Against For Discrimination
A review of Randall Kennedy’s For Discrimination: Race, Affirmative Action, and the Law.

Randall Kennedy


The title, “For Discrimination,” is honest at least — especially compared with most defenses of racial preferences, that being a low bar indeed. And, to be sure, the author, Randall Kennedy, is honest in many ways, particularly in using a cost/benefit framework to assess affirmative action’s justifiability, listing the purported benefits and weighing them against (most of) the undeniable costs. But he is completely incomprehensible in deciding that the purported benefits could possibly outweigh the costs. And I should add that the book is not always fair. Kennedy obviously has a special dislike for Justice Clarence Thomas, the unseemly expression of which detracts from the book’s attempt to strike an evenhanded tone.

Kennedy suggests four possible justifications for racial preferences.

The first one, the only justification available today in university admissions, which is the focus of the book, is diversity. Kennedy doesn’t think much of it. More points to him for honesty. While he no longer quite holds this argument in “disdain,” he “remain[s] doubtful about social scientific ‘proof’ of diversity’s [teaching, learning, and decision-making] value; much of that seems exaggerated and pre-determined with litigation in mind.” Later he criticizes Justice Sandra Day O’Connor for accepting such studies without the appropriate skepticism. He can hardly believe that the diversity justification can meet the constitutional requirement of being “compelling” if it can be accepted only “so long as the demands and expectations imposed on it are not too onerous.”

This is, for me, the key takeaway from this book: A high-profile liberal apologia for racial preferences has little use for their only justification that has any legal viability.

A second justification for racial preferences, their possible prophylactic effect against discrimination, has conceivable but dwindling plausibility in some areas, and none whatsoever in the university-admissions context, where for decades now university officials have been all too happy to engage in politically correct racial discrimination.

The third justification, the integrative effect of racial preferences, is what Justice Lewis Powell called “discrimination for its own sake” in his seminal Bakke opinion. It has never been adopted by the Court as a compelling interest (O’Connor only adverts to something like it in her Grutter opinion).

The fourth justification is clearly Kennedy’s favorite: racial preferences as a kind of reparations for the lingering effects of historical societal discrimination. It, too, has been rejected by the Court, as Kennedy acknowledges. And rightly so — as he denies.

No doubt some of the racial disparities we see now can be traced to slavery and Jim Crow, but there is likewise no doubt (a) that some of the causes of the disparities are self-inflicted (Kennedy says precious little about those causes — and nothing in particular about illegitimacy, the underlying and principal ongoing cause) and (b) that with every tick of the clock, the effect of slavery and Jim Crow recedes.

To his credit, Kennedy admits that many of those receiving preferences today are not among those who suffer most from racial disparities. He fails to acknowledge, however, that pointing to past discrimination does not provide much justification for giving Latinos a preference over Asian Americans — an increasingly common result of preferences, as these two groups are growing much faster than blacks and whites. Indeed, Kennedy generally ignores the problem of discrimination against Asian Americans in university admissions.

In all events, it makes no sense to use race as a proxy for disadvantage. The reparations argument anthropomorphizes different racial groups when in fact many children of all colors are poor for all kinds of reasons. Why pick out one color of children who are (or might be) poor, and whose poverty is (or might be) caused by a particular reason, and give them special treatment?

To elaborate: The government (or a university) has before it a group of socioeconomically disadvantaged 18-year-olds. Some are black and can plausibly trace their poverty to slavery and Jim Crow. Others cannot — maybe they are recent immigrants, for example, or maybe their parents just made bad lifestyle decisions. Some are Hispanic and, with greater difficulty, might be able to trace their poverty to discrimination; more are poor simply because their parents came recently from poor countries. Some are Asian and, in the aggregate, their reasons for poverty mirror those of Hispanics. Some are non-Hispanic whites and can trace their poverty to a variety of reasons — some of which might cause us to feel sympathy for their parents, some not. And there are Arab Americans, Native Americans, and Americans of a variety of mixed ancestries.