioeconomically

ments have a lower floor their own educational advantaged seek out whatever qualitative differ-

attainments must reach to avoid downward mobil- ences there are at that level, using their advan-

ity. Although the theory posits lower cognitive tages to secure quantitatively similar but

ability for students from poor (and racially disem- qualitatively better education (e.g., qualitatively

powered) families, the difference in floors for suc- better, more challenging curricular tracks). Thus,

cess is sufficient to create educational inequality. EMI notes that actors’ foci may shift as qualita-

tive differences supplant quantitative differences

4.13 Effectively Maintained in importance. Alternatively, actors may refer-
Inequality ence qualitative differences even when quantita-
tive differences are common. Either way, EMI

claims that the socioeconomically advantaged

Lucas (2001) proposes Effectively Maintained will use their advantages to secure both quantita-

Inequality (EMI), a general theory of inequality. tively and qualitatively better outcomes.

EMI claims that socioeconomically advantaged Aspects beyond the qualitative/quantitative

actors secure for themselves and their children distinction have not received much attention,

advantage wherever advantages are commonly even though they are constitutive aspects of

possible. The theory further contends that all goods EMI. The theory articulated its decomposition of

have both qualitative and quantitative dimensions. goods into qualitative and quantitative dimen-

This multi-dimensional nature of goods facilitates sions while also identifying an important role for

the intransigence of inequality, for the theory (student) myopia [aka nearsightedness], inequal-

claims that if quantitative differences are common, ity (amongst students) in access to information

the socioeconomically advantaged obtain quanti- that could dispel the myopia, the discretionary

tative advantage. But, if qualitative differences are power of (school personnel) gatekeepers, and the

common, the socioeconomically advantaged possibility of class-based (parental) collective

obtain qualitative advantage. If this is true, consid- action to maintain advantage. School-related

4  Race, Class, and Theories of Inequality in the Sociology of Education 95

Low SES Predicted Values as Coefficient Changes

Predicted Values
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1

-2 -1 0 1 2
Socioeconomic Status Coefficient

Low Outcome Lo-Mid Outcome

Hi-Mid Outcome High Outcome

Fig. 4.6  Low SES predicted values as socioeconomic background coefficient changes

labels are placed in parentheses because they results plotted in Figs. 4.6 and 4.7, Fig. 4.8 sketches
translate the general theoretical postulates into the range of coefficients that satisfy EMI.
the realm of education.
Most theories of inequality would be satisfied if
One important feature of EMI is illustrated the coefficient on social background is positive.
across Figs. 4.6, 4.7, and 4.8. To test for the quali- EMI, however, has a more constrained prediction,
tative hypothesis of EMI, one must use a categori- for it asserts that myopia, differential information
cal dependent variable (e.g., dropout, no academic to dispel myopia, gatekeeper discretion, and class-
course, academic low-track course, academic based collective action all work to keep the social
high-track course) and calculate and compare pre- background coefficient within a smaller band of
dicted outcome category probabilities for those of values. EMI implies that efforts to move the coef-
low and high socioeconomic background. ficient outside of that band will encounter serious
Figures  4.6 and 4.7, for low and high socioeco- resistance (Lucas 2017). Thus, for EMI, most pos-
nomic background students respectively, trace the itive coefficients would be inconsistent with EMI,
predicted probability of entering each of four cat- making it possible for the association between the
egories of an outcome variable as the socioeco- outcome and socioeconomic background to be
nomic background coefficient changes.6 EMI is statistically significant but still not support EMI
supported if the category with the highest pre- (Lucas 2009), rendering EMI falsifiable even
dicted probability differs for those of high and low amidst ubiquitous findings showing a positive
socioeconomic background. Intriguingly, this association between socioeconomic background
means that EMI implies bounds on the socioeco- and education outcomes. Or, in other words, EMI
nomic background coefficient, for only some coef- theory identifies the thresholds at which a society
ficients make the predicted outcome category for shifts from an Effectively Maintained Inequality
those of high socioeconomic background exceed regime to something else.
the predicted outcome category for those of low
socioeconomic background. Given the illustrative The theory specifically addresses change
within an EMI regime by denying its consequen-
6 Three thresholds divide the four categories: −2, 0, and 2. tiality. In a sense, EMI posits a basic cause à la
Lieberson (1985, pp.  185–195)—the aim of
advantaged actors to maintain their advantage.

96 S. R. Lucas and V. Irwin

High SES Predicted Values as Coefficient Changes

Predicted Values
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1

-2 -1 0 1 2
Socioeconomic Status Coefficient

Low Outcome Lo-Mid Outcome

Hi-Mid Outcome High Outcome

Fig. 4.7  High SES predicted values as socioeconomic background coefficient changes

Fig. 4.8  Coefficient values that produce EMI pattern

That cause creates (and thus explains) a diverging professional school, vocational training program,
trajectories pattern such that children of socio- R1 research university, and more. Amidst this
economic advantage transition into occupations plethora of possibilities, the basic cause remains
and earnings niches of socioeconomic advantage operative—advantaged people secure for them-
while their poor peers tend to make other transi- selves advantage wherever advantage is (com-
tions. However, the process by which these tran- monly) possible.
sitions are produced change over time; the stable
pattern exists amidst a plethora of superficial Despite its doubt about overall societal
causes/pathways through which the basic cause change, EMI posits that some individuals will be
maintains consistent force. In the sphere of edu- able to follow more advantaged trajectories than
cation, the superficial causes include the various their disadvantaged origins might suggest. The
levels and kinds of education—high school grad- theory claims that our predictions for disadvan-
uation, Advanced Placement courses, honors, taged students, however, will diverge from those
International Baccalaureate, 4-year college, we make for advantaged students, even after we
small liberal arts college, community college, control for academic achievement. Such patterns

4  Race, Class, and Theories of Inequality in the Sociology of Education 97

reflect the intransigence of inequality and its the mechanisms they highlight can escalate and
intergenerational transmission. rigidify.

4.14 Points of Contact Human capital theory highlights persons’ deci-
Between and Challenges sions to invest (in education), accepting such deci-
of Expansive Theories sions occur under constraint. Both RRA and EMI
also prioritize persons’ decisions to invest under
4.14.1 S elected Points of Contact constraint—RRA with unequal cost constraints,
Across the Theories EMI with unequal information constraints and
unequal discretionary gatekeeper support.
Expansive theories might be arrayed as if each
offers an entirely separable understanding of the The Wisconsin model’s emphasis on encour-
phenomena at issue. Yet, these theories work the agement by others resonates with the social-­
same intellectual terrain, so it should come as no psychological aspects of incorporation theory,
surprise that they connect and reinforce each which can be seen as generalizing the set of sig-
other at some points. To correct the possible ten- nificant others, with oppositional culture, which
dency of seeing each theory in isolation, we note suggests that peer evaluations are an import fac-
a few points of contact across the theories. tor in students’ attitudes toward and behavior in
school, and with RRA, which implies a social-­
First, epigenetics can be interpreted as sug- psychological process through its assertion of a
gesting that educational success partly flows role for students’ assessment of their likelihood
from a genetic basis, but a key part of that basis is of success along various paths.
etched through environmental pathways. That is,
the provision of encouraging environments can Credentialism, in referencing the qualitative
create hormonal responses that coax gene expres- category of professional, highlights ethnic com-
sions conducive to better cognitive performance. petition and professionalization as a resource for
Seen in this way, epigenetics implies an impor- exclusion, in affinity with structural Marxism’s
tant role for encouraging environments, at the recognition of elites’ monopolization of well-­
molecular level and above. In a way, epigenetics remunerated positions, cultural capital theory’s
deepens the importance of the environment, for notice of elites’ erection of arbitrary barriers to
environmental effects are insinuated into the their advantage, and EMI’s reference to a qualita-
organism in a constitutive way. Epigenetics thus tive dimension and class-based collective action
deepens the implications of the Wisconsin model, in the allocation of advantaged positions on that
with its emphasis on significant others’ (i.e., par- dimension.
ents’, teachers’, and peers’) encouragement,
structural Marxism, with its identification of eco- Structural Marxism, privileging distinctions
nomic and education structures that squelch between categorically differentiated economic
human potential, incorporation theory, with its positions and identifying stratified pathways to
distinction between immigrants facing hostile, those positions, resonates with incorporation the-
exclusionary or non-hostile inclusionary ory’s reference to legally-defined distinctions of
responses from natives, and EMI, with its empha- immigrant incorporation.
sis on gatekeeper ability to encourage (open) or
discourage (block) student access to environ- Cultural capital theory, with its emphasis on
ments that encourage increasing performance. translating capital from one field to another, is
Each of these theories identifies a mechanism consistent with incorporation theory’s under-
that may involve an undiscussed epigenetic path- standing of the differential valuation of immi-
way through which intergenerational effects of grants from different origin countries and with
oppositional culture’s understanding of differen-
tial cultural markets.

