The Monkey Cage

Chapter 3 of UNEQUAL DEMOCRACY reports estimates of the impact of various issue preferences on presidential votes among high-, middle-, and low-income white voters. The estimated weight of abortion from 1984-2004 is twice as great for whites in the top third of the income distribution as for those in the bottom third. The increase in the estimated weight of abortion over those 20 years is substantial for the top third and zero for the bottom third. And in 2004, the estimated weight of a “cultural issues” scale including abortion (along with gay marriage, gender roles, etc.) is twice as great for the top third as for the bottom third.

If you prefer education as a measure of “elite” status, my 2006 piece in QJPS on “What’s the Matter with Kansas?” reports parallel analyses differentiating white voters with and without college degrees. From 1984-2004, the weight of abortion is twice as great among people with college degrees. The increase in the estimated weight of abortion over that 20 year period is almost twice as great among people with college degrees. And the estimated weight of abortion in 2004 is four times as great among people with college degrees.

Larry,

Thanks for pointing this out! I read the manuscript of your book (and even cited it in Red State, Blue State), but I hadn’t remembered that analysis, which is definitely relevant to our argument and also consistent with the findings of Ansolabehere, Rodden, and Snyder. Also consistent with what Baldassarri and I found looking at trends in correlations between issue attitudes and partisanship.

But do you really want to say “impact” above? Perhaps “correlation” or “predictive power” or something like that would be safer?