Shifting Tides: The Evolution of Racial Inequality in Higher Education from the 1980s through the 2010s

Abstract

Amid the proliferation of state-level bans on race-based affirmative action in higher education, the U.S. Supreme Court’s decision on June 29, 2023, dismantled race-conscious college admission policies, intensifying concerns about the persistence and potential increase of racial inequality in higher education. The authors analyze four restricted-use national survey datasets to investigate racial disparities in college attendance outcomes from the 1980s through the 2010s. Although college entrance rates increased for all racial groups, Black and Hispanic youth became increasingly less likely than their White peers to attend four-year selective colleges. In the 2010s cohort, Black and Hispanic youth were 8 and 7 percentage points, respectively, less likely than their White counterparts to secure admission to four-year selective colleges, even after controlling for parents’ income, education, and other family background variables. The findings underscore the urgent need for proactive policy interventions to address the widening racial inequality in attending selective postsecondary institutions.

Keywords

race higher education affirmative action college selectivity effectively maintained inequality

Race-based affirmative action in college admission was established to increase the presence of racial minorities historically underrepresented in higher education (Garrison-Wade and Lewis 2004; Hirschman and Berrey 2017). Since its inception in the 1960s, the legitimacy of race-based affirmative action in college admissions has been a subject of heated controversy (Long and Bateman 2020). The Regents of the University of California v. Bakke (1978) and Grutter v. Bollinger (2003) rulings of the Supreme Court recognized the importance of race-conscious admission processes for fostering diversity and inclusion in higher education. However, there has been a lack of consistency in subsequent Supreme Court decisions regarding the support for using race as a factor in college admission. In 2023, the Supreme Court dismantled race-based affirmative action in a case involving Harvard University and the University of North Carolina, through Students for Fair Admissions Inc. v. President & Fellows of Harvard College (2023). Similarly, beginning in the 1990s, an increasing number of state governments have enacted bans on race-conscious college admission (Hirschman and Berrey 2017; Long and Bateman 2020).

Although race-conscious college admission policies were struck down, a multitude of studies found evidence pointing to the endurance, and potential exacerbation, of racial inequality in higher education (Baker, Klasik, and Reardon 2018; Kehal, Hirschman, and Berrey 2021; Long and Bateman 2020; Reardon, Baker, and Klasik 2012; Voss, Hout, and George 2022). White and Asian students have demonstrated higher representation within higher education, whereas racial minority groups, including Black, Hispanic,¹ and Native American students, continue to face heightened structural barriers when seeking admission to colleges, particularly to four-year selective institutions (Espinosa et al. 2019; Irwin et al. 2022). Unfortunately, given the increased socioeconomic importance of college degrees, the underrepresentation of these minority groups within higher education will contribute to perpetuating racial disparities in occupational and other socioeconomic outcomes (Baker et al. 2018; Goldin and Katz 2009; Tamborini, Kim, and Sakamoto 2015).

Supporters of race-conscious college admission bans have often insisted that race-neutral alternative measures, such as affirmative action based on family socioeconomic status (SES), could serve as effective substitutes for race-based affirmative action (Kane 2003; Potter 2014; Reardon et al. 2018). The fact that 27 percent of Hispanic and 31 percent of Black youth lived in poverty in 2016, in contrast to 10 percent of White and Asian youth (Wilson and Schieder 2018), indicates that the diminished levels of K–12 academic achievement among racial minority children are, in part, linked to their family SES. However, the question remains whether SES-based affirmative action in college admission can yield a commensurate level of racial diversity achieved by race-based affirmative action, given the historical racism deeply seated within the United States (Du Bois 1903; Feagin and Ducey 2018; Fredrickson 2015; Kane 2003).

In this research, we use four restricted-use longitudinal datasets from the National Center for Education Statistics (NCES) to track racial disparities in whether and where young adults enrolled at the postsecondary level from the 1980s through the 2010s. Although earlier studies included some of the datasets used in this study (Grodsky 2007; Reardon et al. 2012; Voss et al. 2022), data from the 2010s were often omitted, leaving the most recent trends veiled. Importantly, our analyses indeed demonstrate that racial gaps in attending four-year selective colleges have widened, with the most pronounced disparities emerging in the 2010s. Black and Hispanic students continue to face inequities in higher education. Furthermore, our analyses reveal that the incorporation of controls for parents’ income, education, and other family background variables rarely alters the trends of racial inequality in higher education. The implications of these findings for the ongoing policy debate on affirmative action bans in higher education are discussed.

Literature Review

Policymakers and researchers have recognized that educational expansion fuels economic growth in the United States (Fischer and Hout 2006; Goldin and Katz 2009). Besides the economic benefits, democratic social efforts to provide equal educational opportunities for all children have driven rapid expansion in higher education in the United States, often marked as the “college for all” movement (Goyette 2008). However, it is unclear whether educational expansion guarantees a reduction in racial inequality in higher education.

The maximally maintained inequality (MMI) (Raftery and Hout 1993) and effectively maintained inequality (EMI) perspectives (Lucas 2001) provide seminal sociological insight into the relationship between educational expansion and inequality (Alon 2009; Arum, Gamoran, and Shavit 2007). The MMI perspective suggests that when access to college entrance is expanded, privileged groups maintain their advantage by disproportionately occupying the additional capacity generated by the expansion (Raftery and Hout 1993). Once privileged groups reach a saturation point in terms of college entrance, educational inequality can be maintained through differences in the prestige of the institutions they attend. The EMI perspective proposes that privileged groups maintain their advantage by disproportionately attending selective colleges (Lucas 2001).

Both the MMI and EMI perspectives underscore that educational expansion does not necessarily reduce educational inequality. Although these perspectives are generally applied to studies into educational inequality on the basis of family SES (Alon 2009; Oh and Kim 2020; Torche 2011), these may also be relevant to race-based educational inequality (Baker et al. 2018; Espinosa et al. 2019). That is to say, the expansion of higher education might not have alleviated Black and Hispanic students’ disadvantages in college entrance (i.e., MMI) and/or might have provoked institutional changes growing White and Asian students’ advantages in attending selective colleges (i.e., EMI).

Alon (2009) suggested that EMI in higher education occurs through two mechanisms: adaptation and exclusion. Although privileged groups cultivate their children’s academic currencies through K–12 to meet the evolving admission criteria (i.e., adaptation), they also contribute to changing the contours of the admission criteria favoring their children (i.e., exclusion). Consistent with this theoretical speculation on exclusion, Baker (2019) documented that the spread of state-level bans on race-conscious college admission policies might have been provoked by the scarcity of access to selective colleges for White students. Certainly, considering the important role of race-based affirmative action in the history of the United States (Bowen and Bok 1998), the bans on such policies can play a role in limiting the benefits racial minority children can obtain from the expansion of higher education.