Finally, incorporation theory’s reference to the
differential reception of different immigrants not
only may provide the context within which

98 S. R. Lucas and V. Irwin

­oppositional cultures may arise and take root, but With epigenetics, one challenge is that geneti-
also may matter for EMI’s suggested differential cists have established that many complex tasks
discretionary response of gatekeepers (i.e., gate- require multiple genes acting in concert (Marsh
keepers may respond differently to voluntary and 1997). To discover a genetic connection for such
involuntary immigrants). a complex process as learning and/or education
seems a daunting task. Thus, at present, epi-
The listed points of contact do not exhaust the genetics is a tantalizingly promising theory, its
possible connections between the theories. But, possibility revealed more in our imaginations
they are enough to draw two conclusions. First, than in even the beginnings of research.
even disparate theories may not deny every aspect
of each other, suggesting that if high levels of Human capital theory would seem to require a
hostility are observed in scholars’ debates, those coherent understanding of productivity, but
emotions have more to do with the discussants empirical analysts usually simply assume or
than with the material for discussion. Perhaps assert that earnings track productivity (e.g.,
recognizing theories’ shared elements may Byrus and Stone 1984), a view falsified by
reduce the heat, and increase the light, that dia- decades of sociological research (e.g., Wright
logue can provide. and Perrone 1977; Kalleberg and Griffin 1980;
Spaeth 1985; Halaby and Weakliem 1993). Once
Second, because many theories share some one realizes the uncertainty plaguing the opera-
elements, adjudicating between theories can be tionalization of productivity, the theory’s mecha-
challenging, because shared elements—when nism is no longer clear and the theory’s elegance
confirmed—contribute to concluding in favor of is seriously endangered.
each theory that shares the element. Consequently,
one should expect adjudication to require intense The Wisconsin model foregrounds significant
study and to be difficult. Difficult though it is, others’ influence, making it the chokepoint of
adjudication is an important task. It is to the intergenerational status transmission. Teachers
important task of adjudication to which the pen- are key significant others, and teachers could
ultimate section turns. But first we must consider, encourage all students. If teachers encourage all
why adjudicate? Why not simply accept each students enough but in patterns that lead to the
theory singly, or see each as contributing one equalization of overall encouragement across
piece to our understanding of racial/ethnic and students, downstream outcomes should alter such
class inequality in education? that every child would have and reach high occu-
pational aspirations. Yet, occupational distribu-
4.14.2 C hallenges of the Theories tions are not only a function of young adult
demand for jobs, they also are a function of larger
It may be heartening to observe multiple points macroeconomic features (e.g., trade surpluses
of contact across theories, for their existence may and deficits) as well as employers’ supply of
suggest some degree of consensus, at least within occupational positions, such that it is unlikely
subsets of similar theories. If consensus is that every child, no matter how encouraged, will
­emerging, this may suggest that all is well with attain high status occupations and earnings. One
each theory, and the task now is to simply see response is to interpret the Wisconsin model as a
how they fit together. Alas, such an impression is static summary of relations for a cohort, but such
misleading. The collective points of contact are an interpretation undermines the view of the
important, but they exist alongside another set of model as reflecting a causal theory.
important observations: Although each theory
may appear internally consistent initially, closer Bourdieu has been viewed as identifying the
scrutiny reveals nagging issues with each. process by which oppression is constructed and
maintained by arbitrarily-selected criteria of
merit. Yet, because the theory offers no criteria

4  Race, Class, and Theories of Inequality in the Sociology of Education 99

for what is and is not or can be and cannot be cul- Incorporation theory implies that the condi-

tural capital, anything can be cultural capital, and tions under which immigrant groups entered the

all criteria are arbitrary. While this may make cul- country matter. But, research also shows that

tural capital theory seem to be incredibly broad, changing demographics and policy can greatly

the result is to leave only political grounds for con- reduce the impact of the history of incorporation

testing criteria of merit, i.e., the only way to con- (Lieberson 1980). This raises the question of

test a theory with integrity is to claim one is whether the apparent power of incorporation is

disadvantaged by the criteria. But, as someone real or, instead, epiphenomenal, apparent only

must always be disadvantaged (e.g., someone must because many (most?) groups’ treatment does

be last in line), any given person’s being in the set not change as their incorporation recedes into the

of disadvantaged persons on the basis of some cri- past (e.g., Cubans welcomed, Mexicans

terion is hardly good reason to change the criteria. vilified).

Indeed, even if criteria were to greatly change, the Oppositional culture is based in a claim that

new criteria would still be arbitrary, and thus as communities hold antagonistic views toward

susceptible to Bourdieusian critique as former cri- mainstream success. Yet, research shows late

teria. Thus, cultural capital theory is now and will twentieth-century minority elementary school

always be a critique of the status quo, no matter children seeking to succeed in school (e.g., Tyson

what that status quo is. If the theory cannot extri- 2002), and mid-twentieth-century mainstream

cate itself from this conclusion, it is revealed to be adolescents rejecting school (e.g., Coleman

tautological and thus, ultimately, unilluminating. 1961). Faced with such findings, the origin of

Credentialism is articulated in line with pro- students’ alleged opposition in communities pres-

fessional occupations, but very few credentials ents a serious puzzle for oppositional culture

are actually about traditional or powerful profes- theory for, if opposition does not originate in dis-

sions. It remains to be seen whether the theory’s enfranchised communities and only in disenfran-

social closure mechanism is truly class- and chised communities, how can it explain

racial/ethnic-specific, or even operational, once long-standing group-linked differences in

one broadens the understanding of credential to education?

include the burgeoning number of non-­ Relative risk aversion asserts the existence of

professional certificates so as to reflect the expe- a society-wide consensus as to which positions

rience of the bulk of any cohort. are better, but immigration and concomitant

Structural Marxism is often vilified for an increasing diversity makes the assertion less and

alleged lack of agency (e.g., Giroux 1981), but less secure. The assertion is important because

the actual foundational text rebuts this criticism without it empirical study of RRA mechanisms

(e.g., Bowles and Gintis 1976, pp. 143–144). Far becomes increasingly difficult, or perhaps even

more questionable, however, is whether the the- impossible, owing to challenges of statistical

ory allows non-class-based forms of oppression identification (i.e., too many parameters to

to matter for education (Davies 1995). It would estimate).

be difficult to maintain a structural Marxist posi- Effectively maintained inequality has been

tion while considering the history of Little Rock found in every nation for which studies assessing

and Birmingham, or the way in which post-World it exist (e.g., Lucas 2001 for the United States;

War II economic structure first rejected than Byrne and McCoy 2017 for Ireland; Byun and

embraced women’s paid labor force participa- Park 2017 for Korea; McKeever 2017 for South

tion. And, if one makes space for non-class-based Africa; Weiss and Schindler 2017 for Germany).

grounds for economic action, the theory’s under-

standing of schools is undermined.7 shows, their efforts grow increasingly aspirational and

decreasingly tied to empirical evidence, such that, in the

7 Self-described resistance theorists of a post-Marxist bent main, they fail to satisfy the coherence and falsifiability

claim to resolve this problem, but, as Davies (1995) criteria noted earlier. Thus, we do not include them.

100 S. R. Lucas and V. Irwin

Yet, no research assessing EMI has interrogated 4.15.1 Example 1: “Stratification
EMI’s claim of class-based collective action. Theory, Socioeconomic
While the widespread confirmatory research may Background, and Educational
seem to reflect a powerful theory, failure to assess Attainment: A Formal
its collective action assertion raises questions Analysis”
about the mechanisms the theory identifies.
Lucas (2009) formally translated EMI and
Given the existence of such critical observa- Maximally Maintained Inequality (MMI)
tions for each theory, it appears it would be (Raftery and Hout 1993) into mathematical equa-
worthwhile to assess, and even adjudicate, the tions and then considered those theories in con-
theories. cert with RRA, a theory that had already been
expressed mathematically. Working through the
4.15 A ssessing the Theories equations of these three theories revealed several
useful insights. One important finding is that
We have offered 10 theories of socioeconomic MMI is internally contradictory and tautologous,
and racial inequality in education. The large making it unfalsifiable and thus unworthy of con-
number of theories may reflect real complexity in sideration. For this reason, we did not discuss
the phenomenon. In contrast, however, it may MMI here. Lucas (2009, pp. 491–498) also estab-
instead be a result of sociology’s insufficient lished that EMI is not a tautology, showing that it
attention to the task of critically assessing or is possible to have outcome inequality associated
adjudicating theories. Or, a third option may be with origins yet reject EMI.
more appropriate—it may be that some theories
can be combined, ultimately leading to far fewer Lucas (2009) also found intriguing yet for-
than 10 theories of class and racial/ethnic inequal- merly unrecognized implications of RRA equa-
ity in education. tions, and intriguing possible connections
between RRA and EMI.  First, the analysis
There are at least two ways to proceed. One revealed that RRA implies the existence of a phe-
way is to conduct empirical analyses designed to nomenon Lucas (2009) labelled the Gates
assess two or more theories simultaneously. A Gambit. Essentially, RRA implies that the only
second way is to conduct purely theoretical com- socioeconomically advantaged students who will
parative analyses. Both approaches can reveal exit advanced programs are those who believe
whether a theory is viable and/or whether a com- their chances of matching or exceeding their par-
bination of two theories is worth pursuing. ents’ socioeconomic attainments are better if they
drop out. This pattern was named after Bill Gates,
Alas, purely theoretical assessments of theo- an adolescent of high socioeconomic status who,
ries are rare in the sociology of education. And, despite scoring 1590 on the pre-r­enormed SAT,
while empirical research is dominant, unfortu- dropped out of Harvard to pursue a career in
nately, most contemporary empirical research in computers, a decision that appears to have worked
the sociology of education focuses on establish- for him (Lucas 2009, p. 508, note 5). At the same
ing a given theory, rather than critically adjudi- time, by simplifying RRA equations it was shown
cating multiple theories. Thus, to illustrate the that RRA implies that all other high socioeco-
potential power of work geared to comparing and nomic background students will stay in school
adjudicating theories, we provide three examples, and enter demanding programs, and they will do
one purely theoretical and two empirical. The so without considering their subjective likelihood
purely theoretical work assesses three theories of of succeeding in school. This implication ­tumbles
inequality, of which we will discuss only two. directly out of the equations specifying RRA
The empirical studies can be used to consider (Lucas 2009, pp.  482–483). Thus, despite the
multiple theories as well, even if the original summary claims of the non-mathematical sum-
paper did not.

4  Race, Class, and Theories of Inequality in the Sociology of Education 101

mary of RRA, which state that students consider ticular level of success in school, but students of
their likelihood of success in school as they make lower socioeconomic status act as if they have no
rational choice decisions of whether to continue, such estimate, i.e., they do not reference esti-
the actual equations of the theory imply other- mates of future occupational success. This differ-
wise for particular classes of students. ential is consistent with differential myopia.

