By Richard Sander
Amid the intense heat and noise generated by my report on UCLA’s undergraduate admissions, it is very easy to lose sight of a few basic facts.
First: No one to my knowledge has disputed a single one of the factual statements in my report. For that matter, few people seem to have even read it. I’ve also posted (at the same location) all the data used in my report (provided by UCLA after prolonged negotiations), and the programs used to generate Tables 4 through 8 (Tables 1 through 3 came from readily available, public data).
Second: Much confusion has resulted from a failure to distinguish the general “holistic admissions system” adopted by UCLA beginning with the 2007 entering freshman class, and the initial “holistic scores” generated within that system.
The initial holistic scores are assigned by special readers hired and trained by the admissions office, and are intended to take into account all characteristics of applicants. (It replaced a system where “academic,” “personal achievement” and “life challenges” characteristics were evaluated separately.)
All three of the scholars who have carefully examined the data (Professor Robert Mare, Professor Tim Groseclose, and I) have concluded that the initial holistic scores were assigned in an essentially race-neutral way. That is to say, while some factors (e.g., test scores) favored one racial group, and other factors (e.g., family poverty) favored another, when one controls for all the factors in applicant files, race played no independent role.
The problem is that UCLA’s Admissions Office made a large number of admissions decisions that disregarded the initial holistic scores. Through its still-mysterious final review and supplemental review processes, it either “adjusted” the scores assigned by the initial holistic readers, or disregarded them entirely. These decisions correlated very highly with race ““ that is to say, any number of statistical tests show that black applicants were very disproportionately likely to be admitted through these “post-holistic” adjustments. Again, Mare, Groseclose and I have all found this to be the case.
Many of those ““ including the UCLA administration ““ who have denounced my report suggest that it is “baseless” and in conflict with the Mare report because Mare concluded that “the holistic ranking process … appears to work much as intended.” But there is no conflict.
Mare is referring here to the holistic scoring by the initial holistic readers. He goes on to say in the very same paragraph of his report that “In final and supplemental review, which are conducted by UARS staff, I do find some disparities in outcomes that favor some (racial) groups and disfavor others among applicants who are otherwise similar on their measured characteristics…”
Third: The apparent discrimination engaged in by UCLA admissions officers was overwhelmingly focused on admitting more black applicants. If we compare 2004-06 admissions with 2007-09 admissions, Asian and white applicants experienced about a 10 percent decline in admission rates. Hispanic applicants also experienced about a 10 percent decline. But black applicants experienced about a 30 percent increase in admission rates.
This, of course, is another strong circumstantial reason why even those who are not statistical experts should find UCLA’s conduct highly suspicious and troubling. If the new admissions system begun in 2007 simply did a better job of taking personal disadvantage into account, it should have benefited both black and Hispanic applicants (who, as groups, both experience similar levels of socioeconomic disadvantage in California relative to whites and Asians). But it did not.
Combined with the fact that UCLA ““ and the admissions office in particular ““ was under great political pressure in 2006-07 to produce an increase in black freshman enrollment, we have an obvious motive for improper and illegal behavior. And the statistical analyses pretty much seal the case.
UCLA officials have known about these troubling statistical patterns for a good two years now, ever since Professor Mare produced an early draft of his report. Officials have had a long time to produce data explaining the enormous racial disparities in final and supplemental review decisions, but they have not.
Two final points about me in particular. First, I am more than willing to defend my report and engage in dialogue about the admissions controversy with anyone on campus, at any time that does not conflict with my other teaching and lecturing responsibilities. That is why I attended Monday’s rally outside Kerckhoff Hall and tried ““ unsuccessfully ““ to persuade the leaders of the rally to let me participate in the discussion.
Second, I do believe in the value and contributions of everyone here at UCLA. Our community undoubtedly benefits from diversity of all kinds, including racial diversity. But our community also benefits ““ very fundamentally ““ from a respect for the law.
Sander is a UCLA law professor.