After centuries-old slavery and racism in the United States, the foundation of race-based affirmative action in higher education was established in the civil rights era (Garrison-Wade and Lewis 2004). Race began to be considered a factor in the admission process to remedy past racial discrimination and equalize admission requirements in the 1960s (Hirschman and Berrey 2017). Previous studies have documented that race-based affirmative action in higher education contributes to improving the presence of historically underrepresented racial minorities in higher education (Bowen and Bok 1998; Garrison-Wade and Lewis 2004). However, since their inception, race-conscious college admission policies have sparked furious debates on its legality and justice, coupled with a series of important decisions of the Supreme Court (Long and Bateman 2020).

In its Regents of the University of California v. Bakke ruling in 1978, the Court acknowledged the constitutionality of affirmative action, allowing race to be considered as one factor in admissions decisions. Similarly, the Grutter v. Bollinger ruling in 2003 upheld the use of race in a holistic admissions process. Conversely, starting in California, Texas, Washington, and Florida in the 1990s, a number of states prohibited their universities from using race as a factor in admission decisions (Hirschman and Berrey 2017; Long and Bateman 2020). The Gratz v. Bollinger (2003) ruling struck down a mechanistic points system in undergraduate admissions. The Fisher v. University of Texas at Austin (2013, 2016) rulings underscored that race-conscious admissions policy was permissible as long as it met strict scrutiny standards. The Schuette v. Coalition to Defend Affirmative Action (2014) ruling supported the ban on public colleges and universities’ using race as a factor in admissions. On June 29, 2023, the Supreme Court issued a 6–3 verdict to dismantle race-based college admission processes in a case involving Harvard University and the University of North Carolina. In 2014, only 35 percent of about 1,000 selective institutions reported considering race in college admission (Hirschman and Berrey 2017).

Unfortunately, although race-based affirmative action in college admission has been withdrawn, researchers have discovered that racial minorities continue to be underrepresented in higher education both in terms of admissions and persistence (Baker et al. 2018; Kehal et al. 2021; Long and Bateman 2020; Reardon et al. 2012; Voss et al. 2022). This fact underscores the pressing need for reinstating race-conscious college admission policies or providing alternative policies to substitute them. However, previous research has not found that alternative policies are as effective as race-conscious college admission policies in promoting racial diversity in higher education (Byrd 2021; Long and Bateman 2020; Reardon et al. 2018).

Some opponents of race-based affirmative action in higher education have insisted that such policies can be replaced by SES-based affirmative action (Kane 2003; Potter 2014; Reardon et al. 2018). Racial disparities in K–12 educational outcomes are certainly associated with differences in family SES (Downey 2008; Hattie 2008; Kao and Thompson 2003), indicating that SES-based affirmative action has positive effects, to some extent, on enhancing racial diversity in higher education. However, it should be noted that the educational barriers faced by racial minority children are rooted not only in their disadvantaged family background but also in institutional and direct racism deeply ingrained in the United States (Bloome 2014; Feagin and Ducey 2018; Massey et al. 2011). For instance, there are systematic racial differences in the educational quality of K–12 schools and students’ experiences within those schools (Clotfelter, Ladd, and Vigdor 2005; Condron and Roscigno 2003; Irwin et al. 2022; Logan, Minca, and Adar 2012; Orfield et al. 2014). After decades of social conflict and shifting legal rulings, schools resegregated, reaching 1960s levels of segregation by the mid-1990s (Fiel and Zhang 2019), with the share of predominantly Black or Hispanic schools even increasing since the early 2000s (U.S. Government Accountability Office 2016). Individual disadvantages are compounded in predominantly Black or Hispanic schools because of heightened negative impacts on learning from extraschool inequality, as well as reduced access to resources and dominant forms of capital within these schools (Noguera 2003; Rothstein 2004; Shifrer 2022). White and Asian youth, in contrast, have more access to “high-growth” schools (Hanselman and Fiel 2017). Black and Hispanic youth’s college attendance outcomes may also be uniquely shaped by their disproportionate experiences with exclusionary discipline in school and with the criminal justice system (Wang 2018; Western and Wildeman 2009).

To contribute to the ongoing debate surrounding race-based affirmative action in higher education, this study aims to elucidate the trajectory of racial disparities in whether and where young adults enrolled at the postsecondary level from the 1980s through the 2010s. This observed period spans the decade preceding the emergence of state-level affirmative action bans in the 1990s and the subsequent three decades. Although the data do not allow us to conclusively demonstrate a causal relationship between the trends we observe in racial patterns in college admissions and policy revisions, it is crucial to monitor and document changes in college admission outcomes over time to assess progress toward racial equity.

In addition to the three datasets from the 1980s to the 2000s, covered by prior studies, our research is the first to incorporate the most recent dataset from the 2010s. This addition facilitates updated insights into the evolution of racial inequality in higher education. Furthermore, we consider a comprehensive range of family SES variables, including parents’ income and education, to explore whether the racial patterns in college admissions we found are altered by controlling for related racial differences in family SES.

Analytic Strategy

Data and Sample

In this study, we use four restricted-use datasets: High School and Beyond, the National Education Longitudinal Study of 1988, the Education Longitudinal Study of 2002, and High School Longitudinal Study of 2009. The nationally representative longitudinal surveys each followed a cohort of high school students for nine or more years to examine their high school and postsecondary experiences, respectively representing the cohorts graduating high school in 1982, 1992, 2004, and 2013. We include respondents who graduated or dropped out of high school as of November in the year most of the cohort graduated high school (i.e., excluding those continuing on in high school beyond four years). Results from sensitivity analyses excluding the small number of respondents who dropped out of high school do not alter our main findings at all. The unweighted sample sizes are about 8,900 for High School and Beyond, 9,590 for the National Education Longitudinal Study of 1988, 10,350 for the Education Longitudinal Study of 2002, and 11,240 for the High School Longitudinal Study of 2009. Note that NCES requires that all unweighted frequencies be rounded to the nearest 10. Missing values on independent variables are handled using multiple imputation by chained equations (von Hippel 2007; White, Royston, and Wood 2011).

