Thomas A. Saenz, president of the Mexican American Legal Defense and Educational Fund, served as co-chairman of Proposition 16. Black and Latino students, he said, remain marked by bias from kindergarten to high school, from standardized tests and grades to the expectations of teachers and counselors.
What looks like progress — the growing number of Latino students — is attributable mostly to demographic growth, he said.
“Much of what passes for merit-based admissions is influenced by subconscious bias,” he said. “We have to guard against a coronation of color blindness.”
There is reason to wonder if California’s model is replicable. The state has poured money and effort into attracting diverse students. In a post-affirmative action world, other states might balk at such investments.
Electoral politics are another matter. Those who favored Proposition 16 blamed their loss on confusing ballot language, the difficulty of campaigning during the Covid pandemic and too little voter education.
Ruy Teixeira, a political scientist, takes a different view. He noted polling consistently demonstrates the unpopularity of race-conscious affirmative action.
A Supreme Court death knell, he said, might save Democratic leaders from themselves, untethering them from affirmative action.
“For years, they have said, ‘We must positively discriminate,’” he said. “Maybe they no longer need to die on that hill.”
Ruth Igielnik contributed data reporting. Alain Delaquérière contributed research.
To estimate how demographic groups voted on Proposition 16, The Times combined precinct-level election results from the Statewide Database; a voter file provided by L2, a nonpartisan data vendor; and estimates of the citizen voting-age population by race and ethnicity at the census block level as compiled by the ALARM Project at Harvard University. Those results were then analyzed using multiple methods to determine whether support or opposition to the proposition was tied to factors including the racial and ethnic makeup of each precinct. The analysis included using the eiCompare R package to perform ecological inference using multiple methods; reviewing voting patterns where an ethnic group made up at least 60 percent of the voting population; and regression analysis.
While analyzing precinct-level results can help better understand voting patterns and trends, the conclusions are limited in that there is no way to know how individual voters of certain races or ethnicities voted.
