Democrats' strange victory in the 2022 midterms. By Matthew Ygesias

Democrats' strange victory in the 2022 midterms

The normal backlash materialized — but not in the key races

Interest in election analysis tends to peak the day after the election, which is unfortunate because even under modern conditions, we don’t have accurate vote counts that quickly. Hot takes based on exit polls with very low reliability and impressions tend to congeal quickly.

For my money, the best source of information about who voted for whom is the “What Happened” reports that Catalist puts out months after voting is done. Because Catalist isn’t a media company that’s trying to maximize clicks, they take their time and match survey data with data from their extensive voter file. Their “What Happened 2022” report is now available and I’d recommend that anyone who’s interested in U.S. elections read it.

What really stands out to me this year is something that I think is a little unfashionable but hard to deny: Democrats did a really, really good job, on a technical level, at the blocking and tackling aspects of running effective political campaigns. They raised a lot of money, both from small donors and through large super PACs, and they seem to have been quite effective at translating that money into ads and other campaign messages that voters in key states and districts found persuasive. There are limits to how much this can accomplish, but in close races, it matters a lot.

A different kind of non-wave

After successive down-ballot landslides in 2006, 2010, 2014, and 2018, the notion of a “wave” election became so ingrained in the discourse that I think we sort of lost sight of what a non-wave election might look like. But to understand why observers were initially so impressed by those waves, consider a map of the 1998 Senate elections, which were a wash even though six seats changed hands.

To an extent this is just a different era, one in which Democrats held seats in Arkansas, Louisiana, and South Carolina. But it’s important to note that even back in 1998, South Carolina was more GOP-friendly than Pennsylvania, New Hampshire, Arizona, or Georgia. Democrats gained a seat in Indiana while losing one in Illinois. Bill Clinton won Ohio in 1996 and lost North Carolina, then two years later the Ohio senate seat flipped blue and the North Carolina seat flipped red.

This was, in other words, a non-wave election. The national political climate mattered (the typical midterm backlash was neutralized by a sense that the GOP was overreaching with impeachment), but you can’t really explain the specific results in terms of national trends. There were strong incumbent effects, local news coverage was very important, voters were less cognizant of big-picture partisanship, interest groups did more cross-endorsements — politics was, generally speaking, very complicated. Sometimes a big partisan “wave” would crash across the country (as in 1994) and everything would break one way. But elections would frequently turn on the specific dynamics of something like a politician’s personality.

Or for a truly classic non-wave, look all the way back to 1984 when Ronald Reagan won in a crushing landslide and Democrats gained two net seats in the Senate.

That was a long time ago, but it’s important to understanding the development of the concept of the “wave” because it’s not like 1984 was a close election. It was a huge landslide. But it wasn’t a wave. Down-ballot voters were picking based on candidates or matchups or local concerns. In part that’s because it was a different time. But in 1980, Reagan did have long coattails that generated a wave, and in 1982 there was a midterm backlash wave. So it’s not like waves never happened; it’s just that they only happened sometimes.

The non-wave of 2022 differed in several key waves from those older non-waves. In 2014, Democrats did 4 points worse nationally compared to 2012. Then in 2018, Democrats did three points better nationally than they’d done in 2016. In 2022, Democrats did three points worse nationally than in 2020. Same old, same old. But what made 2022 different, per the Catalist data, is that Democrats did way better than average in precisely the closely contested races where it mattered most.

That’s highly unusual. In both 2018 and 2014, the swings were uniform. The tossup districts are very different from the base districts, but there are a certain number of moderate, cross-pressured, or otherwise persuadable voters in any given district. And in recent elections, the persuadable voters have tended to lean one way or the other basically everywhere. But 2022 wasn’t like that. In the safe districts, a healthy number of Biden voters flipped to the GOP. But in the swing districts, they didn’t. Because most seats are safe, the overall national result (-3) is much closer to the safe seat result (-4) than it is to the tossup seat result (0). But because the tossup seats are the ones that actually matter, the outcome wound up being really strong for Democrats.

Changing women’s minds

So who was persuaded by Democrats’ campaign messages in the swing states? Looking at the national House vote shares by race, Democrats were basically flat with Hispanic voters but lost ground with white, Black, and Asian voters relative to 2020. In the highly contested races, they did about the same with Hispanic voters, slightly better with Black voters, and quite a bit better with white and Asian voters.

Catalist didn’t do gender-by-race breakouts for all groups, but they did for white voters. And while Democrats improved with all four groups below, it was white women specifically — both college graduates and working class — who were much more Dem-leaning in the highly contested races than in the national house races.

These non-college white women are a crucial factor in American politics. People who are white or male constitute a large majority of the American electorate, whereas those of us who are white and male are a distinct minority. Whether intersectional politics actually captures the intersection of its intended beneficiaries matters enormously in terms of whether it’s electorally viable. White women are the largest cross-pressured group in terms of identity politics, and the fact that non-college white women (recall that non-college outnumbers college graduates for all groups) have been relatively conservative in the Trump era is a big practical problem for progressives. If you want to make a feminist pitch, you need one that appeals to working-class white women, too.

To take a random recent example, Karen Attiah wrote a column for the Washington Post with the headline “Guns and abortion laws have made Texas a woman’s nightmare.” I think Texas gets some big stuff in public policy right, but I wholeheartedly agree that the state’s approach to guns and abortion rights is really bad.

