Affirmative action -- non-randomly selected papers

Indigenous Brazilian being expelled from the Supreme Court for disturbing
session that led to the constitutionality of quotas for blacks (source: FSP)

This week the Brazilian Supreme Federal Court decided that racial quota systems for access to the university are constitutional. In the process a lot of arguments were thrown favoring or against such a system, but most of them without much research to back them up. I had the impression that most of the arguments defending quotas interchanged racial and social conditions, reviving the prejudice as a whole in Brazil, that conflates both. Furthermore in a mixed ancestry country like Brazil is very hard to determine objectively if one belongs to a racial minority, but very easy to quantify how poor one is. The discussion is long and I don't plan to do it here (suffices to say that I cannot be against the supreme court's decision, despite I really don't think racial quotas -- or any solution at the superior education level -- will get close to solving the Brazilian educational problems).

Anyway, I decided to look a little bit into http://ideas.repec.org/ for papers about affirmative action, specially for races -- I understand admission quotas are unconstitutional in the US, for instance, but I think that the analyses are close enough. There's too much material for me to read carefully, so here are some articles that I chose based on interesting or apparently neglected conclusions. Be aware that this is not even close to a broad, complete or exhaustive survey, and that I picked the articles up based on one source and reading only the abstracts in most cases. Following the links is essential for understanding the articles -- even in the abstracts there might be relevant information that I didn't reproduce here. Also notice that some articles use empirical data, others simulate, while others still are more argumentative.

Some Experimental Evidence on the Coate and Loury Model of Affirmative Action

Although benign and patronising equilibria are equally plausible in theory, the experiments easily replicate the benign equilibrium, but find that a certainty effect leads to over-investment and divergence from the theoretically predicted patronising equilibrium.

Affirmative action in education: Evidence from engineering college admissions in India

We find that despite poor entrance exam scores, lower-caste entrants obtain a positive return to admission. Our estimates, however, also suggest that these gains may come at an absolute cost because the income losses experienced by displaced upper-caste applicants are larger than the income gains experienced by displacing lower-caste students.

The Informational Value of Affirmative Action in College Admissions

In the theoretical model, the racial identity of the applicants is relevant information for making inferences about an applicant's true academic ability. Preventing admissions officers from using this information results in inefficient selection of applicants, even if diversity does not explicitly enter the objective of the university.

Does Banning Affirmative Action Affect Racial SAT Score Gaps? An Empirical Analysis

My analysis centers on the recent elimination of affirmative action in California and Texas. Using SAT data obtained from The College Board, I compare the changes of the test score gap between minorities and non-minorities over time in the treatment states, relative to a set of control group states. (...)
[from conclusion]  Despite the effects of Texas' Top Ten Percent Plan and redoubled minority recruiting efforts in both states, the SAT remains an important component of an applicant's profile and thus provides relevant evidence about the effects of eliminating affirmative action. I estimate statistically and economically significant increases of the Hispanic­-White test score gap in California, and of the Black­-White test score gap in Texas. Other test score gaps appear to be constant over the time period in question.

The Evolution of a Social Contract: The University of California Before and in the Aftermath of Affirmative Action

This brief essay covers the history of admissions at the University of California (UC), including the development of affirmative action programs in the 1960s and, more recently, the heated political battle over the use of race and gender preferences at the University.
[from conclusion] Precisely because it is problematic, essentially elevating a larger societal goal over the rights of individuals, it remains a vague doctrine that neither UC officials nor the public fully comprehends. While it is steeped in the rationale of egalitarianism, the politics of its logical conclusion are substantial: only by limiting access of one racial group (or one economic category) can another "underrepresented" group gain increased access. This would mean systematically constraining access by "overrepresented" groups ­ namely, Asian-American students. The irony of such a path illustrates the complexity of seeking social redress.

A Cure for Discrimination? Affirmative Action and the Case of California Proposition 209

Employment among women and minorities dropped sharply, a change that was nearly completely explained by a decline in participation rather than by increases in unemployment. This decline suggests that either affirmative action programs in California had been inefficient or that they failed to create lasting change in prejudicial attitudes.

