Voters cast ballots against affirmative action while still favoring relaxed versions of it

Evidence

A Field Poll conducted several days after the vote found that fewer than half of opponents agreed that “affirmative action policies should not be changed.” A third of the opponents further agreed that “affirmative action policies should be relaxed somewhat.” Conversely, about a quarter of the proposition’s supporters agreed that affirmative action policies should not be changed, or should be relaxed but not eliminated (Lempinen, 1996). Some of these voters may simply have been confused about what they were voting for. But many apparently would have preferred a middle ground to the two stark alternatives they were offered. — David Theo Goldberg, John Solomos, A Companion to Racial and Ethnic Studies, 2021

Categorization

  • XREF: Connects to the broader literature on how public opinion on race policy is ‘race-coded’ and multidimensional; also echoes findings in the Bradley Effect / symbolic vs. operative racism debates. — connects to: Voter initiatives countered Grutter while the ruling still guides other states — G. Alan Tarr, Judicial Process and Judicial
, 668
  • QUESTION: The ‘confused voters’ hypothesis — worth chasing how the California ballot-language framing (two stark alternatives) distorted expression of genuinely middling preferences.