But what share of Texas women agree with Attiah and me about this? Greg Abbott’s approval rating is decent. Polling from Pew shows that women in Texas are pretty evenly split over abortion, but that a slight majority favor it being illegal in “all/most cases.” More generally, polls tend to show that men and women don’t differ that much in terms of their views on abortion, but that women care more about abortion

1

(in both directions), and I think that helps explain the midterms. Dobbs was unpopular, banning abortion is unpopular, and where Democrats were able to invest heavily in abortion rights ads, it swung a decent chunk of women into their column.

Of course, not all Democratic ads were about abortion. I wrote last December about Democrats’ best-performing messages, and they really are worth looking at. From a progressive point of view, these messages are mostly compatible with progressive stances on policy issues. But I really do think it’s noteworthy that none of them specifically invoke the progressive stance on a topic that is high on the agenda for progressive advocacy groups. It’s not just that these messages don’t involve defunding the police or talk about race using the vocabulary of a foundation program officer, it’s that they don’t focus on cutting carbon dioxide emissions, expanding subsidies for child care, or as far as I can tell, any of the other things that are driving the progressive policy agenda. That’s fine. Republicans don’t run ads about how they want to cut taxes for billionaires. But I do think anyone who works in or posts about progressive politics should note the contrast between Joe Biden’s 2024 launch video on Twitter (designed to appeal to small donors and build the email list) and his first television ad (reflecting Democrats’ ad testing).

The prospects for 2024

Of course, what everyone really wants to know is what this all means for the 2024 presidential election, and the answer is… it’s really hard to say.

But it’s helpful to draw a distinction between micro-politics and macro-politics. The primary takeaway from the 2022 race is that Democrats put on an impressive micro-political performance — they raised a ton of money, and they spent that money on campaign messages that were well-crafted and that minimized electoral backlash in key races. Republicans, meanwhile, nominated some real stinkers, especially for the Senate. But on macro-political questions like “are people happy with Joe Biden?” Democrats’ performance has so far been much less impressive. After a solid start, Biden has slumped to Trump-esque approval ratings. I don’t think “Joe Biden’s approval rating in May of 2023 is bad” has any real predictive value for November 2024, but it does suggest a bit of an uphill battle.

One piece of good news for the White House is that “Trump-esque approval numbers” aren’t an insuperable obstacle if you are actually running against Trump. But still, they should try to become more popular.

More substantively, the Harvard Cooperative Election Survey has this interesting result: the electorate as a whole places themselves in the middle of the ideological spectrum with the GOP too far right, the Democratic Party too far left, and the Democrats slightly more off-center than the Republicans. But Joe Biden is seen as more moderate than the average Democrat. And while Trump used to be seen as unusually moderate for a Republican, he lost that brand over time. So Biden’s basic brand is good.

This is also why, though I think his age isn’t ideal, it’s not the problem it’s sometimes made out to be. If Biden could de-age by 20 years, he’d be able to do more public events and have a stronger presence in the media, which would be helpful because people like the Biden persona more than they like the generic Democrat. But swapping him out with someone more generic would be bad. And I think the fact that he’s old helps convey moderation — people assume he has the ideas of an old guy, not the ideas of a history grad student.

I think, though, that this does mostly come down to “it’s the economy, stupid.”

In a lot of respects, the economy is doing well. Unemployment is extremely low. Inflation-adjusted GDP is about 5% higher than it was before the pandemic. Nobody is happy about inflation, but while things are more expensive than they used to be and some people have cut back as a result, people aren’t generally getting by with less than before — most people have more stuff than ever.

So what gives? Well, political memories are short. People got a lot of pandemic assistance and they were pretty thrilled with it. At the same time, consumption plummeted during the pandemic thanks to a mix of restrictions and voluntary restraint.

2

So people came out of the pandemic with extra money on hand, they’ve been spending that extra money down, and as a result, prices have been rising faster than incomes. Net-net, it averages out to higher levels of production and consumption than ever before. But nobody today is crediting Biden for money he helped them get two years ago, and the recent trend has been bad.

The good news is that in the very most recent month or two, year-on-year wage growth was back in slightly positive territory.

Because central banks care a lot about inflation, there’s a lot of inflation talk and speculation about the future course of inflation on econ Twitter and in op-ed columns. But you can have positive real income growth with inflation at 5% and you can have negative real income growth with inflation at 3%. In terms of quality of life and the overall political mood, I have to think whether inflation-adjusted income is going up or not is a bigger deal than the inflation rate per se. If a year from now, unemployment stays low and real wage growth has improved, then I think Biden can paint a very happy picture of the Biden economy. If we get a recession or if prices keep rising faster than wages, then things look bleak.

So what’s going to happen? I don’t know — but I would really emphasize that while Democrats are of course always honing their economic message, they also really should do everything they possibly can to bolster supply-driven growth. It’s always fashionable among the commentariat to paint Democrats as politically inept or leaping from blunder to blunder, but the bulk of the evidence indicates that they’ve done a remarkably good job at the micro-politics in an objectively difficult situation. What’s needed, I think, is a bit more ruthlessness about economics vs. interest group desires.

1

For example, when the parties were polarizing on abortion during the 1980s and 1990s, it was more common for men to “follow the leader” and change their stance on abortion to align with their party’s stance, whereas women were more likely to switch parties to join the one they agreed with about abortion.

2

This is a topic for another day, but I think a lot of people have started misremembering exactly how Covid-cautious they were personally and attributing everything to formal rules.