Statistical Models

We first estimate multinomial logistic models for each cohort, as our dependent variable (i.e., college attendance outcomes) is categorical. As the multinomial logit coefficients are hardly interpretable and should not be compared across different models, we subsequently calculate postestimate predicted probabilities to show disparities in college attendance outcomes relative to White young adults within each cohort, and then differences in those disparities for all later cohorts relative to the 1982 cohort.

We construct a four-category dependent variable that combines consideration of college attendance status and selectivity of the institution attended as of November in the year when most of the cohort graduated high school: no college, two-year college, four-year nonselective college, and four-year selective college. The four NCES surveys asked respondents to report their college attendance status at different time points, but they also collected information on the college enrollment date or the enrollment status by month. Using this information, we create a new variable indicating college attendance as of November in the year when most of the cohort graduated high school, which was the earliest month we could consistently identify across the four cohorts (i.e., November in 1982, 1992, 2004, or 2013). As most respondents graduated from high school in May or June and entered college in August or September, the likelihood of students transferring between high school graduation and November of the same year is low. However, we acknowledge that this methodological choice has limitations and may not capture all college attendance patterns.

On the basis of Barron’s Profiles of American Colleges, we classify noncompetitive, less competitive, and competitive institutions as “nonselective” and very competitive, highly competitive, and most competitive as “selective.” In preliminary analyses using all six selectivity levels, patterns were similar for the institutions in the three lower selectivity levels, just as patterns were similar for those in the three higher selectivity levels. Thus, in the main analyses, we use a simplified version of this variable to present more concise results.

The race variable has four categories: White (reference), Black, Hispanic, and Asian. To examine how the inclusion of the controls for family SES alters the patterns of racial disparities we observe, we account for family income percentile (13 categories), parents’ educational level percentile (less than high school, high school, some college, bachelor’s degree, or advanced degree), high school location (urban, suburban, or rural), and high school type (private or public). We also control for gender (male or female). Sensitivity analyses, when conducted separately for men and women, do not result in a substantial variation in the findings presented in this article. All covariates are centered at the mean. Also, considering the survey design, we weighted and adjusted all estimations to be representative of each population.

Results

Table 1 shows descriptive statistics of the analytic variables used in this study. The proportion of young adults who were not enrolled in college dropped from 0.51 to 0.23 between 1982 and 2013, indicating the rapid expansion of higher education during the observed period. Accordingly, the proportions of students who attended two-year and four-year selective colleges grew from 0.18 to 0.30 and from 0.07 to 0.21, respectively. However, the proportions of those who attended four-year nonselective colleges were around 0.25 across cohorts.

Table 1.

Descriptive Statistics.

	1982	1992	2004	2013
College attendance outcomes
Not enrolled	.51	.38	.31	.23
Two-year	.18	.22	.27	.30
Four-year nonselective	.24	.26	.24	.26
Four-year selective	.07	.13	.18	.21
Race
White	.80	.76	.68	.61
Black	.12	.10	.14	.13
Hispanic	.07	.09	.14	.22
Asian	.02	.04	.04	.05
Covariates
Family income (percentile mean)	.44	.44	.44	.47
Parents’ education (percentile mean)	.38	.39	.39	.37
High school location
Urban	.19	.26	.29	.31
Suburban	.49	.42	.51	.45
Rural	.31	.32	.20	.24
Public high school	.90	.91	.92	.91
Woman	.52	.51	.54	.50
n	8,900	9,590	10,350	11,240

Source: U.S. Department of Education, National Center for Education Statistics, High School and Beyond (1980–1982), National Education Longitudinal Study of 1988 (1988–1992), Education Longitudinal Study of 2002 (2002–2004), and High School Longitudinal Study of 2009 (2009–2013).

Note: Means and proportions are weighted and adjusted for survey design and to be representative of the population. Unweighted frequencies are rounded to the nearest 10, as required by the National Center for Education Statistics. “Nonselective” includes noncompetitive, less competitive, and competitive institutions, while “selective” includes very competitive, highly competitive, and most competitive institutions, as recorded in Barron’s Profiles of American Colleges.

One potential scenario is that the aforementioned overall trend might not vary on the basis of race, indicating that the expansion of higher education might result in an equitable advantage for different racial groups. On the other hand, in line with the MMI and EMI perspectives, Black and Hispanic students might have consistently exhibited low probabilities of enrolling in college, with an escalating disadvantage in securing admission to four-year selective institutions. The next phase of this study involves investigating these two competing possibilities through regression models, both without and with the inclusion of control variables. The bottom section of Table 1 presents descriptive statistics for the control variables.

Figure 1 illustrates the predicted probabilities of college attendance outcomes categorized by cohort and race. The full multinomial models are provided in Table A1 in the Appendix. The unadjusted models present estimations without controls, whereas the adjusted models incorporate controls. Despite the broad expansion of education, the unadjusted model indicates that the racial disparities in the likelihood of being unenrolled have rarely diminished over time, in line with the MMI perspective. The adjusted model shows insignificant White-Black and White-Hispanic gaps, suggesting that differences in SES may be important in explaining racial differences in terms of college entry.

Figure 1.

Predicted probabilities of college attendance by race.

However, of greater importance, while the probability of enrolling in two- or four-year nonselective colleges increased over time for Black and Hispanic young adults, the underrepresentation of Black and Hispanic students in four-year selective colleges became more conspicuous, aligning with the EMI perspective. Despite the inclusion of family SES and other covariates, which moderately mitigate racial disparities in attending four-year selective colleges, the trend of widening racial gaps in the probability of securing admission to four-year selective colleges over time is evident in both unadjusted and adjusted models.

Furthermore, we examine the statistical significance of the observed racial disparities and any changes in these disparities. Table 2 presents the racial gaps in college attendance within each cohort, displaying predicted probabilities for White young adults and the differences in predicted probabilities (i.e., marginal effects) relative to White adults for Black, Hispanic, and Asian adults. As an illustration, for the 1982 cohort in the unadjusted models, the predicted probability of attending a two-year college was 18 percent for White youth and two percentage points lower for Black youth than for White youth. Although the controls for family SES and other covariates do influence the size and significance of some marginal effects, both unadjusted and adjusted models yield similar results.

Table 2.

Racial Differences in College Attendance within Each Cohort (Base Category: Not Enrolled).