Notably, this RRA claim is consistent with Such findings provide new, more focused
EMI’s claim that academically mediocre high grounds for empirical research, and, thus, prom-
socioeconomic background students enter ising opportunities for theory adjudication and/or
demanding programs while their equally adept synthesis. For example, the results imply that
low-socioeconomic background peers do not. EMI analysts interested in adjudicating between RRA
highlights the use of non-academic resources (e.g., and EMI should not devote time to assessing the
pressure well-off parents apply to school gatekeep- existence of student myopia, for doing so will not
ers to secure their children’ admission to demand- adjudicate between EMI and RRA because both
ing programs) to predict and explain this pattern. theories predict myopia for some students. Thus,
Thus, the theories are complementary as follows. it appears that assessing the coherence of multi-
ple theories can pay large dividends.
RRA equations imply a pattern of behavior—
the entry of mediocre, well-off students into pro- 4.15.2 Example 2: “A Threat
grams for high achievers—but because RRA in the Air: How Stereotypes
allows entry to demanding programs only on the Shape Intellectual Identity
basis of merit (e.g., prior achievement) and ability and Performance”
to pay, RRA processes of entry deny the possibility
of mediocre well-off students entering demanding Stereotype threat (Steele 1997) occurs when a
educational programs. Thus, RRA equations imply negative stereotype becomes self-relevant and
a behavior, but RRA relations offer no means for fear of fulfilling this stereotype actually impedes
the behavior to be enacted. EMI, however, by not- performance. Stereotype threat has generally
ing the role of gatekeepers holding discretionary been studied in relation to race and gender stereo-
power, provides a way for the implications embed- types in academic performance, but can be applied
ded in RRA equations to be realized. Thus, EMI to any group, including low-income students,
complements RRA by providing a pathway for the who face negative stereotypes about their perfor-
outcome RRA equations predict—mediocre high mance. Studies have triggered stereotype threat
status students’ entry to demanding programs. The both through the labeling of tests as diagnostic of
pathway is gatekeeper discretion. ability (e.g., Steele and Aronson 1995) and
through the presence of a White examiner (e.g.,
This is not the only example of how RRA and Huang 2009); neither of these designs stipulates
EMI may be complementary. Another example the presence of a prejudiced observer or evaluator
flows from EMI’s effort to rebut the neo-classical (e.g., teacher). Thus, the threat is particularly
economic position that students act with fore- insidious, because it does not require the gate-
sight. EMI contended that myopia is differen- keeper with which the person interacts to hold the
tially distributed, and that it is a feature of the stereotype, it is only necessary that a student be
process. It turns out that once one works through conscious of the stereotype. Opportunities for ste-
the equations of RRA, one finds that RRA implies reotype threat to occur are many, extending far
decision processes consistent with differential beyond the school to experiences with family,
myopia. This possible complementarity is friends, co-workers, employers, and more.
­powerful because, as a rational choice theory,
RRA might be expected to deny myopia. Yet, The implications for class and racial/ethnic
simplifying the equations reveals that RRA indi- inequalities in education flow from the flood of
cates that students of well-off parents utilize a stereotypes students encounter daily regarding the
subjective estimate of their likelihood of attain-
ing various occupational positions given a par-

102 S. R. Lucas and V. Irwin

abilities and relative rankings of different groups 4.15.3 Example 3: Unequal
Childhoods
of students. It is possible that a constant low level

of threat underlies some poor and racial/ethnic

minority students’ entire school experience. Schools expect (and generally require) that stu-

Stereotype threat resonates with theories that dents will interact with teachers and other author-

explain educational inequality through expecta- ities in certain ways, but students may not arrive

tions. For example, social-psychological pro- at school equally prepared to do so. Lareau

cesses are the key mechanism of the Wisconsin (2003) suggests that this is related to the way that

model; the model argues that students’ aspira- parents employ language and discipline with

tions are shaped by the influence of significant their children. Lareau identifies two different par-

others, with teachers being an important such enting strategies: concerted cultivation and the

other. Yet, stereotype threat evidence both inten- accomplishment of natural growth. Concerted

sifies the potential role of teachers, while broad- cultivation, the child-rearing strategy associated

ening the sources of influence by noting that with the middle-class, is characterized by highly

expectations of generalized (i.e., nonsignificant) structured time, and eventual conversation and

others can also matter for students’ later attain- negotiation in the practice of discipline. Lareau

ments. Thus, existence of stereotype threat is not argues that such practices reflect and facilitate the

only consistent with the Wisconsin Model, it sug- skills, knowledge, and interpersonal postures

gests an intriguing elaboration of the model; it is rewarded by the school. In contrast, the accom-

an elaboration because it, too, emphasizes social-­ plishment of natural growth, the parenting style

psychological processes at its core. more commonly adopted by working-class and

Stereotypes develop in historical context, and poor families, is characterized by unstructured

education-related racial stereotypes tend to track time, more directive language use, and authori-

with Ogbu’s involuntary (e.g., Black students are tarian discipline. Importantly, Lareau argues that

less motivated and able than White students) and these different patterns of socialization are asso-

voluntary (e.g., “Asian” students are model minor- ciated with different levels of comfort and ability

ities) immigrant designations. In that sense, there in interacting with authority.

is a parallel between the phenomena to which stu- These findings parallel those of Kohn (e.g.,

dents are responding vis à vis stereotype threat 1969) and of Bernstein (e.g., 1971), and contrib-

and according to incorporation theory. However, ute to research traditions on language use in com-

why involuntary/stereotyped students underper- munities and its impact on schooling. For

form differs. Thus, while stereotype threat is con- example, Nystrand and Gamoran (1988, 1991)

sistent with incorporation theory, it is not evidence distinguish authentic and inauthentic questions.

of the reduced school engagement that the theory Authentic questions are questions to which the

suggests. Indeed, a scope condition for stereotype asker does not know the answer. Inauthentic

threat to occur is that the person must care about questions are questions to which the asker does

the domain at issue (Aronson et  al. 1999), and know the answer. Nystrand and Gamoran (1991)

empirical evidence indicates the strongest, not the find that authentic questions are associated with

weakest, students are affected by it (e.g., Steele greater learning.

1999). It is only because the student cares about Research indicates that middle-class and White

success in the domain at issue that anxiety associ- communities tend to use inauthentic questions in

ated with confirming a negative stereotype rises early childhood language training, whereas other

enough to lower performance quality. communities use authentic questions (e.g., Heath

4  Race, Class, and Theories of Inequality in the Sociology of Education 103

1983). When students arrive at school, an institution Scientists, jurists, artists, journalists, designers,
with a predominance of inauthentic questions, some engineers, and other accomplished adults rely on
students, unfamiliar with such an odd language situ- complex forms of communication both to conduct
ation—Why would someone ask me a question to their work and to express their conclusions. The
which I know they know I know they know the language they use—verbal, symbolic, and visual—
answer?—are more likely to be made uncomfort- includes qualifications, nuances, elaborations,
able or unsure. The resulting befuddlement and hesi- details, and analogues woven into extended expo-
tation can quickly set students on a path to failure. sitions, narratives, explanations, justifications, and
dialogue. In contrast, much of the communication
Lareau’s findings would appear to parallel the demanded in school requires only brief answers:
correspondence principal and Kohn’s work in true or false, multiple choice, fill in the blank, or
particular. While Bowles and Gintis focus on the short sentences (e.g., “Prices increase when
socialization that happens within the school, the demand exceeds supply”). (1996, pp. 283–284)
contrast between concerted cultivation and the
accomplishment of natural growth suggests that One implication of the middle-class use of
the divergence in training for class-stratified inauthentic questions in child development is that
positions in adulthood begins before children in order for middle-class children to attain their
enter school. Thus, the predicted reproduction is parent’s occupational positions, their inauthentic-
even more rigid, because working-class students question-based childhood communication pat-
are not only more likely to be placed in substan- terns must someday be undone. In contrast, many
dard academic settings, but Lareau’s findings Black children engage authentic questions at an
suggest that working-class and poor children will early age, meaning that they enter school ready
be less likely to strike the posture that their and able to engage in complex communication
schools value. In this way, we can see how diver- forms, in a sense ahead of the game. But, after
gent child-rearing and language acquisition strat- intense involvement with a school communica-
egies might promote the kind of disjuncture tive environment that re-labels their creativity as
between community and school reflected in deficiency, their linguistic advantage is lost.
Ogbu’s and Carter’s discussions of oppositional
culture. Seen in this way, at least some notable non-­
correspondences are evident, a fact quite consis-
However, arbitrariness of school procedures, tent with Bourdieu’s perspective on cultural
not correspondence, is also evident in such analy- capital, especially the variant highlighting the
ses. Heath (1983) documented the rich language social construction of skill.
use and talent of children raised in homes that use
authentic questions, and how changes in school Lareau (2003) does not support the “burden of
practice made their school achievements improve. acting White” hypothesis, as the findings connect
For every class difference one could consider the child-rearing strategies to class, rather than race,
question of “Which is better?” For example, and also offer no suggestion that either the chil-
Lucas asks: dren or their parents devalue education, only that
they interact differently with school authority.
Are inauthentic questions “better” for teaching chil-
dren? Most analyses say no; although inauthentic Lareau’s work also demonstrates the impor-
questions have their place, they are overused in U.S. tance of significant others’ influence. In concerted
education (Newmann et al. 1996). Further, they fail cultivation and the accomplishment of natural
to match the aim of education in a globalizing, growth, parents set implicit expectations for the
highly competitive, neoliberal, take-­no-­prisoners manner in which children will structure and ori-
economy, and they do not match the aim of many ent their time. Because the former is in line with
parents to empower their children in the social, the expectations of education authorities (e.g.,
political, and economic arenas. (Lucas 2013, p. 71) college admissions officers), middle-class stu-
dents can be expected to attain higher levels of
Newmann, Marks, and Gamoran highlight the education. Moreover, while parents’ encourage-
mismatch, contending that: ment of certain styles of interaction with author-
ity is important, the effect escalates to the extent
that middle-class children are also given greater

104 S. R. Lucas and V. Irwin

access to authority figures at younger ages. Bachman, J.  G., Wallace, J.  M., Jr, O’Malley P.  M.,
Middle-class parents accomplish this by enroll- Johnston, L.  D., Kurth, C.  L., & Neighbors, H.  W.
ing their children in all kinds of organized activi- (1991). Racial/ethnic differences in smoking, drink-
ties, like sports and music lessons. This gives ing, and illicit drug use among American high school
middle-class children many more opportunities to seniors, 1976–89. American Journal of Public Health,
build their comfort with authority figures. 81, 372–377.

