When the case for affirmative action in college and university admissions is argued before the Supreme Court this year, the justices are likely to hear many theories, many assertions -- and little evidence. People who are for or against affirmative action are usually for or against the theory of it. What actually happens under this policy gets remarkably little attention.
There is one study, however, that is virtually certain to be cited as evidence by those defending racial preferences and quotas, because it is regarded by liberals in the media and academia as definitive -- at least by those who confuse voluminous with definitive.
That study is titled "The Shape of the River" by former university presidents William Bowen and Derek Bok. While this book has lots of statistics, none of those statistics is specifically about the ostensible subject of the book -- black students who are admitted to colleges and universities with lower qualifications than white students, as a result of racial preferences or quotas.
This much-heralded study is Hamlet without the prince of Denmark. Bowen and Bok lump together those black students who were admitted under the same standards as white students with those black students who were admitted under lower standards. But the issue is not whether any black students should be admitted to elite colleges. The issue is whether they should be admitted under the same standards as others.
Other studies have confronted that issue. At universities where the test scores of black and white students are similar, their graduation rates have been similar. At universities where there are wide gaps between the average test scores of black and white students, there are usually wide gaps between their graduation rates.
At the flagship University of Colorado campus at Boulder, where the average SAT score of black students was more than 200 points lower than that of white students, only 39 percent of the black students graduated, compared to 72 percent of the whites.
At the University of Colorado at Denver, however, where the difference in SAT scores was only 30 points, half of all black students and 48 percent of all white students graduated within a six-year span. Where there were negligible differences in qualifications, there were negligible differences in results.
Comparisons of other institutions with large and small racial gaps in test scores show very similar patterns -- except for the Bowen and Bok study, in which affirmative action seems to have no negative effects. But one of the remarkable things about the Bowen and Bok study is that the raw data from which their statistics were compiled, and from which their many tables and graphs were put together, remain secret and off-limits to other researchers.
Their refusal to separate out those black students admitted under lower academic standards, combined with their refusal to let others get their hands on the data they used, so that others could make that separation if they wished, make the Bowen and Bok study an odd choice to rely on so uncritically as much of the media and academia have relied on it.
It is hard to explain this reliance on a study conducted in such an unusual way, and whose results are so different from other studies conducted according to the usual scholarly practices -- except that the Bowen and Bok book told many people what they so much wanted to hear.
The sample used in the Bowen and Bok study was as questionable as their methods. That sample consisted of 24 private colleges and universities -- and only four state institutions. But only 9 percent of black college students attend private institutions, which constitute the bulk of the Bowen-Bok sample.
Nearly two-thirds of the black students in the study by Bowen and Bok had at least one parent who had graduated from college, which is by no means the norm for black students attending college. What all this shows is that highly atypical blacks attending a highly atypical sample of institutions have a different pattern from most blacks attending most institutions.
The 14th Amendment's requirement for equal treatment should not be overthrown or evaded because of one contrived "study."