	Unadjusted Models				Adjusted Models
	1982	1992	2004	2013	1982	1992	2004	2013
Outcome: two-year college
White (reference)	.18 (.01)***	.21 (.01)***	.26 (.01)***	.26 (.01)***	.18 (.01)***	.21 (.01)***	.26 (.01)***	.28 (.01)***
Black	−.02 (.02)	−.04 (.03)	.00 (.02)	.06 (.02)*	−.01 (.02)	−.04 (.02)	−.01 (.02)	.02 (.02)
Hispanic	.01 (.02)	.09 (.03)***	.08 (.02)***	.15 (.03)***	.04 (.02)	.08 (.03)**	.06 (.02)**	.10 (.02)***
Asian	.10 (.05)*	.05 (.03)	.00 (.02)	.01 (.03)	.10 (.05)*	.07 (.03)*	.00 (.02)	.02 (.03)
Outcome: four-year nonselective college
White (reference)	.18 (.01)***	.21 (.01)***	.26 (.01)***	.26 (.01)***	.18 (.01)***	.21 (.01)***	.26 (.01)***	.28 (.01)***
Black	−.04 (.02)*	−.01 (.03)	.03 (.02)	.03 (.02)	.02 (.02)	.02 (.03)	.07 (.02)***	.07 (.02)**
Hispanic	−.12 (.01)***	−.11 (.02)***	−.12 (.01)***	−.08 (.02)***	−.09 (.02)***	−.06 (.02)*	−.09 (.02)***	−.04 (.02)
Asian	.01 (.03)	−.04 (.03)	−.04 (.03)	−.02 (.03)	.01 (.03)	−.04 (.03)	−.03 (.03)	.00 (.03)
Outcome: four-year selective college
White (reference)	.08 (.01)***	.14 (.01)***	.21 (.01)***	.25 (.01)***	.07 (.00)***	.14 (.01)***	.19 (.01)***	.22 (.01)***
Black	−.04 (.01)***	−.08 (.01)***	−.15 (.01)***	−.15 (.02)***	−.02 (.01)	−.05 (.02)***	−.11 (.01)***	−.08 (.02)***
Hispanic	−.04 (.01)***	−.07 (.01)***	−.13 (.01)***	−.14 (.02)***	−.01 (.01)	−.03 (.02)	−.06 (.01)***	−.06 (.02)***
Asian	.11 (.03)***	.13 (.03)***	.12 (.03)***	.14 (.03)***	.08 (.02)***	.09 (.03)***	.13 (.03)***	.12 (.03)***

Note: We provide predicted probabilities for White young adults and the difference relative to White adults for Black, Hispanic, and Asian adults. Numbers in parentheses are standard errors. Estimations are weighted and adjusted for survey design and to be representative of each population. Adjusted models include controls for family income, parents’ education, high school location, public high school, and gender. All covariates are centered at their means. “Nonselective” includes noncompetitive, less competitive, and competitive institutions, while “selective” includes very competitive, highly competitive, and most competitive institutions, as recorded in Barron’s Profiles of American Colleges.

p < .05, **p < .01, and ***p < .001 (two-tailed tests).

In the most consistent pattern, the predicted probability of attending a four-year selective college remains significantly lower for Black and Hispanic youth compared with White youth in the two youngest cohorts, even in adjusted models. Specifically, according to the unadjusted model, in the 1982 cohort, Black and Hispanic youth were four percentage points less likely to enter four-year selective colleges. The gaps between White-Black and White-Hispanic students in attending such institutions grew to 15 percent and 14 percent, respectively, in the 2013 cohort. The adjusted models indicate that the inclusion of controls for family SES mitigated the disadvantages faced by Black and Hispanic youth in entering four-year selective colleges, reducing them to statistically nonsignificant levels in the 1982 cohort. However, even with the inclusion of controls, the White-Black and White-Hispanic gaps in entering four-year selective colleges remained statistically significant in the 2013 cohort. In this cohort, Black and Hispanic youth were still eight and seven percentage points, respectively, less likely than their White counterparts to secure admission to four-year selective colleges.

Moreover, Table 3 presents variations in the extent of racial disparities in college attendance outcomes across later cohorts in comparison with the 1982 cohort. For instance, in the unadjusted models, the White-Black gap in the predicted probability of attending a two-year college was two percentage points wider in the 1992 cohort than in the 1982 cohort; however, this change is not statistically significant. While the unadjusted models indicate a significant decrease in the relative disadvantages faced by Black and Hispanic young adults in attending two-year or four-year nonselective colleges compared with their White counterparts, these differences in disparities lose significance in the adjusted models. Importantly, the racial gaps in the predicted probabilities of attending a four-year selective college widened significantly. From 1982 to 2013, the White-Black and White-Hispanic gaps in entering such institutions increased by about 11 percentage points without controlling for family SES. Even with the controls, the White-Black and White-Hispanic gaps increased by seven and five percentage points, respectively. The results are consistent with the EMI perspective and also suggest that only a portion of racial disparities in attending selective institutions can be attributed to differences in family SES. Our supplementary analyses on gender differences (available upon request) indicate that the MMI and EMI trends are observable for both men and women.

Table 3.

Differences in Racial Disparities in College Attendance in the 1992, 2004, and 2013 Cohorts Relative to the 1982 Cohort.

	Unadjusted Models			Adjusted Models
	1992	2004	2013	1992	2004	2013
Outcome: two-year college
Black	−.02 (.03)	.02 (.02)	.08 (.03)**	−.03 (.03)	−.01 (.03)	.02 (.03)
Hispanic	.08 (.03)*	.07 (.03)*	.14 (.03)***	.05 (.03)	.03 (.03)	.06 (.03)
Asian	−.04 (.05)	−.10 (.05)	−.08 (.05)	−.03 (.06)	−.10 (.05)	−.08 (.06)
Outcome: four-year nonselective college
Black	.03 (.03)	.07 (.03)**	.07 (.03)**	.01 (.03)	.05 (.03)	.05 (.03)
Hispanic	.02 (.03)	.00 (.02)	.05 (.02)*	.02 (.03)	−.01 (.02)	.05 (.03)
Asian	−.05 (.04)	−.05 (.04)	−.02 (.04)	−.05 (.04)	−.04 (.04)	−.01 (.04)
Outcome: four-year selective college
Black	−.04 (.02)*	−.11 (.01)***	−.11 (.02)***	−.04 (.02)	−.09 (.02)***	−.07 (.02)***
Hispanic	−.03 (.02)	−.09 (.02)***	−.11 (.02)***	−.01 (.02)	−.04 (.02)*	−.05 (.02)*
Asian	.02 (.04)	.01 (.04)	.03 (.04)	.01 (.03)	.05 (.03)	.04 (.03)

Note: The base outcome category is not enrolled. Numbers in parentheses are standard errors. Estimations are weighted and adjusted for survey design and to be representative of each population. Adjusted models include controls for family income, parents’ education, high school location, public high school, and gender. All covariates are centered at their means. “Nonselective” includes noncompetitive, less competitive, and competitive institutions, while “selective” includes very competitive, highly competitive, and most competitive institutions, as recorded in Barron’s Profiles of American Colleges.

p < .05, **p < .01, and ***p < .001 (two-tailed tests).