4.16 C oncluding Remarks Becker, G.  S. (1962). Investment in human capital: A
theoretical analysis. Journal of Political Economy, 70,
Evidence indicates that class and racial/ethnic S9–S49.
inequality in education is ubiquitious or perhaps
even universal. Analysts have proposed multiple Becker, G. S., & Tomes, N. (1986). Human capital and the
theories to explain the documented inequalities rise and fall of families. Journal of Labor Economics,
and their intransigence. Even so, many theories 4, S1–S39.
suggest mechanisms that might be manipulable
enough to reduce, or even eliminate, class- and Benavot, A. (1983). The rise and decline of vocational
racial/ethnic-linked educational inequality. Yet, education. Sociology of Education, 56, 63–76.
prior to the challenge of constructing the political
will to engage such mechanisms, analysts must Bernstein, B. (1971). Class, codes and control. Volume 1:
intensify their efforts to assess the theories Theoretical studies towards a sociology of language.
through which those potential mechanisms are Boston: Routledge/Kegan Paul.
identified. As analysts deepen their engagement
with this task, it is likely that some theories will Blau, Peter M., & Duncan, O. D., with the collaboration
be found wanting. At the same time, new, more, of Andrea Tyree. (1967). The American occupational
full comprehension of the maintenance of structure. New York: The Free Press.
inequality may come within reach. In this way,
sociologists may contribute to closing the gap not Bobo, L., & Licari, F. C. (1989). Education and political
only between classes and racial/ethnic groups in tolerance: Testing the effects of cognitive sophistica-
achievement and attainment but, also, to reduc- tion and target group affect. Public Opinion Quarterly,
ing the gap between humans’ cognitive potential 53, 285–308.
and realized cognitive achievement. Perhaps the
possible gains to such a closure, and the prospect Bourdieu, P. (1986). The forms of capital. In J. Richardson
of sociologists contributing to such an enterprise, (Ed.), Handbook of theory and research for the
will spur the next adjudicatory steps in the sociology of education (pp.  241–258). New  York:
research agenda of sociologists of education. Greenwood Press.

References Bourdieu, P., & Passeron, J.-C. (1977). Reproduction in
education, society, and culture (2nd ed.). London:
Ainsworth-Darnell, J.  W., & Downey, D.  B. (1998). SAGE Publications.
Assessing the oppositional culture explanation for
racial/ethnic differences in school performance. Bowles, S., & Gintis, H. (1976). Schooling in capitalist
American Sociological Review, 63, 536–553. America. New York: Basic Books.

Aronson, J., Lustina, M. J., Good, C., Keough, K., Steele, Breen, R., & Goldthorpe, J. H. (1997). Explaining educa-
C.  M., & Brown, J.  (1999). When White men can’t tional differentials: Towards a formal rational action
do math: Necessary and sufficient factors in stereotype theory. Rationality and Society, 9, 275–305.
threat. Journal of Experimental Social Psychology, 35,
29–46. Byrne, D., & McCoy, S. (2017). Effectively maintained
inequality in compulsory and post-compulsory educa-
tion in the Republic of Ireland. American Behavioral
Scientist, 64, 49–73.

Byrus, R.  T., & Stone, G.  W. (1984). Microeconomics
(2nd ed.). Glenview: Scott Foresman.

Byun, S.-y., & Park, H. (2017). When different types of
education matter: Effectively maintained inequal-
ity of educational opportunity in Korea. American
Behavioral Scientist, 64, 94–113.

Carnoy, M., & Rothstein, R. (2013). What do interna-

tional tests really show about U.S. student perfor-
mance? Washington, DC: Economic Policy Institute.
Carter, P.  L. (2003). “Black” cultural capital, status posi-
tioning, and schooling conflicts for low-income African
American youth. Social Problems, 50, 136–155.
Carter, P. L. (2005). Keepin’it real: School success beyond
Black and White. New York: Oxford University Press.
Carter, P. L. (2006). Straddling boundaries: Identity, cul-
ture, and school. Sociology of Education, 79, 304–328.
Cherlin, A., & Walters, P.  B. (1981). Trends in United
States men’s and women’s sex-role attitudes: 1972 to
1978. American Sociological Review, 46, 453–460.

4  Race, Class, and Theories of Inequality in the Sociology of Education 105

Coleman, J. S., with the assistance of Johnstone, J. W. C, Grubb, W. N. (1992). The economic returns to baccalau-
& Jonassohn, K. (1961). The adolescent society: The reate degrees: New evidence from the class of 1972.
social life of the teenager and its impact on education. Review of Higher Education, 15, 213–231.
New York: The Free Press.
Grubb, W. N. (2002). Learning and earning in the middle,
Collins, R. (1974). Where are educational requirements part I: National studies of pre-baccalaureate educa-
for employment highest? Sociology of Education, 47, tion. Economics of Education Review, 21, 299–321.
419–442.
Halaby, C. N., & Weakliem, D. L. (1993). Ownership and
Collins, R. (1977). Some comparative principles of edu- authority in the earnings function: Nonnested tests
cational stratification. Harvard Educational Review, of alternative specifications. American Sociological
47, 1–27. Review, 58, 16–30.

Collins, R. (1979). The credential society: An historical Harris, A.  L. (2006). I (Don’t) hate school: Revisiting
sociology of education and stratification. San Diego: oppositional culture theory of Blacks’ resistance to
Academic. schooling. Social Forces, 85, 797–834.

Daniels, M., Devlin, B., & Roeder, K. (1997). Of genes Harris, K.  M., Duncan, G.  J., & Boisjoly, J.  (2002).
and IQ.  In B.  Devlin, S.  E. Fienberg, D.  P. Resnick, Evaluating the role of “nothing to lose” attitudes on
& K.  Roeder (Eds.), Intelligence, genes, & success: risky behavior in adolescence. Social Forces, 80,
Scientists respond to the bell curve (pp.  45–70). 1005–1039.
New York: Springer.
Hauser, R.  M., Tsai, S.-L., & Sewell, W.  H. (1983). A
Davies, S. (1995). Leaps of faith: Shifting currents in model of stratification with response error in social
critical sociology of education. American Journal of and psychological variables. Sociology of Education,
Sociology, 100, 1448–1478. 56, 20–46.

Defina, R., & Hannon, L. (2016). Social status attainment Heath, S.  B. (1983). Ways with words: Language, life,
and racial category selection in the contemporary and work in communities and classrooms. New York:
United States. Research in Social Stratification and Cambridge University Press.
Mobility, 44, 91–97.
Hedström, P. (2005). Dissecting the social: On the prin-
Downey, D.  B., Ainsworth-Darnell, J.  W., & Qian, Z. ciples of analytic sociology. New  York: Cambridge
(2009). Rethinking the attitude–achievement paradox University Press.
among Blacks. Sociology of Education, 82, 1–19.
Herrnstein, R.  J., & Murray, C. (1994). The bell curve:
Esping-Anderson, G., & Wagner, S. (2012). Asymmetries Intelligence and class structure in American life.
in the opportunity structure: Intergenerational mobil- New York: Free Press.
ity trends in Europe. Research in Social Stratification
and Mobility, 30, 473–487. Howell, C., & Turner, S.  E. (2004). Legacies in Black
and White: The racial composition of the legacy pool.
Featherman, D. L., & Hauser, R. M. (1978). Opportunity Research in Higher Education, 45(4), 325–351.
and change. New York: Academic.
Huang, M.-H. (2009). Race of the interviewer and the
Feliciano, C., & Lanuza, Y.  R. (2017). An immigrant Black–White test score gap. Social Science Research,
paradox? Contextual attainment and intergenerational 38, 29–38.
educational mobility. American Sociological Review,
82, 211–241. Jacoby, D. (1991). The transformation of industrial
apprenticeship in the United States. Journal of
Fischer, C., Hout, M., Jankowski, M.  S., Lucas, S.  R., Economic History, 51, 887–910.
Swidler, A., & Voss, K. (1996). Inequality by design:
Cracking the bell curve myth. Princeton: Princeton Jaynes, G. D., & Williams, R. M., Jr. (1989). A common
University Press. destiny: Blacks and American society. Washington,
DC: National Academy Press.
Flores-Gonzàlez, N. (2002). School kids/street kids:
Identity development in Latino students. New  York: Jencks, C., & Phillips, M. (1998). The Black–White test
Teachers College Press. score gap. Washington, DC: Brookings Institution
Press.
Fordham, S., & Ogbu, J. U. (1986). Black students’ school
success: Coping with the burden of acting White. Jencks, C., Perman, L., & Rainwater, L. (1988). What is
Urban Review, 18(3), 176–206. a good job? A new measure of labor-market success.
American Journal of Sociology, 93, 1322–1357.
Giroux, H. A. (1981). Hegemony, resistance, and the para-
dox of educational reform. Interchange, 12(2), 3–26. Jensen, A. (1969). How much can we boost IQ and scho-
lastic achievement? Harvard Educational Review, 39,
Goldberger, A.  S. (1991). A course in econometrics. 1–123.
Cambridge, MA: Harvard University Press.
Jones, J.  H. (1981). Bad blood: The Tuskeegee syphi-
Goodman, J.  D. (1979). The economic returns of edu- lis experiment: A tragedy of race and medicine.
cation: An assessment of alternative models. Social Washington, DC: The Free Press.
Science Quarterly, 60, 269–283.
Kalleberg, A., & Griffin, L. (1980). Class, occupation,
Gottfredson, D.  C. (1981). Black–White differences in and inequality in job rewards. American Journal of
the educational attainment process: What have we Sociology, 85, 731–768.
learned? American Sociological Review, 46, 542–557.
Kalmijn, M. (1994). Assortative mating by cultural and
Greenwood, E. (1957). Attributes of a profession. Social economic occupational status. American Journal of
Work, 2(3), 45–55. Sociology, 100, 422–452.

