Expectations for the 2022 midterms were set early on. The Democratic Party was supposed to easily lose control of the House, and potentially the Senate, as a result of lackluster approval ratings for Joe Biden and concerns about issues like crime and inflation.
But then came the end of Roe v. Wade. Tom Bonier, a Democratic strategist and CEO of TargetSmart, determined that the Supreme Court decision on abortion was having a significant effect on voters and that a potential Dobbs effect could increase turnout and combat the predicted red wave. After Democrats pulled off unprecedented wins for a party in power, Bonier’s hypothesis seems to have been validated.
I recently spoke with Bonier about his initial impressions of the midterm results, what impact he believes the Dobbs decision had, and whether the polling was better than it was in 2016.
It’s been several days since Election Day, and more of the returns have come in. What is your initial takeaway from the results we’ve seen?
My takeaway at this point is that we saw the trend starting after the Dobbs decision that was initially evident in Kansas, where the pro-choice vote won by about 18 points. Then what we saw in the data was this massive gender gap of women registering to vote. We were seeing younger women, especially, voting at unprecedented rates, and we saw this gender gap around the country, in special elections in New York’s 19th District, in Alaska. It was reflected in the Democratic fundraising advantage and then in the early vote when Democrats were coming out at very high rates. It was very much reflected in these results that there was, in the end, an anti-extremist, anti-MAGA coalition that delivered what I think for most Americans was a surprise result. But in the end, if you’d been paying attention to the data we’ve been tracking over the last three months, it wasn’t especially surprising.
Why do you think the red wave didn’t materialize?
I think Dobbs effectively coalesced this anti-extremist majority around issues that weren’t quite breaking through into the public consciousness. That had two effects: It energized a lot of voters, generally Democratic voters who otherwise wouldn’t have participated in a midterm election; generally, that’s younger voters, especially younger women, but also voters of color. And it created a larger universe of swing voters you generally wouldn’t see in a midterm election. To the extent that you have swing voters in the midterms, they’re generally swinging away from the party in power, especially when you look at the various fundamental indicators that tend to be predictive of midterm performance — presidential approval ratings, inflation, things like that. Generally, those factors coalesce in such a way that the party out of power wins significant majorities among independent voters. This narrative of Republican extremism that was really crystallized by the Dobbs decision resulted in these independent and swing voters voting overwhelmingly for Democratic candidates. In the end, I think that’s what blunted the red wave and the reason we didn’t see it materialize.
As we got closer to Election Day, there was speculation that the Dobbs effect was waning. Is there any evidence that this occurred?
There were two different worlds we were seeing at that point: There was the world of all of the hard data I mentioned — voter-registration data, early-vote data, fundraising data — that was all pointing in one direction, which was a high level of intensity and engagement from generally Democratic-leaning voters. There was the second world that was limited exclusively to polling data, and even amid the polling, there wasn’t a consensus opinion. You had one world within the broader universe of polling that was depicting a red wave, and then you had a more consistent narrative that was really more aligned with the other data that was showing not nearly as dynamic an environment.
I think what happened was, No. 1, there was a flood of the lower-quality Republican-leaning or explicitly Republican surveys that was really drowning out the much more limited and sporadic universe of quality polling. There was at some level an interpretation bias, where the media in general — you saw this even among Democratic operatives — was perhaps putting too much emphasis on and giving too much credibility to these Republican polls. The precedent would have suggested this would be a red wave.
We all lived through the 2016 election, where the polling showed a Democratic bias. Even in 2020, the polls, which were better at that point, still showed a Democratic bias. I think when these polls showed a movement toward Republicans, there was a tendency among the media and operatives to believe them and to lend them far too much credibility. And that had the result of creating a false sense of momentum for Republicans, which I believe helped them. In the end, it didn’t help them enough to hold off this Democratic rally and this Democratic result, but it might have helped them hold on in some of these close races.
You mentioned 2016. How was the quality of the polling we saw this cycle compared with 2016?
When you look at the Senate battleground, for example, the average error at this point looks to be about four points on the margin. A lot of people look at that and say, Well, that’s not bad, but a third of those polls had an error of five points or higher. So the challenge as we look forward is asking, What did we learn from this, and how will we change the way we interact with polls? We can’t think of any individual state as potentially being off by that average error of four points because if we assume all of the states are within that range, it generally lends a certain amount of credibility to the polls and you feel confident about that. But in a third of these states, there’s going to be error that’s significant enough, and you don’t know which third of the states will fall into that category. Then you can’t have confidence in any of the results. That’s a big problem for polling in general. I mean, again, there are lessons that can be learned and applied here, and the biggest lesson is not lending credibility to low-quality partisan polls that lack transparency. Hopefully, that’s the takeaway. But it remains to be seen if those changes will be put into place.
The consistent pattern is that turnout is low during midterm elections, particularly among certain voting groups like young people. What can we conclude about the turnout this year?
At this point, it’s impossible to answer that question with any high level of confidence. The good news is that we will be able to relatively soon. Every state will produce what we call “vote-history files” that tell us who voted in the election. But the challenge is it’ll take three or four months before all of that data is released. Some states, like Georgia, will be releasing that data within the next week or so. But we won’t have the full picture nationally at this point.
I do think it’s safe to say — given the result and what we saw in the early vote and the voter-registration data and in the special elections and the Kansas election, where we do have the vote history — that young voters were highly engaged in this election. I don’t know that we’ll see them reaching 2018 levels of turnout because it’s important to keep in mind that 2018 was really the inverse of this election. Harvard’s polling director, John Della Volpe, I have to credit for this. When people ask him what percent of younger voters turned out, he says, “Enough to win.” And that’s really the case. The Democratic coalition, the anti-extremist coalition in this election, was so incredibly diverse in terms of age, race, gender, ethnicity. That was clear in all the data we were seeing leading up to Election Day, and it’s what we expect to see in the vote-history data as we compile it.
The data is still coming in, and it’ll take time to pore through it and analyze it. What are you looking to see and learn from this election?
I’m most interested in getting a better sense of exactly what this coalition looked like and how it was comprised. Like I said, I believe it was incredibly diverse. I think we’ll see a lot of variation by state: Michigan and Governor Whitmer winning by wide margins, Democrats sweeping the competitive congressional races and picking up the statehouse; Pennsylvania, another state where the results are so clearly a blue wave for Democrats. And then there are places like New York, where the results weren’t quite as positive. That suggests there were some uneven effects. So we’ll be looking at the individual turnout data to better understand what happened and then, perhaps more important, how and why.
The big question here is to what extent the issue of choice impacted voters, perhaps to varying degrees. We saw this in the voter-registration data, where it looked like voters were being more motivated by the Dobbs decision in red states and in states where choice was on the ballot, like Michigan, Kansas, or Kentucky. And then perhaps less so in some blue states, which would be consistent with what we saw in New York. As we think about the 2024 election, we think about the likely Republican primary and the fact that choice will again be in the forefront of the national conversation. Do we see choice emerging, even in blue states, as a motivating factor that matches the extent to which it clearly has been motivating voters in these purple and red states?