A Very Quick Comparison of 2024 and 2020 Exit Polls
What they suggest about the shifting party coalitions
My impression is that the key narratives developing around the 2024 presidential election point to widescale cross-national trends that are working both against incumbents generally and against the educated liberal elite in particular.
I don’t have a hot take on those accounts, but I can say that both are roughly consistent with what I’m seeing in the available results from the Edison Research exit polls (used by NBC, CNN, etc.) and the AP/Fox voters surveys run by NORC.
Here’s a look at 2024 compared with 2020. You can see that there’s a Republican-leaning shift for most groups. This is consistent with some large-scale force pushing against Harris and in favor of Trump – could relate to general anti-incumbent sentiment, or to anti-liberal-elite sentiment, or to some kind of general weakness in Harris and her campaign relative to Trump and his campaign.
We can also see that some groups moved more than others and, indeed, some groups even managed to swim against the tide and shift left. I’ve tagged the ones that have the biggest Republican shifts, and I’ve also tagged the biggest shifts to Democrats (some of which are D+0, where the group merely marginally favored Democrats in 2024 relative to 2020 – at the least, they were groups that resisted whatever general forces were pushing most groups to the right).
The biggest Republican shifts:
Latino men
Ages 18 to 29
No college attendance
Non-whites without 4-year degrees
Latina women
Black men
Those with incomes less than $100k
The biggest Democratic shifts (or, at least, the smallest Republican shifts):
LGBT
Those with incomes above $100k
Secular whites and Jews
Ages 65+
White men with 4-year degrees
White women with 4-year degrees
People with advanced/graduate degrees
I’d group some of these together. LGBT and secular whites and Jews have in common (1) that they are most liberal on issues relating to discrimination on the basis of sexual orientation and religion, as well as on sexual and reproductive issues such as abortion. At higher education levels, these groups tend to produce full-on across-the-board liberals. At lower education levels, they often combine being sexual/religious/reproductive liberals with being more conservative on issues of race, immigration, and income redistribution.
Contrast these groups with Latino men, Latina women, and Black men – all of which appear to have had larger shifts to the right. These minority groups often combine being rather liberal on issues of race, immigration, and income redistribution with being more conservative on issues relating to discrimination on the basis of sexual orientation and religion, as well as on sexual and reproductive issues such as abortion.
This tension has long existed within the Democratic coalition between prioritizing sexual/religious issues and prioritizing racial/economic issues. Some of these shifts in 2024 suggest that that tension is being resolved in favor of Democrats’ focus on sexual/religious issues.
And then there are big shifts (and non-shifts) that look to be more straightforwardly about socioeconomic status. Democrats gained ground in 2024 with high-income folks and highly educated folks (especially whites). Republicans had their biggest gains with groups that have the least education and less income (especially racial minorities). It seems roughly plausible to me to combine the socioeconomic story with the LGBT/secular story to tell one big story about how the working class continues to expand its revolt against elite liberals. It also seems plausible to me that it’s really two things going on simultaneously: (1) A realignment in which Democrats are attracting secular liberals who lean conservative on race/economics at the expense of Republicans attracting race/economic liberals who lean conservative on sexual/religious issues, and (2) a general trend where those least protected from the economic harms of the past few years moved against incumbents and/or liberal elites while those most protected from economic challenges and have been doing quite well tended to support Democrats.
The shift of young people to the right simultaneously with the small shift of seniors to the left might factor in here, in favor, I suspect, of the second story, but I’ll need to get individual-level data to really sort that out.