Limitations

We acknowledge that certain unobserved variables (e.g., family wealth) could potentially revise our findings, although the restricted-use datasets used in this study provide more comprehensive measures of family SES compared with those found in most national datasets. It should also be noted that controlling for racial differences in family SES is an insufficient method to evaluate the effectiveness of SES-based affirmative action as a replacement for race-based affirmative action (Kane 2003). Furthermore, our delineation of racial inequality relies on the four racial categories within the datasets. It is important to note that racial categories are a social construct, and there may be shifts in how respondents identify with these categories (Byrd 2021; Teranishi et al. 2020). We also were unable to incorporate some racial groups, such as Native Americans, due to their limited sample size.

Additionally, we recognize that certain young adults might enter college or transfer later in their adult lives. However, incorporating data from the four national datasets required a focus on young adulthood. By focusing on those who enroll in college soon after high school, we capture a more expansive view of racial inequality because those who enroll in college later in adulthood likely experience fewer educational and occupational benefits (Oh and Kim 2020; Taniguchi 2005).

Finally, although data limitations prevent us from establishing causal relationships between race-based affirmative action bans and the observed changes over cohorts, this research documents a significant shift in racial disparities in college admission outcomes from the 1980s through the 2010s. Black and Hispanic students are becoming increasingly less likely to be admitted to selective colleges compared with their White or Asian peers. In the following section, we further discuss the implications of our findings.

Discussion

Motivated by the ongoing controversy surrounding the proliferation of race-affirmative action bans in higher education, this study delineates the evolution of racial disparities in college attendance outcomes between the 1980s and 2010s. Although the college enrollment rate increased for all racial groups, the expansion of education hardly changed the existing racial gaps in college entrance, aligning with the MMI perspective. More notably, the gaps between White-Black and White-Hispanic students in attending selective four-year colleges have widened, providing support for the EMI perspective. The results suggest that the MMI and EMI perspectives assist in comprehending not only educational inequalities on the basis of family SES but also those rooted in race (Baker et al. 2018; Espinosa et al. 2019).

Corroborating earlier findings on the growing racial inequality at the postsecondary level, especially within selective institutions (Baker et al. 2018; Kehal et al. 2021; Long and Bateman 2020; Reardon et al. 2012), this research offers timely updates on the evolving nature of racial inequality in higher education by integrating the latest dataset from the 2010s. The educational stratification based on race, directing White and Asian youth toward four-year selective colleges while steering Black and Hispanic youth toward two-year or nonselective four-year colleges, became even more pronounced in the 2010s. Furthermore, the observed trend is not substantially altered by controlling for students’ family SES. The inclusion of these controls partially mitigates racial disparities in college attendance outcomes, particularly in the 1980s and the 1990s. However, even when adjusting for all the covariates observed in this study, significant racial disparities in securing access to four-year selective colleges persisted in the 2000s and the 2010s. More importantly, the incorporation of family SES controls rarely alters the trend of widening racial disparities in attending four-year selective colleges. The recent decision of the Supreme Court in 2023 is likely to contribute to maintaining, if not further exacerbating, the dearth of racial diversity in selective postsecondary institutions.

Some supporters of race-based affirmative action bans insist that other avenues still remain open for postsecondary institutions to address racial inequality in higher education. Colleges are encouraged to account for differences in the prestige of students’ precollege educational institutions or differences in their resources for learning (Akos et al. 2023; Fletcher and Tienda 2009). Given the high degree of segregation by race in neighborhoods and K–12 schools, and the profound implications of this segregation for educational, occupational, and health outcomes (Noguera 2003; Rothstein 2004; Shifrer 2022), the applicants’ former educational institution may be one effective proxy for race in college admissions. Similarly, a more holistic approach to admissions—one that considers an applicant’s achievements in the context of their background and circumstances, rather than relying on metrics such as standardized test scores—is suggested as an avenue for addressing inequity in admission to selective colleges (Akos et al. 2023; Coleman and Keith 2018; Jelks 2010).

In addition, it is imperative to narrow racial disparities in educational outcomes at the K–12 and earlier levels (Condron and Roscigno 2003; Logan et al. 2012; Rothstein 2004). As long as race remains a determinant of children’s opportunities to develop academic capacities or credentials, the legitimacy of considering race in college admissions to remedy the failures of the precollege education system will continue to be controversial. A fundamental policy reform for enhancing racial diversity in higher education must commence prior to the college admission process.

Unfortunately, however, as suggested by the MMI and EMI perspectives (Lucas 2001; Raftery and Hout 1993), racial inequality in higher education is likely to persist, particularly in the absence of race-conscious policy interventions. Historically disadvantaged groups are prone to exclusion in college admission processes, whereas advantaged groups tend to sustain their advantages through swift adaptation to new education policies (Alon 2009). To facilitate evidence-based policymaking for greater diversity and inclusion, future studies should continue monitoring the evolving nature of racial inequality in higher education. Additionally, it will be crucial to compare the effectiveness of race-conscious admission policies and their alternatives in enhancing racial diversity.

During the past several decades, the decisions made by the Supreme Court regarding the constitutionality of race-conscious college admission policies have shown inconsistency. However, it is apparent that the persistent educational inequality against racial minority children and youth is incongruent with the principles of equal rights protection for all citizens. The racial disparities observed in higher education are deeply connected to the historical racism embedded in the United States over many centuries, which is rarely altered without proactive policy interventions (Andersen and Collins 2020; Feagin and Ducey 2018).

Although comprehensive solutions have been absent, race-conscious college admission policies have served as a temporary measure to rectify historical shortcomings in the U.S. education system. Unfortunately, it has been abolished without compelling alternative strategies. Racial disparities in educational outcomes can be attributed, in part, to family SES. However, it is crucial to recognize that racial differences in family SES are also a result stemming from historical racism. Given the increasing importance of higher education in socioeconomic attainments throughout a lifetime, promoting equal access to higher education stands as one of the effective strategies to narrow racial gaps in family SES for future generations. This study emphasizes the urgency of proactive policy interventions to address the dearth of racial diversity in colleges, especially in selective postsecondary institutions.

Footnotes

Appendix

Table A1.

Log Odds from Multinominal Logit Estimations of Racial Differences from White Youth in College Attendance Outcomes (Base Category: Not Enrolled).

