Researchers Find Substantial Drop in Use of Affirmative Action in College Admissions

University of Washington researchers Grant H. Blume and Mark C. Long have produced the first empirical estimates using national-level data to show the extent to which levels of affirmative action in college admissions decisions changed during the period of 1992 to 2004. Blume and Long’s study, “Changes in Levels of Affirmative Action in College Admissions in Response to Statewide Bans and Judicial Rulings,” was recently published online in Educational Evaluation and Policy Analysis (EEPA), a peer-reviewed journal of the American Educational Research Association (AERA).

The authors find a significant decline nationally in the level of affirmative action used by selective public colleges from 1992 to 2004. This decline is attributable to institutions in the eight states affected by statewide affirmative action bans or Circuit Court rulings during the period (Alabama, California, Florida, Georgia, Louisiana, Mississippi, Texas, and Washington).

While selective public institutions in the eight states, which Blume and Long label “post affirmative action states,” ceased giving preferences to minority applicants in their admissions decisions, selective public institutions in other states continued to do so.

“There’s been a lot of debate about whether or not states are complying with the law,” said Blume. “Our research shows that they are.”

Blume and Long define affirmative action as the higher probability of admission for an underrepresented minority (black, Hispanic, or Native American) student compared with a similar nonminority (white or Asian) student at the same institution. They found substantial and significant preference being given to minority students in 1992 at highly selective institutions in all states, but by 2004 there were dramatic declines in preference given to minority students in the eight affected states.

Outside of the eight post-affirmative action states, at selective public institutions (i.e., colleges and universities where an enrolled first-year student has a median SAT score of 1,100 or higher) a minority applicant was 19 percentage points more likely to be admitted than a comparable non-minority applicant in 1992, and this advantage remained at 18 percentage points in 2004.

The researchers also found spillover effects for students living in states bordering the eight affected states, particularly in Arizona, Nevada, and Idaho. Because these neighboring states lack highly selective institutions, students who apply to selective institutions in nearby California and Washington were affected by the decline in affirmative action in those two states.

“These spillover effects are important to consider, especially in light of the growing number of states that have more recently enacted statewide bans (including Michigan, Nebraska, Arizona, and Oklahoma) as these bans will likewise have regional implications for college applicants,” said Long.

To determine the magnitude of changes in affirmative action in states affected by bans and court rulings, Blume and Long examined nationally representative data on admissions decisions from 1994 and 2004. In Texas, the 1997 Hopwood v. Texas ruling effectively banned affirmative action in college admissions; voter referendums in California in 1996 and in Washington in 1998 and administrative decisions in Florida in 1999 had the same result. The Hopwood and 2001 Johnson v. Board of Regents of the University of Georgia rulings against affirmative action applied to public colleges in Alabama, Georgia, Louisiana, and Mississippi.