106 S. R. Lucas and V. Irwin

Kalmijn, M. (2001). Assortative meeting and mating: Lucas, S. R. (2017). An archaeology of effectively main-
Unintended consequences of organized settings for tained inequality. American Behavioral Scientist, 61,
partner choices. Social Forces, 79, 1289–1312. 8–29.

Kerckhoff, A.  C. (1976). The status attainment pro- Lucas, S. R., & Beresford, L. (2010). Naming and clas-
cess: Socialization or allocation? Social Forces, 55, sifying: Theory, evidence, and equity in education.
368–381. Review of Research in Education, 34, 25–85.

Kerckhoff, A. C., & Campbell, R. T. (1977). Black–White Lumey, L.  H. (1992). Decreased birthweights in infants
differences in the educational attainment process. after maternal in utero exposure to the Dutch famine of
Sociology of Education, 50, 15–27. 1944–1945. Paediatric and Perinatal Epidemiology, 6,
240–253.
Kimbro, R. T., Bzostek, S., Goldman, N., & Rodríguez, G.
(2008). Race, ethnicity, and the education gradient in MacLeod, J. (1987). Ain’t no makin’ it: Aspirations and
health. Health Affairs, 27, 361–372. attainment in a low-income neighborhood. Boulder:
Westview Press.
Kitagawa, E. M., & Hauser, P. M. (1968). Education dif-
ferentials in mortality by cause of death: United States Marsh, D. G. (1997). Approaches toward the genetic anal-
1960. Demography, 5, 318–354. ysis of complex traits: Asthma and atopy. American

Klegon, D. (1978). The sociology of professions: An Journal of Respiratory and Critical Care Medicine,
emerging perspective. Work & Occupations, 5, 156, S133–S138.
259–283. McBean, A.  M., & Gornick, M. (1994). Differences by
race in the rates of procedures performed in hospitals
Kohn, M. (1969). Class and conformity: A study in values. for Medicare beneficiaries. Health Care Financing
Homewood: Dorsey Press. Review, 15(4), 77–90.
McKeever, M. (2017). Educational inequality in apart-
Kohn, M., & Schooler, C. (1969). Class, occupation, heid South Africa. American Behavioral Scientist, 64,
and orientation. American Sociological Review, 34, 114–131.
659–678. McNeil, L. M. (1981). Negotiating classroom knowledge:
Beyond achievement and socialization. Journal of
Kramer, R., DeFina, R., & Hannon, L. (2016). Racial Curriculum Studies, 13, 313–328.
rigidity in the United States: Comment on Saperstein McNulty, T.  L., & Bellair, P.  E. (2003). Explaining
and Penner. American Journal of Sociology, 122, racial and ethnic differences in adolescent violence:
233–246. Structural disadvantage, family well-being, and social
capital. Justice Quarterly, 20, 1–31.
Lamont, M., & Lareau, A. (1988). Cultural capital: Meaney, M. (2010). Epigenetics and the biological defi-
Allusions, gaps, and glissandos in recent theoretical nition of gene × environment interactions. Child
developments. Sociological Theory, 6, 153–168. Development, 81, 41–79.
Mitchell, D.  E., & Kerchner, C.  T. (1983). Labor rela-
Lareau, A. (2003). Unequal childhoods: Class, race, and tions and teacher policy. In L. S. Shulman & G. Sykes
family life. Berkeley: University of California Press. (Eds.), Handbook of teaching and policy (pp.  214–
238). New York: Longman Publishing Group.
Lareau, A., & Weininger, E.  B. (2003). Cultural capital Morgan, S.  L. (2004). Methodologist as arbitrator: Five
in educational research: A critical assessment. Theory models for Black–White differences in the causal
and Society, 32, 567–606. effect of expectations on attainment. Sociological
Methods and Research, 33, 3–53.
Lieberson, S. (1980). A piece of the pie: Blacks and NCES. (2013). Performance of U.S. 15-year-old stu-
White immigrants since 1880. Berkeley: University of
California Press. dents in mathematics, science, and reading literacy
in an international context: First look at PISA 2012.
Lieberson, S. (1985). Making it count: The improvement Washington, DC: U.S. Department of Education.
of social research and theory. Berkeley: University of Newmann, F. M., Marks, H. M., & Gamoran, A. (1996).
California Press. Authentic pedagogy and student performance.
American Journal of Education, 104, 280–312.
Loury, G.  C. (1992). Incentive effects of affirmative Nystrand, M., & Gamoran, A. (1988). A Study of instruc-
action. Annals of the American Academy of Political tion as discourse. Washington, DC: U.S. Department
and Social Science, 523, 19–29. of Education, Office of Educational Research and
Improvement. Education Resources Information
Lucas, S.  R. (2001). Effectively maintained inequality: Center ED328516.
Education transitions, track mobility, and social back- Nystrand, M., & Gamoran, A. (1991). Instructional dis-
ground effects. American Journal of Sociology, 106, course, student engagement, and literature achieve-
1642–1690. ment. Research in the Teaching of English, 25,
261–290.
Lucas, S. R. (2009). Stratification theory, socioeconomic Ogbu, J.  U. (1978). Minority education and caste (Vol.
background, and educational attainment: A formal 581). New York: Academic Press.
analysis. Rationality & Society, 21, 459–511.

Lucas, S.  R. (2013). Just who loses? Discrimination in
the United States (Vol. 2). Philadelphia: Temple
University Press.

Lucas, S.  R. (2016). Where the rubber meets the road:
Probability and non-probability moments in experi-
ment, interview, archival, administrative, and eth-
nographic data collection. Socius: Sociological
Research for a Dynamic World, 2. https://doi.
org/10.1177/2378023116634709.

4  Race, Class, and Theories of Inequality in the Sociology of Education 107

Ogbu, J.  U. (1987). Variability in minority school Sørensen, A.  B. (2000). Toward a sounder basis for
performance: A problem in search of an explana- class analysis. American Journal of Sociology, 105,
tion. Anthropology & Education Quarterly, 18, 1523–1558.
312–334.
Spaeth, J.  (1985). Job power and earnings. American
Ogbu, J. U. (2003). Black American students in an affluent Sociological Review, 50, 603–617.
suburb: A study of academic disengagement. Mahwah:
Erlbaum Associates. Steele, C. M. (1997). A threat in the air: How stereotypes
shape intellectual identity and performance. American
Paulsen, R. (1991). Education, social class, and participa- Psychologist, 52, 613–629.
tion in collective action. Sociology of Education, 64,
96–110. Steele, C.  M. (1999). Thin ice: Stereotype threat and
Black college students. The Atlantic, 284, 44–54.
Perreira, K. M., Harris, K. M., & Lee, D. (2006). Making
it in America: High school completion by immigrant Steele, C. M., & Aronson, J. (1995). Stereotype threat and
and native youth. Demography, 43, 511–536. the intellectual test performance of African Americans.
Journal of Personality and Social Psychology, 69,
Raftery, A. E., & Hout, M. (1993). Maximally maintained 797–811.
inequality: Expansion, reform, and opportunity in
Irish education, 1921–75. Sociology of Education, 66, Sweeney, M., & Cancian, M. (2004). The changing
41–62. importance of White women’s economic prospects
for assortative mating. Journal of Marriage and the
Rumbaut, R. G. (1999). Assimilation and its discontents: Family, 66, 1015–1028.
Ironies and paradoxes. In C. Hirschman, P. Kasinitz,
& J.  DeWind (Eds.), The handbook of international Tomes, N. (1981). The family, inheritance, and the inter-
migration: The American experience (pp.  172–195). generational transmission of inequality. Journal of
New York: Russell Sage Foundation. political Economy, 89, 928–958.

Saperstein, A., & Penner, A.  M. (2010). The race of a Tyson, K. (2002). Weighing in: Elementary-age students
criminal record: How incarceration colors racial per- and the debate on attitudes toward school among
ceptions. Social Problems, 57, 92–113. Black students. Social Forces, 80, 1157–1189.

Saperstein, A., & Penner, A.  M. (2012). Racial fluidity Tyson, K., Darity, W., Jr., & Castellino, D. R. (2005). It’s
and inequality in the United States. American Journal not “a Black thing”: Understanding the burden of act-
of Sociology, 118, 676–727. ing White and other dilemmas of high achievement.
American Sociological Review, 70, 582–605.
Saporito, S., & Sohoni, D. (2006). Coloring outside the
lines: Racial segregation in public schools and their Useem, E. (1992). Middle schools and math groups:
attendance boundaries. Sociology of Education, 79, Parents’ involvement in children’s placement.
81–105. Sociology of Education, 65, 263–279.

Saporito, S., & Sohoni, D. (2007). Mapping educational Weil, F.  D. (1985). The variable effects of education on
inequality: Concentrations of poverty among poor and liberal attitudes: A comparative-historical analysis
minority students in public schools. Social Forces, 85, of anti-semitism using public opinion survey data.
1227–1253. American Sociological Review, 50, 458–474.

Schwartz, C.  R., & Mare, R.  D. (2005). Trends in edu- Weiss, F., & Schindler, S. (2017). EMI in Germany—
cational assortative marriage from 1940 to 2003. Qualitative differentiation in a tracked education sys-
Demography, 42, 621–646. tem. American Behavioral Scientist, 64, 74–93.