	1982	1992	2004	2013
Unadjusted models
Outcome category: two-year college
Race
Black	−.29 (.13)*	−.47 (.19)*	−.32 (.10)**	−.02 (.13)
Hispanic	−.22 (.13)	.14 (.14)	−.19 (.11)	.20 (.13)
Asian	1.01 (.26)***	.70 (.18)***	.34 (.19)	1.10 (.22)***
Constant	−1.00 (.05)***	−.56 (.06)***	−.07 (.05)	.19 (.05)***
Outcome category: four-year nonselective college
Race
Black	−.35 (.10)***	−.31 (.14)*	−.22 (.10)*	−.12 (.13)
Hispanic	−.95 (.13)***	−.69 (.16)***	−1.13 (.12)***	−.60 (.14)***
Asian	.60 (.21)**	.30 (.18)	.16 (.20)	.98 (.20)***
Constant	−.66 (.04)***	−.29 (.05)***	−.08 (.05)	.26 (.05)***
Outcome category: four-year selective college
Race
Black	−.98 (.22)***	−1.11 (.19)***	−1.54 (.15)***	−1.09 (.16)***
Hispanic	−.93 (.24)***	−.82 (.20)***	−1.36 (.13)***	−1.10 (.16)***
Asian	1.47 (.24)***	1.11 (.18)***	.79 (.17)***	1.48 (.20)***
Constant	−1.85 (.08)***	−.94 (.07)***	−.26 (.06)***	.18 (.06)**
Adjusted models
Outcome category: two-year college
Race
Black	−.04 (.13)	−.39 (.19)*	−.23 (.11)*	.08 (.13)
Hispanic	.02 (.13)	.30 (.15)*	−.07 (.12)	.27 (.11)*
Asian	1.03 (.26)***	.71 (.19)***	.43 (.20)*	1.09 (.21)***
Female	.46 (.08)***	.27 (.09)**	.18 (.07)*	.23 (.09)**
Family income (percentile)	.63 (.18)***	−.03 (.18)	.24 (.14)	.79 (.19)***
Parents’ education (percentile)	1.40 (.16)***	1.21 (.21)***	.46 (.14)***	.53 (.18)**
High school location (reference: urban)
Suburban	.18 (.13)	.12 (.12)	.08 (.11)	−.19 (.14)
Rural	−.05 (.14)	−.01 (.13)	.15 (.13)	−.40 (.14)**
Public high school	−.25 (.15)	−.48 (.20)*	−.28 (.13)*	−.56 (.18)**
Constant	−1.87 (.21)***	−.68 (.24)**	−.23 (.16)	.33 (.21)
Outcome category: four-year nonselective college
Race
Black	.05 (.11)	−.13 (.19)	.02 (.11)	.24 (.14)
Hispanic	−.62 (.14)***	−.33 (.17)	−.81 (.14)***	−.22 (.14)
Asian	.67 (.22)**	.30 (.22)	.34 (.21)	1.09 (.22)***
Female	.32 (.07)***	.49 (.09)***	.37 (.07)***	.46 (.09)***
Family income (percentile)	.90 (.15)***	−.03 (.18)	1.22 (.13)***	1.45 (.18)***
Parents’ education (percentile)	2.35 (.13)***	2.69 (.20)***	1.44 (.14)***	1.56 (.19)***
High school location (reference: urban)
Suburban	.00 (.11)	−.05 (.13)	−.11 (.09)	−.16 (.13)
Rural	.07 (.12)	.05 (.13)	−.10 (.12)	−.27 (.13)*
Public high school	−.82 (.13)***	−1.05 (.16)***	−.62 (.12)***	−1.31 (.16)***
Constant	−1.48 (.17)***	−.61 (.20)**	−.73 (.16)***	.15 (.20)
Outcome category: four-year selective college
Race
Black	−.28 (.23)	−.76 (.25)**	−1.09 (.17)***	−.50 (.17)**
Hispanic	−.47 (.25)	−.32 (.25)	−.76 (.15)***	−.43 (.17)*
Asian	1.50 (.25)***	1.05 (.24)***	1.08 (.21)***	1.61 (.21)***
Female	.16 (.11)	.41 (.11)***	.36 (.07)***	.53 (.12)***
Family income (percentile)	1.51 (.26)***	.87 (.24)***	2.13 (.16)***	2.06 (.21)***
Parents’ education (percentile)	3.40 (.24)***	3.91 (.28)***	2.65 (.16)***	2.75 (.26)***
High school location (reference: urban)
Suburban	.15 (.23)	−.01 (.19)	−.12 (.12)	−.16 (.16)
Rural	−.39 (.27)	−.35 (.21)	−.27 (.17)	−.53 (.16)***
Public high school	−1.44 (.21)***	−1.32 (.19)***	−1.05 (.13)***	−1.70 (.18)***
Constant	−3.03 (.36)***	−2.04 (.25)***	−1.66 (.18)***	−.57 (.25)*

Note: Numbers in parentheses are standard errors. Estimations are weighted and adjusted for survey design and to be representative of each population. All covariates are centered at their means. “Nonselective” includes noncompetitive, less competitive, and competitive institutions, while “selective” includes very competitive, highly competitive, and most competitive institutions, as recorded in Barron’s Profiles of American Colleges.

p < .05, **p < .01, and ***p < .001 (two-tailed tests).

Acknowledgements

This study benefited from suggestions from Xuan Dinh, Affie Eyo-Idahor, Suzy Fly, Daniel Mackin Freeman, J. R. Romanko, Rachel Springer, Hannah Sean Ellefritz, Rory Williamson, and Kaitlin Yeomans.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Science Foundation (DRL-1652279) and the National Institutes of Health-funded Build EXITO program at Portland State University (UL1GM118964).

Ethical Approval

Dr. Shifrer maintains these datasets through a restricted-use data license from the Institute for Education Sciences in accordance with their regulations to protect respondent anonymity. Portland State University’s institutional review board decided that this research project, consisting wholly of secondary data analysis, represents minimal risk and qualified for an exemption from institutional review board oversight under 45 CFR 46.

ORCID iDs

Byeongdon Oh

Dara Shifrer

1

We use Hispanic rather than Latinx in this study to be consistent with the phrasing respondents encountered in the surveys.

