UCLA

Background

The percentage of U.S. physicians who identify as being from an underrepresented racial or ethnic group remains low relative to their proportion in the U.S. population. How this percentage may have been affected by state bans on affirmative action in public postsecondary institutions has received relatively little attention.

Objective

To examine the association between state affirmative action bans and percentage of enrollment in U.S. public medical schools from underrepresented racial and ethnic groups.

Design

Event study comparing public medical schools in states that implemented affirmative action bans with those in states without bans.

Setting

U.S. public medical schools.

Participants

21 public medical schools in 8 states with affirmative action bans matched to 32 public medical schools in 24 states without bans from 1985 to 2019.

Measurements

Percentage of total enrollment from racial and ethnic groups underrepresented in medicine (Black, Hispanic, American Indian or Alaska Native, and Native Hawaiian or other Pacific Islander).

Results

The percentage of enrollment from underrepresented racial and ethnic groups was 14.8% in U.S. public medical schools in the year before ban implementation in states with bans. The adjusted percentage of underrepresented students in ban schools decreased by 4.8 percentage points (95% CI, -6.3 to -3.2 percentage points) 5 years after ban implementation relative to the year before implementation, whereas the adjusted percentage in control schools increased by 0.7 percentage point (CI, -0.1 to 1.6 percentage points), for a relative difference, or difference-in-differences estimate, of -5.5 percentage points (CI, -7.1 to -3.9 percentage points).

Limitation

Inability to account for the effect of these bans on undergraduate enrollment.

Conclusion

State affirmative action bans were associated with significant reductions in the percentage of students in U.S. public medical schools from underrepresented racial and ethnic groups.

Primary funding source

None.