Sewell, W.  H., & Hauser, R.  M. (1980). The Wisconsin Willis, P. (1977). Learning to labour. Farnborough: Saxon
Longitudinal Study of social and psychological factors House.
in aspirations and achievement. Research in Sociology
of Education and Socialization, 1, 59–99. Wolfle, L.  M. (1985). Postsecondary educational
attainment among Whites and Blacks. American
Silberberg, E. (1990). The structure of economics: A Educational Research Journal, 22, 501–525.
mathematical analysis (2nd ed.). San Francisco:
McGraw-Hill. Wright, E.  O., & Perrone, L. (1977). Marxist class cat-
egories and income inequality. American Sociological
Review, 42, 32–55.

Educational Achievement 5
and Attainment Differences
Among Minorities and Immigrants

Phoebe Ho and Grace Kao

Abstract 5.1 I ntroduction
The U.S. student population is increasingly
comprised of racial/ethnic minority and immi- Recent estimates show that nearly half of the 50
grant students. Drawing on national-level million students enrolled in public elementary
data, we document the gaps in educational and secondary schools in the U.S. are racial and
achievement and attainment for minority and ethnic minorities. Specifically, the student popu-
immigrant students that are apparent at all lev- lation in public schools is 51% White, 16%
els of education, from early education through Black, 24% Hispanic, 5% Asian/Pacific Islander,
postsecondary schooling. These achievement and 1% American Indian/Alaska Native.1 In some
gaps reflect, in part, the broader racial and eth- of the largest urban school districts in the U.S.,
nic hierarchy of the U.S., but the experiences the student population is already “majority
of immigrant-origin minority students addi- minority” (Aud et al. 2010). Moreover, racial and
tionally contribute to the complexity of racial ethnic differences in academic achievement and
and ethnic stratification in education. Though attainment are longstanding and continue to be
research shows that socioeconomic status the subject of much research and debate (Kao and
accounts for much of the differences in Thompson 2003; Noguera 2008). The U.S. stu-
achievement, factors such as schools and dent population also includes a significant num-
teachers, peer relationships, and neighbor- ber of children of immigrants. Nearly one in four
hoods and communities may also contribute to children have at least one immigrant parent
the variation in academic outcomes. (Fortuny et al. 2009), and by 2050, an estimated
one in three children will come from immigrant
families (Passel 2011). Further, the children of
immigrants are highly diverse—about 58% are
Hispanic, 19% are Asian, 16% are White, and 9%
are Black (The Urban Institute n.d.).

P. Ho (*) 1 The U.S. Department of Education is the source for much
Department of Sociology, University of Pennsylvania, of the data presented in this chapter and typically com-
Philadelphia, PA, USA bines Asian and Pacific Islander populations into one cat-
e-mail: [email protected] egory. We recognize that this broad category masks
considerable diversity and, where possible, we present
G. Kao data for sub-groups.
Department of Sociology, Yale University,
New Haven, CT, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 109
B. Schneider (ed.), Handbook of the Sociology of Education in the 21st Century, Handbooks
of Sociology and Social Research, https://doi.org/10.1007/978-3-319-76694-2_5

110 P. Ho and G. Kao

Scholars have proposed various scenarios for 5.2 E arly Education

how the U.S. racial and ethnic hierarchy might

change due to the diversity of immigrants, and Enrollment in early education helps children pre-

how such changes are likely to affect different pare academically for entry into formal school-

groups (Lee and Bean 2010). However, the recent ing. In the fall of 2014, about 41% of White 3- to

rise of anti-immigrant rhetoric and a new politi- 5-year-olds were enrolled in preschool, followed

cal administration that favors restrictive immi- by 40% of Asians, 39% of Blacks, 32% of

gration policies have arguably made the U.S. less Hispanics, and 31% of American Indians/Alaska

welcoming of immigrants more generally. As a Natives. Among children attending preschools,

result, immigrant children may face greater greater proportions of minority children did so

obstacles in the near future. While some cities for the full day compared to White children

such as San Francisco, Seattle, and Philadelphia (Kena et  al. 2016). Immigrant parents are less

and a number of college campuses have declared likely to enroll their children in center-based care

themselves as sanctuary sites, proposed policies (Karoly and Gonzalez 2011). For minority and

that target individuals from specific countries and immigrant children, access to early education

undocumented individuals threaten educational may help them adapt to the “middle-class main-

opportunities. Elsewhere, this volume examines stream” norms expected by schools (Entwisle

undocumented children, who will suffer the and Alexander 1993). Access to early education

greatest impact of the current administration’s can strengthen the English language skills of

focus on the deportation of undocumented adults. children with immigrant parents (Karoly and

A non-trivial share of native-born children from Gonzalez 2011). Moreover, early childcare cen-

immigrant families come from families with ters serve as important facilitators of social capi-

mixed legal statuses (Fix and Zimmermann tal, providing mothers with access to a broader

2001). In such families, children with legal status network of parents and resources (Small 2009).

may have a parent, sibling, or other close relative There is some evidence that Black children

who is undocumented. Such families are at risk receive lower-quality care than White children in

of being separated and face significant challenges early education programs and that providing uni-

that will likely affect their children’s educational versal, quality early childhood education would

achievement. substantially reduce early achievement gaps for

Researchers commonly use educational both Black and Hispanic students (Magnuson

achievement and attainment measures to gauge and Waldfogel 2005).

the integration of minorities and immigrants. It is The Early Childhood Longitudinal Study

critical to understand the educational outcomes Birth Cohort of 2001 (ECLS-B 2001) is a nation-

of children of minority native-born and foreign-­ ally representative study conducted by the

born parents, especially in the context of growing Department of Education that administered tests

racial tensions. In this chapter, we compile data of letter and number and shape recognition to a

from U.S. Department of Education reports and sample of children who were about 4 years of age

studies to present an overview of racial, ethnic, in 2005–06. Overall, about 33% of children were

and immigrant differences in achievement and proficient in letter recognition and 65% were pro-

attainment from early education to postsecondary ficient in number and shape recognition. Race

completion. We then place educational outcomes and ethnic differences are already apparent at this

in context by drawing upon prior reviews of lit- early age. Asian children had the highest rates of

erature and highlighting illustrative examples of proficiency in both letter (49%) and number and

current empirical research. We do not focus on shape recognition (81%), followed by White chil-

gender differences or the experiences of undocu- dren (37% and 73%, respectively). In letter rec-

mented youth because other chapters in this vol- ognition, Black children had a proficiency rate of

ume do so. 28%, followed by 23% for Hispanic children, and

5  Educational Achievement and Attainment Differences Among Minorities and Immigrants 111

19% for American Indian/Alaska Native chil- peers. These differences have remained largely
dren. For number and shape recognition, Black unchanged over the past decade.
children had a proficiency rate of 55%, followed
by 51% for Hispanic children, and 40% for There are also stark differences in NAEP
American Indian/Alaska Native children (Aud scores by English language learner (ELL) status
et al. 2010). (Fig.  5.2).2 On average, non-ELL 4th-graders
outperform their ELL peers in both reading and
Studies have linked parenting behaviors and math, though differences are larger in reading
infant health to racial and ethnic differences in scores. In reading, non-ELL 4th-graders scored
early cognitive ability using ECLS-B data (Gibbs an average of 226 compared to 189 for their ELL
et  al. 2016; Lynch 2011). Lynch (2011) found peers. In math, non-ELL students had an average
that Black infants had poorer health (e.g., prema- score of 243 while ELL students had an average
ture birth, lower birth weight) than White infants. score of 218. The ELL disadvantage is present
Asian infants had better health and Hispanic across racial/ethnic groups. Further, racial/ethnic
infants did not differ from White infants. differences in ELL student performance mirror
Accounting for infant health explained a large those of non-ELL students, with Asian/Pacific
portion of the Black, but not Hispanic, disadvan- Islander and White ELL 4th-graders outperform-
tage in early educational outcomes and some of ing their Black and Hispanic ELL peers.
the Asian advantage. Other studies have found
that when socioeconomic factors, such as family Similar racial and ethnic patterns are seen in
income and parents’ education are taken into NAEP 8th-grade reading and math assessment
account, much of the gap in early educational trends (Fig.  5.1). Results from the 2015 assess-
outcomes for minority and immigrant children is ment show that Asian/Pacific Islander students
accounted for (Entwisle and Alexander 1993; have the highest average reading and math scores
Glick and Bates 2010). Understanding early dif- (280/306), followed by White students (274/292).
ferences in child developmental outcomes has Hispanic and American Indian/Alaska Native
implications for achievement gaps that are found students had similar reading and math scores
later in life, when children enter schools (Torche (253/270 and 252/267, respectively) while Black
2016). students had the overall lowest scores (248/260).
These racial/ethnic differences in reading and
5.3 P rimary and Secondary math achievement are also found among high
Education schoolers (Fig. 5.1). In the 2013 NAEP reading
assessment of 12th-graders, White students had
5.3.1 Test Scores the highest average score (297), followed by
Asian/Pacific Islander (296), American Indian/
Trends in reading and math performance of 4th-­ Alaska Native (277), Hispanic students (276),
graders in the main National Assessment of and Black (268) students. In math, Asian/Pacific
Educational Progress (NAEP) show persistent Islander students had the highest average score
differences by race/ethnicity (Fig. 5.1). In 2015, (172), followed by Whites (162), American
Asian/Pacific Islander 4th-graders had the high-
est achievement, with an average NAEP reading 2 We acknowledge that the term English language learner
score of 239 and an average NAEP math score of (ELL) is an imprecise measure of students’ immigrant sta-
257, followed by White students (232 and 248, tus. Unfortunately, the federal data used in this chapter do
respectively). In reading/math, Black (206/224), not provide measures of student or parent place of birth.
Hispanic (208/230), and American Indian/Alaska There may be immigrant students who are fluent in
Native (205/227) 4th-graders scored similarly, English and thus not classified as ELL and native-born
but below their White and Asian/Pacific Islander students who are classified as ELL. An ELL student, as
defined by the National Center for Education Statistics
(NCES), is one who has “sufficient difficulty speaking,
reading, writing, or understanding the English language.”

