The electorate is becoming less racially polarized

Demographics isn't destiny

I was doing takes back in 2019 and 2020 about Donald Trump’s successful efforts to court Black and Hispanic voters, so I wasn’t surprised when he improved his numbers with non-white voters in the last election.

But the message was poorly received by progressives at the time, and there’s been a lot of denials ever since, to the point that things like this week’s New York Times poll showing further Trump gains with Black and Hispanic voters come as a shock in certain quarters. I think that’s largely because this trend deals a blow to many progressives’ self-image in a way that’s out of proportion to its objective electoral significance. But self-image is important, and I want to make a few related points about these findings:

1. Every pollster sees broadly the same trend toward racial realignment, but not every pollster sees it on the same scale as the NYT — this is part of an interesting larger conversation about methodology.

2. If the Times’ findings are correct, they are broadly good news for Kamala Harris, in the context of her small-but-steady lead in national polls.

3. It’s probably also good news for American society if voting patterns are less stratified by race.

4. Voters of color moving toward Trump is, in part, the result of the deliberate strategic choice by Democrats to dramatically increase the extent to which they run on abortion rights — a choice that, on the net, is smart.

5. Notwithstanding my relatively complacent attitude with regard to 1-4, the rightward shift of voters of color is genuinely difficult for many progressives to reconcile with their big picture view of American politics, and it underscores the need to try to be chill and assemble a big tent to win.

Before we delve in, though, it’s worth recalling that not only are most Americans white, but white Americans vote at a higher rate than non-white ones (mostly because we are older and better educated) and are also overrepresented in key electoral college battleground states. Play around a little with the 538 Swingometer and you can see a scenario in which Trump gets to dead even with the Hispanic vote (versus losing by 37 points in 2020), but in exchange, Harris gains two points with white voters:

Or imagine Trump gaining a staggering 38 points with Black voters, but losing three with white ones:

There’s just a lot of white people in America.

Black Democrats have some conservative opinions

Especially with regard to African Americans, the basic dilemma facing Democrats is that Joe Biden’s 90 percent vote share — to say nothing of the even higher numbers achieved by Hillary Clinton and Barack Obama — is hard to sustain indefinitely.

Black Americans, like all groups of Americans, hold varied opinions on the broad range of topics discussed in national politics. There are even some questions where Black opinion is, on average, more conservative than white opinion. The General Social Survey, for example, asks the question, “And what about a man who admits that he is a homosexual? If some people in your community suggested that a book he wrote in favor of homosexuality should be taken out of your public library, would you favor removing this book, or not?”

Very few people favor censoring this book, but the censorship option is twice as popular among Black respondents as among white ones.

What’s particularly striking is that among Black voters, censoring gay books is dramatically more popular than the Republican Party!¹ If you want a nuanced exploration of why many African-Americans with fairly conservative views on a lot of issues vote for Democrats anyway, I strongly recommend the book Ismail White and Chryl Laird published on this subject in 2020. If you want a not-so-nuanced explanation, though, it’s that they think Republicans are racist.

Note that the responses to the book question reflect not an idiosyncratic view about public libraries, but the fact that Black Americans are generally more skeptical of homosexuality.

A virtue of the GSS is that it has long time series data on these kind of questions, but that means that it unfortunately does not have questions about more contemporary topics, like puberty blockers or who should play on which sports team. But while I would say these issues have not proven to be the big vote-getters that Republicans keep hoping for, it’s precisely among non-white Democrats that you’d expect to see inroads made. This is even more pointedly true, though, of another issue — abortion rights — where it’s Democrats who’ve been telling everyone that it’s incredibly important to vote for them in order to secure women’s right to reproductive autonomy.

Abortion rights depolarizes the electorate

Republicans have been hoping to weather the Dobbs backlash by pointing out that Democrats have fairly extreme views on abortion themselves. The problem for the GOP is that the public has become much more liberal on this topic over the past 10 years, and abortion should be legal “if a woman wants it for any reason” is now a majority position.

In this case, it’s not that Democrats’ view on abortion is unpopular with Black voters; it’s not even that it’s less popular with Black voters than it is with white ones.

It’s just that the baseline is really different. A Democrat who got 55 percent of the white vote would be turning in the greatest performance with that demographic since Lyndon Johnson. But a Democrat who got 56 percent of the Black vote would be turning in the worst performance with that demographic since at least Adlai Stevenson and probably Al Smith. With Hispanics, meanwhile, “legal for any reason” is underwater at 45 percent.

