The case for affirmative action in higher education

Affirmative action is under attack, and the author responds, explaining affirmative action's historical principles and purposes, dispelling its myths, and discussing why it must remain a part of higher education.     

State affirmative action bans and STEM degree completions

This paper investigates the effect of statewide affirmative action bans on minority STEM degree completions at US public four-year colleges. The number of minority students completing STEM degrees at highly selective colleges falls by 19% five years after affirmative action bans, while there is no change in the total number of students completing STEM degrees. This indicates that a nontrivial number of minority students only admitted to highly selective colleges because of affirmative action graduate in STEM during periods of race preferences in college admissions. There is no convincing evidence of effects at moderately selective colleges. These findings speak to the recent debate about the extent to which minority students admitted to top ranked colleges due to affirmative action may have higher probabilities of graduating in the sciences if they had attended lower ranked colleges. Results are presented with the caveats that changes in race reporting caused by affirmative action bans may upwardly bias estimated effects, and that estimated aggregate effects may not fully capture all student-level responses.     

Academic and diversity consequences of affirmative action in Brazil

Since 2001, Brazilian universities have been implementing affirmative-action policies to correct the racial, social and ethnic disparities in university admissions. An examination of the social-inclusion policies at three public universities in Brazil – the University of Brasilia, the Federal University of Bahia and the State University of Campinas – was undertaken to analyse: (1) the impact of affirmative-action policies on campus diversity, (2) the academic performance of quota and non-quota students and (3) the predictive validity of the college entrance exam as a selection criterion. Using institutional enrolment, course grades in several academic programmes and retention indicators, the study finds that quota students do reasonably well and thus do not diminish the quality of their institutions. Academic performance varies by programme of study but, in general, variation is surprisingly slight. This latter finding suggests that the predictive validity of the college entrance exam is weak.     

The impact of Israel's class-based affirmative action policy on admission and academic outcomes

In the early to mid-2000s, four flagship Israeli selective universities introduced a voluntary need-blind and color-blind affirmative action policy for students from disadvantaged backgrounds. The program allowed departments to offer admission to academically borderline applicants who were above a certain threshold of disadvantage. We examine the effect of eligibility for affirmative action on admission and enrollment outcomes as well as on academic achievement using a regression discontinuity (RD) design. We show that students who were just barely eligible for this voluntary policy had a significantly higher probability of admission and enrollment, as compared to otherwise similar students. The affirmative action program also led to higher rates of admission to the most selective majors. Moreover, after enrollment, AA-eligible students are not falling behind academically, even at the most selective majors. Our results suggest the potential for a long-lasting impact of class-based preferences in admission on social and economic mobility.     

Autonomy versus affirmative action: What price social justice?

This study measures all the marginal direct and indirect costs of an affirmative action/equal employment opportunity (AA/EEO) program at a single institution through cost analysis procedures identifying cost generating variables in three categories: personnel, operating expenses, and capital outlay replacement value. Costs were weighed against benefits, as measured by number of women and minorities hired. Since costs and benefits should be evaluated only in relation to effectiveness of implementation procedures, program administration was assessed through a series of interviews with faculty chairs. Total compliance costs incurred for the university in one budget cycle amounted to 0.4% of the total institutional budget, and the great majority of these costs were expended indirectly, in faculty time. Representation of women and minorities was increased by approximately 1.9%, despite lack of strong leadership from the central administration and haphazard program administration. Although the costs incurred in implementing AA/EEO are negliglible compared to other social justice legislation, educators still decry its expense. Thus, we argue that educators are not concerned with the actual dollar outlay for AA/ EEO but instead use its intangible educational costs as a symbolic issue in their struggle with the state over university autonomy.     

Not just racial quotas: affirmative action in Brazilian higher education 10 years later

This paper examines how affirmative action in Brazilian public universities has evolved from the start of the new millennium up to recent years. After an overview of the existing policies in public universities, we explain these patterns based on an analysis of the processes internal and external to the universities. Although these policies were initially thought of as racial quotas, class-based criteria have predominated, while the racial criteria became relatively circumscribed. Within the institutions, affirmative action brought new concerns about how to support and retain low-income students, as well as a discussion about racism, although the latter remains controversial.     

On discarding affirmative action: An address to the regents of the University of California

Editor's Note: On 20 July 1995 the Board of Regents of the University of California voted to end the use of race, gender, and ethnicity in student admissions and in faculty and staff hiring. Before the vote, thirty-one “elected and other public figures” were scheduled to address the Regents briefly on the subject of affirmation action. Governor Pete Wilson (ex officio president of the Board) was allocated several of the thirty-one slots. Among those he invited to speak was NAS member John M. Ellis. We present Professor Ellis’s remarks below.     

The redistributive equity of affirmative action: Exploring the role of race,socioeconomic status,and gender in college admissions

This paper contributes to research on affirmative action by examining issues of equity in the context of racial quotas in Brazil. We study the experience of the University of Brasilia, which established racial quotas in 2004 reserving 20% of available admissions slots for students who self-identified as black. Based on university admissions data and a student survey conducted by the authors, we find evidence that race, socioeconomic status, and gender were considerable barriers to college attendance and achievement. For example, first-difference regressions involving pairs of siblings indicate that black identity and gender had a negative effect on entrance exam scores. Moreover, we compare displaced and displacing applicants and find that racial quotas helped promote equity to some extent. Nevertheless, the scale and scope of redistribution were highly limited, and the vast majority of Brazilians had little chance of attending college, suggesting that more still needs to be done.