Author Biographies

Byeongdon Oh is a first-generation immigrant scholar from South Korea, currently serving as an assistant professor of sociology – social justice and change at the State University of New York Polytechnic Institute. He also serves as a computational social science research faculty member in the Social Science D-Lab at the University of California–Berkeley. With a diverse background encompassing experiences as a journalist, military officer, and policy researcher, he has cultivated his sociological insight into social diversity and inclusion. His research focuses on social stratification, education, race/ethnicity, and quantitative and computational methods. He has been grateful to have received funding from the National Science Foundation for his doctoral and two postdoctoral projects. For more information, visit his Web site ().

Ned Tilbrook is a data analyst in the Office of Assessment and Research and a PhD candidate in the Department of Sociology at Portland State University, where he also received his master’s degree in 2020. His dissertation research uses nationally representative data to examine social reproduction in higher education, in particular, the experiences of first-generation college students and qualitative stratification by college majors with a focus on science, technology, mathematics, and engineering. He has also researched college graduate underemployment using qualitative interviews and food access in Portland using geographic information system methods. His work has been funded by the National Science Foundation and published in Social Science Research and Sociology of Education.

Dara Shifrer is an associate professor of sociology at Portland State University. She uses statistical and qualitative approaches to investigate how inequality outside of schools shapes youth’s educational outcomes, and physical and mental health (e.g., neurodevelopmental disabilities, locus of control). She also investigates interventions (e.g., teachers and school-based health centers) that reduce these inequalities. Prior to entering academia, she taught math at two middle schools at the opposite ends of the socioeconomic spectrum. A CAREER award from the National Science Foundation supports her research on the factors that reduce disparities in adolescents’ math and science progress at the intersection of race, disability, and SES. She is a research affiliate at Rice University’s Houston Education Research Consortium.

References

Akos

Patrick

Clayton

Taffye

Kretchmar

Jen

Perez

Lou

Pierce

Catherine

Polk

Barbara

Simmons

Kara

, et al. 2023. Exploring Race-Neutral Alternatives in Undergraduate Admissions. Chapel Hill: University of North Carolina at Chapel Hill.

Alon

Sigal

. 2009. “The Evolution of Class Inequality in Higher Education.” American Sociological Review 74(5):731–55.

Andersen

Margaret L.

Collins

Patricia Hill

. 2020. Race, Class, and Gender: Intersections and Inequalities. 10th ed. Boston: Cengage.

Arum

Rina

Gamoran

Adam

Shavit

Yossi

. 2007. “More Inclusion than Diversion: Expansion, Differentiation, and Market Structure in Higher Education.” Pp. 1–35 in Stratification in Higher Education: A Comparative Study, edited by Shavit

Arum

Gamoran

. Stanford, CA: Stanford University Press.

Baker

Dominique J.

2019. “Pathways to Racial Equity in Higher Education: Modeling the Antecedents of State Affirmative Action Bans.” American Educational Research Journal 56(5):1861–95.

Baker

Rachel

Klasik

Daniel

Reardon

Sean F.

. 2018. “Race and Stratification in College Enrollment over Time.” AERA Open 4(1):1–28.

Bloome

Deirdre

. 2014. “Racial Inequality Trends and the Intergenerational Persistence of Income and Family Structure.” American Sociological Review 79(6):1196–225.

Bowen

William G.

Bok

Derek

. 1998. The Shape of the River: Long-Term Consequences of Considering Race in College and University Admissions. Princeton, NJ: Princeton University Press.

Byrd

W. Carson

. 2021. Behind the Diversity Numbers: Achieving Racial Equity on Campus. Boston: Harvard Education Press.

10.

Clotfelter

Charles T.

Ladd

Helen F.

Vigdor

Jacob

. 2005. “Who Teaches Whom? Race and the Distribution of Novice Teachers.” Economics of Education Review 24(4):377–92.

11.

Coleman

Arthur L.

Keith

Jamie Lewis

. 2018. “Understanding Holistic Review in Higher Education Admissions: Guiding Principles and Model Illustrations.” New York: College Board.

12.

Condron

Dennis J.

Roscigno

Vincent J.

. 2003. “Disparities Within: Unequal Spending and Achievement in an Urban School District.” Sociology of Education 76(1):18–36.

13.

Downey

Douglas B.

2008. “Black/White Differences in School Performance: The Oppositional Culture Explanation.” Annual Review of Sociology 34:107–26.

14.

Du Bois

W.E.B

. 1903. The Souls of Black Folk. Boston: Bedford.

15.

Espinosa

Lorelle L.

Turk

Jonathan M.

Taylor

Morgan

Chessman

Hollie M.

. 2019. “Race and Ethnicity in Higher Education: A Status Report.” Washington, DC: American Council on Education.

16.

Feagin

Joe R.

Ducey

Kimberley

. 2018. Racist America: Roots, Current Realities, and Future Reparations. 4th ed. New York: Routledge.

17.

Fiel

Jeremy E.

Zhang

Yongjun

. 2019. “With All Deliberate Speed: The Reversal of Court-Ordered School Desegregation, 1970–2013.” American Journal of Sociology 124(6):1685–719.

18.

Fischer

Claude S.

Hout

Michael

. 2006. Century of Difference: How America Changed in the Last One Hundred Years. New York: Russell Sage.

19.

Fisher v. University of Texas at Austin, 570 U.S. 297 (2013).

20.

Fisher v. University of Texas at Austin, 579 U.S. 364 (2016).

21.

Fletcher

Jason M.

Tienda

Marta

. 2009. “High School Classmates and College Success.” Sociology of Education 82(4):287–314.

22.

Fredrickson

George M.

2015. Racism: A Short History. Princeton: NJ: Princeton University Press.

23.

Garrison-Wade

Dorothy F.

Lewis

Chance W.

. 2004. “Affirmative Action: History and Analysis.” Journal of College Admission 184:23–26.

24.

Goldin

Claudia

Katz

Lawrence F.

. 2009. The Race between Education and Technology. Cambridge, MA: Belknap.

25.

Goyette

Kimberly A.

2008. “College for Some to College for All: Social Background, Occupational Expectations, and Educational Expectations over Time.” Social Science Research 37(2):461–84.

26.

Gratz v. Bollinger, 539 U.S. 244 (2003).

27.

Grodsky

Eric

. 2007. “Compensatory Sponsorship in Higher Education.” American Journal of Sociology 112(6):1662–1712.

28.

Grutter v. Bollinger, 539 U.S. 306 (2003).

29.

Hanselman

Paul

Fiel

Jeremy E.

. 2017. “School Opportunity Hoarding? Racial Segregation and Access to High Growth Schools.” Social Forces 95(3):1077–1104.

30.

Hattie

John A. C.