Affirmative Action Policy and Effort Levels. Sequential-Move Contest Game Argument

We also prove that in such cases a significant role is played by the level of the heterogeneity of individuals. In particular, in contrast to the present-in-the-literature predictions, we demonstrate that as a consequence of the interplay of these two factors, the response to the implementation of the affirmative action policy option may be the decrease in the total equilibrium effort level of the contestants in comparison to the unbiased contest game.

But Affirmative Action hurts Us! Race-related beliefs shape perceptions of White disadvantage and policy unfairness

We also find that White disadvantage mediates the relationship between the combined effects of race-based AAPs, MR beliefs, and CRD beliefs and the perceived fairness of the organization's selection and promotion policies. Our findings suggest that race-based AAPs do not necessarily lead to perceptions of White disadvantage, but are contingent upon the interpretive lens of Whites' MR and CRD beliefs, and also offer practical insights for preventing negative reactions to race-based AAPs.

Anti-Discrimination Versus Anti-Poverty: Does Affirmative Action Hurt the Poor?

The author relied upon data from the Current Population Survey (CPS) 1975-1999 and a generalized least squares model to test the relationship between the percent of Blacks who qualify as poor in a given year and several measures of affirmative action. The author finds a mixed but largely null relationship between Black poverty and affirmative action and concludes that affirmative action programs have little impact on Black poverty levels.

Does affirmative action work?

After four decades, we are still debating how much impact affirmative action can and should have on opportunities and outcomes at work.
[from text] One concern that has been raised is that it creates de-facto quotas. But work by Jonathan Leonard shows that federal contractors tend to fall short of their employment goals for women  and minorities, suggesting that they are indeed goals and not quotas.

An efficiency argument for affirmative action in higher education

The individual decision to undertake education is inefficient because people fail to account for the fact that their getting education increases the chances that their children will also gain access to education. This intergenerational externality is higher for people from the disadvantaged minority, provided that the difference in expected utility for children of uneducated and educated individuals is larger within this group. This provides an argument for affirmative action in higher education, in the form of larger subsidies to individuals from the minority group, which is exclusively based on efficiency considerations.

Affirmative Action: Diverging Perspectives

The question of merit, therefore, demands a sustained and careful analysis and indeed needs to answer - how to detect merit?; how to define merit?; and how to ascertain and establish a proper sphere of influence for merit?

Affirmative Action and Efficiency in Education

The main result is that the high ability individuals from groups with relatively few high ability individuals should receive more education than equally able individuals from groups with a more favourable distribution of abilities.

Peer effects in affirmative action: Evidence from law student performance

At these schools, we find no consistent evidence that having additional students from one's ethnic group raises a student's performance. Instead, we find some evidence that having additional ethnic peers lowers performance - albeit by a very small amount.

Affirmative Action and Its Mythology

This paper enumerates seven commonly held but mistaken views one often encounters in the folklore about affirmative action. 1) Affirmative action may involve goals and timelines, but definitely not quotas. 2) Color-blind policies offer an efficient substitute for color-sighted affirmative action. 3) Affirmative action creates opportunities but does not undercut incentives. 4) Passing equal opportunity laws is enough to ensure racial equality. 5) The earlier affirmative action is used in education or career development, the better. 6) Many whites are directly affected by affirmative action policies designed to increase representation of minorities. 7) Affirmative action always helps its intended beneficiaries.

College admissions with affirmative action

Then it shows that the student-applying deferred acceptance mechanism makes truthful revelation of preferences a dominant strategy for every student if each college's preferences satisfy responsiveness over acceptable sets of students that respect its type-specific quotas. These results have direct policy implications in several entry-level labor markets (Roth 1991). Furthermore, a fairness notion and the related incentive theory developed here is applied to controlled choice in the context of public school choice by Abdulkadiroglu and Smez (2003).