112 P. Ho and G. Kao

Fig. 5.1  Trends in NAEP reading and math scores by the right. Authors’ compilation of data from the NAEP
race/ethnicity. (Broken lines are due to lack of data for that Data Explorer (NDE), U.S.  Department of Education,
year. In 2005, the math portion of the NAEP for 12th-­ Institute of Education Sciences, National Center for
graders was redesigned with a new scoring scale—scores Education Statistics (https://nces.ed.gov/nationsreport-
from 2005 onwards are graphed on the secondary axis to card/naepdata/))

Indian/Alaska Native (142), Hispanic students ELL peers in both 8th and 12th grade, on average
(141), and Black students (132). and across racial/ethnic groups (Fig. 5.2). Among
8th-graders, non-ELL students had an average
There are large differences in both reading and reading score of 268 compared to a score of 223
math scores between non-ELL students and their

5  Educational Achievement and Attainment Differences Among Minorities and Immigrants 113

Fig. 5.2  Average NAEP reading and math scores in 2015 U.S.  Department of Education, Institute of Education
by ELL status and race/ethnicity. (Authors’ compilation Sciences, National Center for Education Statistics (https://
nces.ed.gov/nationsreportcard/naepdata/))
of data from the NAEP Data Explorer (NDE),

for ELL students. In math, non-ELL students had for ELL students. This pattern of ELL disadvan-
a score of 284 compared to 246 for their ELL tage holds across racial and ethnic groups in both
peers. Among 12th graders, non-ELL students 8th and 12th grade. However, racial and ethnic
had an average reading score of 290 compared to gaps among ELL students are generally smaller
237 for their ELL peers. In math, non-ELL stu- than those found among non-­ELL students.
dents scored an average of 155 compared to 109

114 P. Ho and G. Kao

5.3.2 High School Grades Hispanic (2.84) students, while Black students,
and Coursework on average, have the lowest GPAs (2.69).

The NAEP High School Transcript Study (HSTS) ELL students earn somewhat lower grades
collects transcript data on a nationally represen- than their non-ELL peers (Fig. 5.4). The average
tative sample of graduating U.S. high school stu- GPA for ELL students in 2009 was 2.75, 0.25
dents. Data from HSTS show that the racial and points lower than that of non-ELL students. For
ethnic and immigrant differences in test scores some racial/ethnic groups, ELL students earn
are mirrored in students’ grades and coursework comparable or even higher grades than their non-­
as well. Between 1990 and 2009, the average ELL peers. For example, Black ELL students
GPA of all students increased slightly, but racial/ have an average GPA of 2.75, higher than the
ethnic differences persist (Fig. 5.3). Asian/Pacific 2.69 average for non-ELL Black students.
Islander students maintain the highest GPAs Hispanic ELL students have an average GPA that
(3.26  in 2009), followed by White (3.09), is 0.18 points lower than their non-ELL counter-
American Indian/Alaska Native (2.87), and parts, smaller than the average non-ELL/ELL dif-
ference, and much smaller than the 0.30 point

Fig. 5.3  Trends in high
school achievement by
race/ethnicity. (Authors’
compilation of data from
the NAEP Data Explorer
(NDE), U.S. Department
of Education, Institute of
Education Sciences,
National Center for
Education Statistics
(https://nces.ed.gov/
nationsreportcard/
naepdata/))

5  Educational Achievement and Attainment Differences Among Minorities and Immigrants 115

Fig. 5.4  High school
achievement in 2009 by
ELL status and race/
ethnicity. (Authors’
compilation of data from
the NAEP Data Explorer
(NDE), U.S. Department
of Education, Institute of
Education Sciences,
National Center for
Education Statistics.
Data for White students
did not meet reporting
standards and are thus
not shown (https://nces.
ed.gov/
nationsreportcard/
naepdata/))

difference between Asian/Pacific Islander ELL in honors, Advanced Placement (AP), and

and non-ELL students. Moreover, among ELL International Baccalaureate (IB) courses can give

students, racial/ethnic differences in grades are them access to higher-quality instruction and

less pronounced. Black ELL students have an indicate their college readiness to postsecondary

average GPA comparable to the ELL student institutions. The increasing relevance of advanced

average while Hispanic ELL students have an coursework for high school students is evident in

average GPA just 0.08 points lower than the ELL the steep growth over the past two decades in the

average. In contrast, among non-ELL students, average number of advanced credits earned by

Black and Hispanic students have average GPAs students (Fig. 5.3). In 1990, with the exception of

that are 0.31 and 0.15 points lower than the non-­ Asian/Pacific Islander students who earned

ELL average, respectively. slightly less than 1.5 credits, all student groups

Because students are likely to encounter some accumulated on average less than one advanced

form of tracking once they enter formal school- course credit, defined as an honors, pre-AP/AP,

ing, it is important to examine differences in or pre-IB/IB course. By 2009, all racial and eth-

coursework. For high school students, enrolling nic groups of students on average had more

116 P. Ho and G. Kao

advanced course credits. However, the gaps as a GED certificate. In 2014, the average SDR
between racial/ethnic groups also sharply was about 7%, but this varied significantly by
increased. Asian/Pacific Islander students earned race, ethnicity, and nativity. Overall, Asian youths
an average of nearly seven advanced course cred- had the lowest average SDR (3%), followed by
its, while White students earned an average of Whites (5%), Blacks (7%), and Hispanics (11%).
just over four credits. Black, Hispanic, and However, among Hispanic and Asian subgroups,
American Indian/Alaskan Native students all average SDRs varied considerably. Among
accumulated on average between 2.5 and 3 Hispanics, Central American groups, such as
advanced course credits, less than half that of Guatemalans (29%) and Hondurans (20%), gen-
Asian/Pacific Islander students. erally had average SDRs higher than the Hispanic
average while South Americans, such as
The gap in advanced course credits between Colombians and Peruvians (both 3%), generally
non-ELL and ELL students is also substantial had lower average SDRs. The average SDR for
(Fig.  5.4). On average, non-ELL students had Mexicans (11%) was similar to the Hispanic
about four advanced course credits, compared average. Among Asians, average SDRs for
to less than one credit for ELL students. Black Nepalese (20%) and Burmese (28%) were much
and Hispanic ELL students earned an average higher than the average Asian SDR.  Hmong
of less than one advanced course credit, while (6%), Cambodian (8%), and Laotian (9%) youth
their non-­ELL counterparts accumulated an also had average SDRs higher than the Asian
average of between 2.5 to 3 credits, respec- average (Kena et al. 2016). These widely varying
tively. The ELL to non-ELL gap in credits estimates highlight the limitations of broad racial/
earned is especially large among Asian/Pacific ethnic categories such as Hispanic and Asian
Islander students—non-ELL students earned when analyzing educational outcomes, although
about seven credits compared to fewer than data limitations often preclude disaggregation by
two for ELL students. Thus, though ELL stu- subgroups.
dents had GPAs that were fairly comparable to
their non-ELL peers, they are less likely to Among U.S.-born youth, Asians had the low-
accumulate advanced credits. est average SDR (2%), followed by Whites (4%),
Blacks and Pacific Islanders (both 7%), Hispanics
5.4 H igh School Completion (8%), and American Indians/Alaska Natives
and College Readiness (11%). Among foreign-born youth, Asians and
Whites had average SDRs comparable to their
The Averaged Freshman Graduation Rate U.S.-born counterparts (3% and 4%, respec-
(AFGR) is a measure used by the Department of tively). Black immigrant youth had a slightly
Education that estimates on-time high school lower average SDR (6%) than their U.S.-born
graduation with a regular diploma. In 2013–14, peers while immigrant Hispanics and Pacific
the overall AFGR was estimated to be 82%. Islanders had much higher average SDRs (21%
Asian/Pacific Islander students had the highest and 23%, respectively) (Kena et  al. 2016).
AFGR—89%—followed by White students, at However, because the SDR measure is population-­
87%. Hispanic students had an AFGR of 76%, based and includes a broad age range, it likely
followed by Black (73%) and American Indian/ includes many immigrants who never attended
Alaska Native (70%) students (Kena et al. 2016). schools in the U.S. (Aud et al. 2010; Oropesa and
Landale 2009).
Another measure of high school completion is
the “status dropout rate” (SDR) which relies on Students who intend to enter postsecondary
census data to estimate the percentage of 16- to schooling usually have to take the SAT and/or the
24-year-olds who are not enrolled in school and ACT. Across SAT test subjects, White and Asian/
who have not received either a regular high Pacific Islander students have higher average
school diploma or an equivalent credential, such scores than Black, Hispanic, and American
Indian/Alaska Native students (The College

