The 2022 Electorate Was the Most Educated in American History
According to a Census Bureau survey, the 2022 electorate was the most educated in American history. 74.1% of the electorate reported that they either had some college education, graduated college, or had post-graduate education.
What does this mean for future elections?
A simple trend shows that the share of the electorate has steadily climbed from about half of the electorate in the mid-1980s to about three-quarters today.
There is some variation between midterm (blue dots) and presidential (orange dots), with a slightly more educated midterm electorate than presidential electorate. Both types of elections have been on an upward education trend.
Education is highly correlated with turnout, with higher educated voters having two or more the turnout of lower education voters. The slight relative increase in the share of the higher-educated electorate in a midterm election has to do with relative turnout rate changes of the groups of differing education levels. For example, turnout rates for lower education eligible voters decreased by a little under half between 2020 and 2022 while they decreased by about only about a quarter for higher education eligible voters.
The trend of a more educated electorate is occurring against the backdrop of more people in the general population seeking college education. It should not come as a surprise that as more people seek a college education we should see the electorate become more educated as well.
A safe prediction, then, for the 2024 presidential election is that the percentage of the electorate that is college-educated will continue the upward trend, but will fall between the share present in 2020 and 2022. My guess would be about 75% of the 2024 electorate will have some college, be a college graduate, or a post-graduate education.
This matters for two reasons.
First, the more wonky election polling reason is that a number of people have rightly criticized election polls has having respondents who are too-educated, throwing off candidate horserace numbers. One way to correct for this problem is to weight a survey to census education numbers. I support this approach, and I further recommend interpolating forward education levels since there is a clear upward trend.
One may wonder if this education bias is present in the census numbers. Likely it is not, or at least not to a large degree. The Census Bureau expends considerable resources to ensure the Current Population Survey is representative of the United States population because the primary purpose of the survey is to estimate the county’s unemployment rate — the voting and registration supplement is simply a byproduct. The Census Bureau conducts face-to-face interviews, makes repeated contact attempts, and there is actually a federal law that requires people to respond to the survey. Election polls rarely have resources to even make several call-back attempts to unreachable people on the first or second try. This is not to say there are not problems with the census survey, just that the survey is likely providing a better estimate of education levels than a typical election poll.
Second, is the political ramifications.
According to the exit polls, in the 2004 election, Bush and Kerry tied among college graduates, both receiving 49% support. In the 2020 election. Biden received 50% support among college graduates to Trump’s 43%.
Now, the exit polls are like other election polls in that they are not weighted by education and thus may have a pro-education bias. Still, any number of polls and measurement approaches confirm the shift in support among education levels in the Trump era. As Trump said, “I love the poorly educated.” He and Gov. DeSantis have doubled and tripled down on appealing to non-college educated voters by portraying colleges as bastions of wokeness. This anti-intellectualism may appeal to the Republican base, but it is performative theater before a shrinking audience (both Trump and DeSantis are Ivy-league educated). It is driving away higher education voters, and this has consequences not only because these person continue to increase their share of the electorate but also because these are the people most likely to vote.
Being against education may be a winning strategy for a Republican primary, but it is increasingly untenable for a general election.