Affirmative Action in Higher Education in India and the US: A Study in Contrasts

The following paper makes an attempt to understand the nuances of a caste-based reservation policy in higher education in light of recent controversies, court verdicts, a subsequent amendment to the constitution in India; and affirmative action policies, court verdicts, and alternatives to affirmative action in certain universities in the US. The objective is to bring out commonalities and contrasts between the two countries in terms of legal, political, socio-cultural, economic, and psychological perspectives.

A Theory of Affirmative Action in College Admissions

Such a "handicapping" rule increases competition and induces candidates to invest more in educational attainment. These results reconcile the often-assumed conflicts between diversity and academic quality. However, we also show that the non-minority responds to the affirmative action admissions more aggressively, which tends to widen the racial test score gap.

Does Banning Affirmative Action Harm College Student Quality?

When it cannot use affirmative action, the admissions office's preferred rule generally does not select the best-qualified candidates from either group: it randomizes over candidates to achieve diversity, at the expense of within-group selection. A ban always reduces diversity, and may also lower average quality. Moreover, even when a total ban on affirmative action raises average quality, a partial ban may raise average quality even more.

Job Queues, Discrimination, and Affirmative Action

We find that a job queue existed before the implementation of affirmative action and that affirmative action changed the ordering of the job queue in manner that had a negative impact on nonwhite males. We find evidence that affirmative action may have increased the unemployment rate of nonwhite males by increasing their duration of unemployment.

The Distributional Consequences of Diversity-Enhancing University Admissions Rules

High-income majority citizens, who tend to have better academic qualifications, favour more diversity under affirmative action, which imposes its costs on marginal majority candidates. Lower-income majority citizens prefer less diversity under affirmative action and would rather achieve diversity by de-emphasizing academic qualifications. Increasing income inequality among majority citizens tends to reduce the median citizen's support for affirmative action.

Could affirmative action be efficient in higher education?

Many states are outlawing the use of affirmative action, which has lead to the use of inefficient measures as a substitute. I show that schools still use affirmative action in the application process because of a down-stream demand from employers.

Rethinking Admissions: US Public Universities in the Post-Affirmative Action Age

The paper also discusses the use of standardized tests in judging students' readiness for university-level work, and especially changes to the SAT tests that have come about partly in response to UC policies. The paper concludes by assessing the ongoing debates over racial preferences in college admissions.

Diversity and Affirmative Action in Higher Education

We find that minority students pay lower tuition and attend higher-quality schools because of affirmative action. A ban of affirmative action will lead to a substantial decline of minority students in the top-tier colleges.

What determines public support for affirmative action?

When the education premium is relatively low, the matching efficiency gains provided by affirmative action policies are relatively high compared to the opportunity cost of not acquiring education, and the majority supports broader affirmative action. In contrast, when the education premium is high, the opportunity cost of not acquiring educated is high relative to the matching efficiency gains provided by affirmative action policies, and the majority's support for affirmative action is weaker. With endogenous wages, the negative relationship between the returns to education and affirmative action is reinforced.

Reversal of Envy

I define envy as occurring when people with the same abilities have different incomes because of unequal education and employment opportunities. This is the case when, because of adverse discrimination, low-caste people are denied access to education and public-sector jobs, and also when, because of compensating discrimination, it is high-caste people who are correspondingly denied equal access.

Will Affirmative Action Policies Eliminate Negative Stereotypes?

They study the joint determination of employer beliefs and worker productivity in a model of statistical discrimination in job assignments. The authors prove that, even when identifiable groups are equally endowed ex ante, affirmative action can bring about a situation in which employers (correctly) perceive the groups to be unequally productive, ex post.

Maintaining the Legitimacy of the High Court:Understanding the "25 Years" in Grutter v. Bollinger

Given the divided opinion of the American public on the issue of affirmative action in higher education, it was expected that the Court's rulings would ignite fervor amongst individuals on either side of the debate, whichever way the decisions came out

Affirmative action results from the University of Campinas.
Comparison between students without bonus (top bars), with bonus from public school (middle bars) and bonus from self-declared race (bottom bars), according to course conclusion rate (top panel) and estimated grade repetition (bottom panel). (source: FSP)