To be clear, I am not saying that Democrats are making a mistake by emphasizing this issue. If you match the Swingometer to the pure popularity of “abortion legal for any reason,” Harris wins handily. This is a good campaign strategy.

But tradeoffs are real in the electorate as well as in policymaking. By drawing attention to abortion rights in 2024, Democrats push up their appeal with white voters relative to baseline, and push down their appeal with Black and Latin voters relative to baseline. With Hispanic voters, that’s because the position is unpopular. With African-Americans, it’s just because the baseline is so incredibly high.

The problem of subtractive intersectionality

As I said at the top, narrowly construed, the NYT’s thesis that Democrats have lost a lot of ground with nonwhite voters is good news. That’s because national polls normally show Harris with a 2-3 point lead in the popular vote. That’s a decent lead! The reason the election is razor close is that Harris is expected to face a big disadvantage in the electoral college. But Michigan, Pennsylvania, and Wisconsin are all whiter than the US population average², so if it’s true that Harris is doing unusually poorly with non-white voters because she’s emphasizing issues like abortion that increase her appeal to white voters, that implies her electoral college handicap has diminished.

I am not 100 percent confident that this is actually what’s happening. The New York Times polling is showing something slightly different from many other pollsters, but that’s because of a methodological choice. My guess (and it’s just a guess) is that the Times is getting it right.

The issue here isn’t that the loss of non-white support is per se a huge crisis. But I’ve written extensively over the years about the progressive mobilization myth — the delusion that there’s some public opinion cheat code that involves transforming the electorate or counting on the demographic tides to sweep conservatives into the dustbin of history. This might work if all coalition politics were purely additive, if you could give feminists abortion rights and Hispanics lax immigration policies, and criminal justice reform to African-Americans and the PRO Act to union members and student loan relief to recent college grads, and just add all of that up to get to 55% percent of the electorate that way.

In practice, though, things can go in the other direction. If you push too hard on climate change, you can alienate low-income voters who place a higher value on material prosperity. If you do what LGBT groups or feminists want, that may put you into conflict with culturally conservative Black and Hispanic voters. Jews and Muslims both affiliate with the Democratic Party because we value pluralism as minority groups, but we tend to disagree about Israel/Palestine. You can sometimes manage these tensions by giving more to various groups, but sometimes what it takes is to do a bit less. My basic view is that the best way for Democrats to do better with nonwhite voters isn’t really to do more group-specific pitches, it’s to be a bit more moderate on climate and sex and gender issues. Similarly, the best way to do better with blue-collar union members probably isn’t to triple-down on pro-union stuff (Democrats are clearly the more pro-union party), it’s to be more moderate on issues related to race, guns, and immigration. Less can be more!

But beyond that, the super high level of Black support that Democrats have traditionally enjoyed is the flip side of American society being highly racist and segregated. You ultimately want to see a less racially polarized electorate, with people voting less on identity.

A hopeful future

Former Slow Boring intern Milan Singh has been busy back at school, among other things launching the Yale Youth Poll, which is now the largest survey of young voters in America.

It shows Harris up 10 points with Hispanic voters under 30, up 12 points with white voters under 30, and up 50 points with young Black voters.

Compared to the national electorate, those are terrible numbers with non-white voters, but incredible numbers with white ones. It’s a vision of an electorate that is dramatically less divided by race than the traditional American electorate. It skews to the left because young people are more liberal than the general public on LGBT issues and climate change. But the minority of young people of color who have conservative views seem to be comfortable voting Republican, while young white people are much less right-wing than old ones.

This is ultimately a much more hopeful and constructive vision of politics than the kind of never-ending race war that both “emerging Democratic majority” triumphalists and “great replacement” paranoiacs have spent a lot of the past decade fighting about. And it reflects the fact that younger people are growing up in a less racist, more integrated society, one in which white people are more open to progressive ideas but nonwhite people feel comfortable affiliating with the right if that’s who they agree with more on policy. If this ends up narrowly costing Harris the election, I’ll be very sad about that. But as a broad social trend, it’s change for the better.

1. The question is phrased in a very weird way, but I think the racial gap in responses is still illustrative of a difference in underlying attitudes.

2. Because the four largest states — California, Texas, Florida, and New York are all more diverse than average, the majority of states end up being whiter than average.