Trump and the trend

Almost everyone who wrote about the 2016 and 2020 elections said that Donald Trump appealed to the "white working class."  In the Washington Post, Nicholas Carnes and Noam Lupu argue that almost everyone was wrong:  the headline is "Trump didn’t bring White working-class voters to the Republican Party. The data suggest he kept them away."  A smaller heading under that says "White working-class voters had been moving to the Republican Party for years. Trump stopped the trend."  Those headlines are a bit misleading (they were probably written by editors)--according to their figures, Trump's 2016 share of the white working class vote was the highest of any Republican in the period covered by their data (which in 1980), and his share in 2020 was the second highest.  But his strong performance was just a continuation of a long-term trend--2016 and 2020 don't stand out as unusual.  

I looked at this issue in a  couple of posts, and said that there was a long-term trend in the effects of college education on vote, but that 2016 was unusual--Trump's relative performance among less educated voters was much better than any previous Republican.  Why were my conclusions different?  One difference is that I looked at relative performance--Carnes and Lupu looked at actual share of the "white working class" vote.  Another is data sources--they used the American National Election Studies and I used exit polls.  The ANES data are generally regarded as of higher quality (they are a national probability sample and have a high response rate) but the samples are smaller, so they are more affected by sampling error.  Finally, Carnes and Lupu take account of both education and income:  they define the working class as people without a college degree who are below the median income.  Taking account of income is good, but cutting the size of the "working class" also means more sampling error.  The result is that there's a good deal of noise in their estimates, making it harder to see if any election deviates from the trend. 

An alternative way to take about of both income and education is to estimate the effects of both together.  I did this using the General Social Survey, partly because the samples are usually larger, but mostly because I'm more familiar with that data set.  I did logistic regressions of Democratic vs. Republican votes (omitting third parties) on a dummy variable for college graduates vs. all others and the logarithm of household income (adjusted for inflation).  The estimated effects of college degree:

Higher values mean that people with a college degree are relatively more likely to support the Democrats and people without are relatively more likely to support the Republicans.  There is an upward trend, but a clear jump in 2016--it's well above the predicted values from a time trend fitted to the 1968 to 2012 data.  

Now the effect of income:  all values are negative, meaning that higher income goes with support for the Republican candidate (controlling for education).   There is no apparent trend--the highest value (greatest Republican appeal to low-income voters) was in the 1968 election, but that was a small sample.  The 2016 value was not unusual.  

 

 So Trump had unusually strong relative appeal to voters with less education, but not to voters with lower incomes.  Rather than trying to translate that into a generalization about the working class, I'd just talk about education and income.  

    The "relative appeal" connects to my first point about the Carnes & Lupu analysis.  Trump got about the same share of the vote as Mitt Romney had in 2012, but he did better among people without a college degree and worse among college graduates.  Of course, that's not all because of the difference between Trump and Romney--they had different opponents and were running under different circumstances.  The comparison we really want is to an "ordinary" Republican running against Hillary Clinton in 2016, and of course we can't know that.  But still, it would seem reasonable to start by assuming that the difference was partly appeal to less educated voters and partly repelling more educated voters. Most of the commentary seemed to focus on the first part and ignore the second.  So although I disagree with Carnes and Lupu on one point--the distinctiveness of 2016--I think they made a contribution in raising questions about whether Trump had a special appeal to the "working class."

 

 Note:  if would be reasonable to focus on groups defined by a combination of income and education if there was some interaction between them--more exactly, a three-way interaction so that income/education groups shifted position from one election to the next.  There is some evidence of an interaction between income and education--income seems to have more effect among people without a college degree--but no evidence that the relationship changes from one election to the next.  That is, the effect of education changes, the effect of income changes, but there's no evidence that the effect of the combination changes.