5  Educational Achievement and Attainment Differences Among Minorities and Immigrants 117

Board 2015). For the ACT, the percentage of an estimate of the percentage of 18- to 24-year-
2015 high school graduates who met ACT col- olds enrolled in college. In 2014, the average col-
lege readiness benchmarks also varied by race/ lege participation rate was about 40%. Asians had
ethnicity, with a higher percentage of White and the highest college participation rate (65%), fol-
Asian students meeting benchmarks than other lowed by Whites (42%), Pacific Islanders (41%),
racial/ethnic minority students (ACT, Inc. 2015). Hispanics and American Indians/Alaska Natives
Factors such as high school coursework and track (both 35%), and Blacks (33%) (Kena et al. 2016).
placement likely shape students’ preparedness Studies using nationally representative longitudi-
for college entrance tests. nal data find that differences in college enrollment
between White and minority students are largely
Researchers have also examined access to explained by differences in socioeconomic status
resources such as SAT/ACT test preparation and family background (Bennett and Xie 2003;
courses and private tutors. Some studies have Charles et al. 2007).
shown that minority students are more likely than
their White peers to use such strategies to improve Among White students enrolled in college in
their performance (Alon 2010; Buchmann et  al. 2013, about 35% attended a 2-year public institu-
2010; Byun and Park 2012; Espenshade and tion. This is in contrast to 49% of all Hispanic
Radford 2009). However, studies of low-income students enrolled in college who attended 2-year
urban Black and Hispanic youth show that such public institutions. About 45% of American
students generally report limited knowledge about Indian/Alaska Native college students, 39% of
college entrance exams and their importance in Black students, and 38% of both Asian and
college admissions and have less access to test Pacific Islander students attended public 2-year
preparation resources (Deil-Amen and Tevis colleges. About 40% of White and 44% of Asian
2010; Walpole et  al. 2005). While special pro- college students were enrolled in 4-year public
grams that seek to improve the college readiness institutions compared to 31% of both Pacific
of underrepresented minority students may be Islander and Black students and 34% of both
helpful, they likely offer fewer resources than Hispanic and American Indian/Alaska Native
what is available to students in high academic students. About 18% of White college students
tracks (Ochoa 2013). Cram schools often found in enrolled in private, not-for-profit 4-year institu-
Chinese and Korean ethnic communities may tions, followed by 14% of Asian students, 13% of
offer even less wealthy Asian American students both Black and Pacific Islander students, 11% of
access to supplementary education services (Byun American Indian/Alaska Native students, and
and Park 2012; Lee and Zhou 2015), but these 10% of Hispanic students. Pacific Islander stu-
resources are less readily available to other minor- dents had the highest rate of enrollment in pri-
ity students (Zhou and Kim 2006). vate, for-profit schools (19%), followed by Black
students (15%), American Indian/Alaska Native
5.5 P ostsecondary Enrollment students (10%), Hispanic students (9%), White
and Completion students (6%), and Asian students (4%) (Musu-­
Gillette et al. 2016).
5.5.1 P ostsecondary Enrollment
In 2007–08, nearly one quarter of undergradu-
The immediate college enrollment rate, or the per- ates had at least one immigrant parent. For some
centage of graduating high school students groups, immigrant generational status is espe-
enrolled in 2- or 4-year colleges the following fall, cially salient to their postsecondary experiences.
was approximately 68% in 2014. Asian students For example, among Asian college students,
had the highest immediate enrollment rate (85%), more than half (55%) were foreign-born and
followed by Whites (68%), Blacks (63%), and another 38% had at least one immigrant parent.
Hispanics (62%). The college participation rate is Among Hispanic college students, 21% were
foreign-born and 45% had at least one immigrant
parent. Enrollment patterns among first and

118 P. Ho and G. Kao

second generation immigrant Hispanic college uation rates. Asian students had the highest
students were comparable—for both groups, 6-year graduation rate (71%), followed by Whites
51% were enrolled in community college, 36% in (63%), Hispanics (53%), Pacific Islanders (50%),
nonprofit 4-year schools, and 12% in for-profit Blacks (41%), and American Indian/Alaska
schools. Among foreign-born Asian college stu- Native students (41%) (Snyder et  al. 2016).
dents, 54% were enrolled in community colleges Another measure of college attainment is the per-
and 38% in nonprofit 4-year schools, compared centage of adults over the age of 25 who have a
to 40% and 55%, respectively, of second postsecondary degree. In 2013, about 30% of
generation Asian college students. About 7% of adults had a bachelor’s degree or higher. Among
foreign-born and 5% of U.S.-born Asian college Asians, 52% earned a bachelor’s or higher, fol-
students were enrolled in for-profit schools lowed by Whites (33%), Blacks (19%), Pacific
(Staklis and Horn 2012). Islanders (16%), American Indian/Alaska Natives
(15%), and Hispanics (14%). The broad catego-
The type of institution students attend matters ries of Hispanic and Asian mask considerable
for their graduation rates—when comparing sim- variation by sub-groups. For example, 32% of
ilar students attending differently selective insti- South Americans and 25% of Cubans are college
tutions, researchers found that minority students graduates compared to 10% of Mexicans and 8%
have a higher likelihood of graduating if they of Salvadorans. Among Asian sub-groups, 73%
attend a more selective institution (Alon and of Asian Indian and 52% of Chinese adults have
Tienda 2005). Some research has shown that a college degree compared to 28% of Vietnamese
Black and Hispanic applicants to highly selective adults (Musu-Gillette et al. 2016).
schools receive an admissions advantage in terms
of their ACT/SAT scores (though Asians do not) In 2008, the percentage of U.S.-born adults
(Espenshade and Radford 2009). However, high over the age of 25 with at least a bachelor’s degree
schools vary in the amount of support they pro- was about 28% and 24% for the foreign-­born.
vide to students to help them navigate the transi- Among Hispanics, about 13% of the U.S.-born
tion to postsecondary schooling, which may and 12% of the foreign-born earned a college
result in underrepresented minority students degree. U.S.- and foreign-born Asians students
applying to less selective schools than they are also had comparable rates of college degree
actually qualified for (Roderick et al. 2011). The attainment overall (50% and 49%, respectively).
concentration of immigrant students in commu- Though there are considerable variations in col-
nity colleges is also an area of ongoing research lege degree attainment among both U.S.-born and
concern, including issues of access, affordability, foreign-born Hispanic and Asian sub-g­roups,
and language learning (Teranishi et al. 2011). within sub-groups rates of college degree attain-
ment by nativity are similar. For example, 10% of
5.5.2 P ostsecondary Completion U.S.-born and 9% of foreign-born Hondurans
earned a college degree, and about 50% of U.S.-
For students attending a 4-year college full-time born and 51% of foreign-born Korean adults are
for the first time in 2006, the average graduation college graduates (Kao et al. 2013).
rate after 4 years was 39%. About 46% of Asian
students and 43% of White students graduated 5.6 T he Importance of Race,
within 4 years. Hispanic students had an average Ethnicity, and Nativity
4-year graduation rate of 29%, and for Pacific
Islander, American Indian/Alaska Native, and At every level of education and across multiple
Black students, the corresponding rates were educational outcomes, patterns of racial and eth-
24%, 22%, and 21%. Not surprisingly, 6-year nic stratification are apparent. In general, Black,
graduation rates are higher overall (60%) and for Hispanic, and American Indian/Alaska Native
all racial/ethnic groups compared to 4-year grad- students experience poorer educational outcomes

5  Educational Achievement and Attainment Differences Among Minorities and Immigrants 119

relative to more advantaged groups such as White their ethnic identity and immigrant origins, dis-

and Asian students. Students identified as English tancing themselves from native Blacks. These

Language Learners (ELL) on average also fare students believed that doing well in school would

worse than non-ELL students, although racial pay off. Poorer second generation youths tended

and ethnic differences among ELL students typi- to identify with native Black peers and believed

cally, though not always, mirror those found they would have limited opportunities for upward

among non-ELL students. In this section, we mobility and did not do as well in school. Matute-­

describe how these racial, ethnic, and immigrant Bianchi (1986) found similar patterns among

differences in educational outcomes fit into the Mexican-descent students in central California—

larger debates around racial relations in the academically successful first and second genera-

U.S. We also highlight some of the issues that set tion students used their immigrant and ethnic

children of immigrants apart from their peers culture to distinguish themselves from less aca-

with native-born parents. demically successful Chicanos and “cholos.” In

Scholars envision various ways in which the contrast to the negative stereotypes about Black

U.S. racial and ethnic hierarchy may shift due to and Hispanic students’ academic abilities, the

demographic changes, including the growing size general academic success of Asian students has

and diversity of the immigrant population. Some led to the “model minority” stereotype that paints

scholars believe that “[c]hildren of Asian, black, all Asian students as naturally high-achieving.

mulatto, and mestizo immigrants cannot escape However, the stereotype can be harmful to Asian

their ethnicity and race, as defined by the main- groups that do not fare as well academically

stream” and that discrimination will likely affect because their struggles may be overlooked in

these students’ academic performance (Portes schools (Lee 2005; Ngo and Lee 2007; Teranishi

et al. 2005). Others argue that boundaries between 2010), and also contributes to perceptions of

Whites and Asian and Latino groups are more Asian students as overly competitive academi-

likely to erode over time than Black–White lines cally and less well-rounded (Jiménez and

(Lee and Bean 2010), suggesting more positive Horowitz 2013; Kao 1995; Oakes and Guiton

outcomes for non-Black minorities. Still others 1995; Ochoa 2013).

believe that a tri-racial hierarchy is more likely— In addition to their experiences with the racial

with lighter-skinned minorities (such as East and ethnic hierarchy of the U.S., children of

Asians and White Latinos) earning “honorary immigrants are also affected by generational sta-

White” status and darker-skinned minorities tus. The proportions of first, second, and third

forming a disadvantaged “collective Black” generation and higher varies considerably across

group (Bonilla-Silva 2004). groups. Among Hispanic youth, about 6% are

How the minority children of immigrant par- first-generation, 51% second generation, and

ents adapt to the U.S. racial and ethnic hierarchy 42% third generation or higher. For Asian youth,

is important for understanding their educational the corresponding estimates are 13%, 65%, and

outcomes (Kao et al. 2013). Some research sug- 20%; for Black youth 2%, 12%, and 86%; for

gests that academically successful first and sec- White youth less than 1%, 7%, and 92%. These

ond generation minority youth assert a more generational differences matter for student out-

“traditional” identity that they contrast with the comes. Among first-generation youth, the age of

“Americanized” values of their less successful arrival matters for language acquisition and

co-ethnics (Lee 2005; Louie 2012; Matute-­ socialization (Rumbaut 2004). Research is mixed

Bianchi 1986; Waters 1994). In interviews with on whether the first or second generation immi-

West Indian and Haitian youths in New  York, grants experience better educational outcomes

Waters (1994) found that although second gen- (Baum and Flores 2011; Coll and Marks 2012;

eration youth all realized they were likely to be Crosnoe and Turley 2011; Duong et  al. 2015;

perceived as native Blacks by others, those from Kao and Tienda 1995; White and Glick 2009). An

middle-class backgrounds tended to emphasize ongoing research concern is the notion of “immi-