2008. Visible Learning: A Synthesis of over 800 Meta-Analyses Relating to Achievement. New York: Routledge.

31.

Hirschman

Daniel

Berrey

Ellen

. 2017. “The Partial Deinstitutionalization of Affirmative Action in U.S. Higher Education, 1988 to 2014.” Sociological Science 4:449–68.

32.

Irwin

Véronique

De La Rosa

Josue

Wang

Hein

Sarah

Zhang

Jijun

Burr

Riley

Roberts

Ashley

, et al. 2022. “Report on the Condition of Education 2022.” NCES 2022-144. Washington, DC: U.S. Department of Education, National Center for Education Statistics.

33.

Jelks

Jammie

. 2010. “Race-Conscious Admission Alternatives.” ACPAAdmin 8(1).

34.

Kane

Thomas J.

2003. “The Long Road to Race-Blindness.” Science 302(5645):571–73.

35.

Kao

Grace

Thompson

Jennifer S.

. 2003. “Racial and Ethnic Stratification in Educational Achievement and Attainment.” Annual Review of Sociology 29:417–42.

36.

Kehal

Prabhdeep Singh

Hirschman

Daniel

Berrey

Ellen

. 2021. “When Affirmative Action Disappears: Unexpected Patterns in Student Enrollments at Selective U.S. Institutions, 1990–2016.” Sociology of Race and Ethnicity 7(4):543–60.

37.

Logan

John R.

Minca

Elisabeta

Adar

Sinem

. 2012. “The Geography of Inequality: Why Separate Means Unequal in American Public Schools.” Sociology of Education 85(3):287–301.

38.

Long

Mark C.

Bateman

Nicole A.

. 2020. “Long-Run Changes in Underrepresentation after Affirmative Action Bans in Public Universities.” Educational Evaluation and Policy Analysis 42(2):188–207.

39.

Lucas

Samuel R.

2001. “Effectively Maintained Inequality: Education Transitions, Track Mobility, and Social Background Effects.” American Journal of Sociology 106(6):1642–90.

40.

Massey

Douglas S.

Charles

Camille Z.

Lundy

Garvey

Fischer

Mary J.

. 2011. The Source of the River: The Social Origins of Freshmen at America’s Selective Colleges and Universities. Princeton, NJ: Princeton University Press.

41.

Noguera

Pedro

. 2003. City Schools and the American Dream: Reclaiming the Promise of Public Education. New York: Teachers College Press.

42.

Byeongdon

Kim

Changhwan

. 2020. “Broken Promise of College? New Educational Sorting Mechanisms for Intergenerational Association in the 21st Century.” Social Science Research 86:102375.

43.

Orfield

Gary

Frankenberg

Erica

Jongyeon

Kuscera

John

. 2014. “Brown at 60: Great Progress, a Long Retreat and an Uncertain Future.” Los Angeles: The Civil Rights Project, University of California–Los Angeles.

44.

Potter

Halley

. 2014. “What Can We Learn from States That Ban Affirmative Action?” New York: The Century Foundation.

45.

Raftery

Adrian E.

Hout

Michael

. 1993. “Maximally Maintained Inequality: Expansion, Reform, and Opportunity in Irish Education, 1921–75.” Sociology of Education 66(1):41–62.

46.

Reardon

Sean F.

Baker

Rachel

Klasik

Daniel

. 2012. “Race, Income, and Enrollment Patterns in Highly Selective Colleges, 1982–2004.” Stanford, CA: Center for Education Policy Analysis, Stanford University.

47.

Reardon

Sean F.

Baker

Rachel

Kasman

Matt

Klasik

Daniel

Townsend

Joseph B.

. 2018. “What Levels of Racial Diversity Can Be Achieved with Socioeconomic-Based Affirmative Action? Evidence from a Simulation Model.” Journal of Policy Analysis and Management 37(3):630–57.

48.

Regents of the University of California v. Bakke, 438 U.S. 265 (1978).

49.

Rothstein

Richard

. 2004. Class and Schools: Using Social, Economic, and Educational Reform to Close the Black–White Achievement Gap. New York: Teachers College Press.

50.

Schuette v. Coalition to Defend Affirmative Action, 572 U.S. 291 (2014).

51.

Shifrer

Dara

. 2022. “Contextualizing Educational Disparities and the Evaluation of Teacher Quality.” Social Problems 69(2):841–86.

52.

Students for Fair Admissions Inc. v. President & Fellows of Harvard College, 600 U.S. ____ (2023).

53.

Tamborini

Christopher R.

Kim

ChangHwan

Sakamoto

Arthur

. 2015. “Education and Lifetime Earnings in the United States.” Demography 52(4):1383–1407.

54.

Taniguchi

Hiromi

. 2005. “The Influence of Age at Degree Completion on College Wage Premiums.” Research in Higher Education 46(8):861–81.

55.

Teranishi

Robert T.

Nguyen

Bach Mai Dolly

Alcantar

Cynthia M.

Curammeng

Edward R.

. 2020. Measuring Race: Why Disaggregating Data Matters for Addressing Educational Inequality. New York: Teachers College Press.

56.

Torche

Florencia

. 2011. “Is a College Degree Still the Great Equalizer? Intergenerational Mobility across Levels of Schooling in the United States.” American Journal of Sociology 117(3):763–807.

57.

U.S. Government Accountability Office. 2016. “Better Use of Information Could Help Agencies Identify Disparities and Address Racial Discrimination.” GAO-16-345. Washington, DC: U.S. Government Accountability Office.

58.

von Hippel

Paul T.

2007. “Regression with Missing Ys: An Improved Strategy for Analyzing Multiply Imputed Data.” Sociological Methodology 37(1):83–117.

59.

Voss

Kim

Hout

Michael

George

Kristin

. 2022. “Persistent Inequalities in College Completion, 1980–2010.” Social Problems. doi:10.1093/socpro/spac014.

60.

Wang

Jackie

. 2018. Carceral Capitalism. Los Angeles: Semiotext(e).

61.

Western

Bruce

Wildeman

Christopher

. 2009. “The Black Family and Mass Incarceration.” Annals of the American Academy of Political and Social Science 621(1):221–42.

62.

White

Ian R.

Royston

Patrick

Wood

Angela M.

. 2011. “Multiple Imputation Using Chained Equations: Issues and Guidance for Practice.” Statistics in Medicine 30(4):377–99.

63.

Wilson

Valerie

Schieder

Jessica

. 2018. “The Rise in Child Poverty Reveals Racial Inequality, More Than a Failed War on Poverty.” Washington, DC: Economic Policy Institute.