The Conversation About Which Voters Are to 'Blame' Makes Me Want to Stick My Finger Through My Eyeball

This is not a morality play, there is no demographic Electoral College, and exit poll data is garbage.

Dear readers,

One of my least favorite parts of the immediate aftermath of any election is the argument among Democrats over demographics: whose votes we lost and why. This is not because I hate talking about demographics — I love talking about demographics. It’s because the way liberals talk about the demography of the vote after the election is routinely very stupid.

The first problem with the conversation is that it’s built around shitty data: The exit polls are unrepresentative of the actual electorate — they’re representative of who likes to talk to exit pollsters — and they’re especially unrepresentative before they’ve been adjusted to align with the overall election result. So commentators pick up the initial, unadjusted exit polls and make impassioned arguments based on numbers that turn out to be wrong. One “finding” I keep seeing people cite from this year’s initial exits is that Harris gained eleven points compared to Biden 2020 among voters in households with incomes over $100,000. If you even glance at the electoral map you can see this must be wrong — if Harris surged ahead with affluent households, how did she underperform Biden almost everywhere, including wealthy places? How is it that her margin of victory in Greenwich, Connecticut was 9 points narrower than Biden’s four years ago? The real answer seems to be that last cycle’s exit poll was wrong — Biden ran much better among affluent voters than the exit poll found — and so pundits are comparing old garbage to what is quite possibly new garbage and generating a garbage statistic to throw around like it’s the truth.

Several months from now we will have good data from the Census Bureau and Catalist and the Pew Research Center, and we’ll be able to make more accurate assessments of how the vote moved among various demographic groups. But by then, it will almost be too late. Any false narratives forming now (while people are desperate to understand what went wrong) will have gotten baked in to the cake; they will have already shaped people’s views about how to run the next campaign.

ChatGPT image generation for the prompt “blame game,” with subsequent nudging about making the game a board game.

The second reason I hate this way-too-premature argument over whose votes we lost is that people turn it into some sort of morality play and therefore get inappropriately emotional about it. If you talk about Kamala Harris losing disproportionate ground with Hispanic voters — which she clearly did; you see this from the maps without even consulting the exit polls — that somehow becomes blaming Hispanics for her loss, which is unfair because white voters were actually a lot more likely than Hispanics to vote for Trump. There’s a weird “demographic electoral college” element to these arguments, as though white women have 199 electoral votes and Harris needed a few more points of their support to clinch those. But it is not “white women” or “men” or “Latinos” or any other demographic that caused Harris to lose — it is the entire coalition of voters that voted against her, which comes from all different demographic groups, just in different proportions.

The third reason I hate the argument over demographics is why people (by which I mean liberals) turn it into a morality play: they don’t like the implications when you look more dispassionately at what has happened. Donald Trump, the master offender who launched his first presidential campaign by declaring that Mexico is sending us rapists, has somehow presided over a major racial depolarization of American politics. He has been far better than past Republican presidential candidates at winning non-white voters. The reason for that appears to be that voting behavior is becoming more aligned to ideology — the subset of Asian and Hispanic and black voters with more conservative political attitudes are increasingly choosing the more conservative candidate. In a way, Elie Mystal is right — “the ‘solidarity’ between ‘people of color’ has been significantly damaged,” inasmuch as non-white voters are apparently becoming less inclined to set ideology aside in order to vote in a bloc. Where he goes wrong is treating that as something that separates Hispanic voters from black ones — Trump’s inroads with non-white voters are extending into the black community, too, though not to quite the same degree as with Hispanics.

The problem for liberals with admitting that this racial depolarization is happening, and that we should probably find ways to win back the votes we’re losing from non-white voters, is that the problems can’t be addressed with their favorite playbook. Saying that Trump is racist doesn’t seem to do anything about this problem — in fact, Trump himself saying racist things doesn’t seem to do anything to address the problem. Accusing non-white Trump voters of “identifying with whiteness” or “pulling up the ladder behind them” is definitely not going to help matters. To win these voters back, we’re going to have to compromise on ideology — that is, take them seriously as individuals with policy preferences that are not necessarily left-wing, rather than assuming we can get their votes because of the color of their skin.

The funny thing about the “blame” frame is that it allows liberals, who usually talk about how important it is to listen to non-white people, to disregard the preferences of the increasing share of non-white voters who have left the Democrats.¹ To say that black Americans have done their part and it’s on whites to do their part is effectively to say that black voters should not be courted — that Democrats should focus their campaign on winning more white votes next time. I think some of the people who talk this way do so because they envision a white electorate that behaves like the progressive white people around them — that is, one that is interested in listening to lectures about “whiteness” and how to atone for it — but of course, if marginal white voters were like that, Trump would have lost.

In any case, I look forward to having the good data in a few months so we can take a close look at where and how Harris lost, so that Democrats can make good decisions going into 2028 about how to regain ground with the demographic groups where we’re slipping. But in the meantime, we’re going to be stuck with these dumb takes.

---

By the way, I have a column for The New York Times today about how Joe Biden set us all up for failure in this election. You can go read it here.

It’s really very frustrating — as you know, I was a big fan of Biden’s, and I felt he was the right person to nominate in 2020. I touch on this briefly in the piece, but he really seemed to take leave of his senses once COVID hit and he locked up the 2020 nomination. The whole theory of his campaign until that point had been that Democrats needed to chart a moderate course to build a broad coalition, and then the pandemic happened and he became convinced that he had a mandate to move left and be a “transformational” president like Franklin Roosevelt. He also picked a running mate who had made all the ideological mistakes in her losing campaign that he was wise enough to avoid in his. This all positioned the Democratic Party very poorly when it became necessary to defend the presidency in an unfavorable political and economic environment and it became clear that his faculties were declining and he could not be the candidate to do so. It was such a display of hubris, and it wasn’t necessary — if only he’d picked Amy Klobuchar for the ticket, governed as the moderate he was in the primary, and then announced after the midterms that he wouldn’t run again, we’d probably be on the way to inaugurating President Klobuchar right now.

What a mess.

Very seriously,

Josh

1. Gustavo Arellano has a very good column about this for the Los Angeles Times. Don’t get distracted by the headline; he’s not one of the commentators with his head in the sand.