Democrats and liberal political junkies were rattled by a New York Times/Siena College poll that was written up in the newspaper as “Republicans Gain Edge as Voters Worry About Economy, Times/Siena Poll Finds.” The poll’s top line found that 49 percent of likely voters planned to vote for a generic Republican candidate in the 2022 midterms. A slightly smaller portion of those surveyed, 45 percent, said they planned to vote for a generic Democrat instead.
That was bad enough from the Democratic point of view. But even worse was the finding of a 32-point swing among independent women from Democrats to Republicans in one month—from +14 for Democrats to +18 for Republicans. If that’s true, that’s pretty devastating.
So—is it? Interviews I did with over a half-dozen pollsters in the hours after the Times published its findings yielded a reinforced sense of caution. These pollsters, both Democratic and Republican, didn’t refute the poll’s findings. But they warned against treating it as the be-all-and-end-all oracle on what will happen in next month’s vote.
The consensus argument was that any national poll doesn’t factor in various dynamics of individual races. Beyond that, some of the most significant public policy events of the last year have added an unpredictable element to the 2022 midterms that polling has a hard time picking up. Nobody expected an anti-abortion referendum in Kansas to be rejected as decisively as it was. And now Congresswoman Mary Peltola’s victory in the special election for Alaska’s at-large congressional district demonstrated that even in a red-leaning state, in a year where Republicans should be wiping the floor with Democrats on every level of electoral politics, nothing is certain.
“A national poll looks at national mood,” Democratic pollster Ruth Bernstein told me. “It doesn’t necessarily translate into what’s going to happen in individual congressional districts or individual states. So yeah, overall mood: Gas prices are up, and people are maybe a little more focused on the economy right now, but there are individual dynamics in individual races.”
Democratic pollster Anna Greenberg was skeptical about the demographic breakdowns. “I just think we always need to be careful to overinterpret one single survey. I think that when you argue about polling, you’re usually losing, so I say that with that caveat,” she said. “But I am skeptical that Democrats and Republicans are breaking even among women. I’m looking at it right now. Forty-seven [percent] Dem, 47 Republican. Have we ever since the ’90s had a situation where women didn’t vote more Democratic than men did? Similarly Republicans are probably doing better with men. So there’s that. And then you’ve got that Democrats are going to get more than 78 percent of Black voters. Republicans are not going to get 18 percent of Black voters … so I don’t entirely trust this poll.”
A Democrat would be forgiven for freaking out about the finding that there has been a major swing toward Republicans among independent women just a few months after the Dobbs decision. “You’re looking at 95 interviews with independent women. So you have a pretty big margin of error on this poll, plus comparing to independent women in other polling,” Bernstein said. “Is it really [a] trend or is it a margin of error? And 95 women independents from across the country doesn’t really tell us what’s happening in individual races.”
Greenberg was even more direct: “I’m polling 25 polls a week right now in congressional and Senate races, and I haven’t seen a 30-point shift or whatever they say among independent women.”
Democratic pollster Peter Brodnitz said that a better indicator is looking at gender and education. “If they find that college-educated white women are not heavily Democratic right now, then their data is just not representative of what’s happening in campaigns that I know of,” Brodnitz said. “And that’s the thing about their female independent switch. I don’t know what that is exactly. It might be that there’s something that’s going on with white women.”
The Times poll isn’t the only public survey with foreboding numbers for Democrats and only slightly better predictions for Republicans. But, with only a few weeks out, the overall consensus of the pollsters I talked with for this story, and veteran strategists in general, is that the 2022 election cycle will not be a wave election cycle. If Republicans retake control of Congress, it will likely be by slim margins. If Democrats are able to keep their slim majority in the Senate, it will still be by a whisker.
“Republicans should be running away with this election. With [President] Biden’s job approval in the low 40s, with inflation at a 40-year high and not moderating, with crime a serious problem in many cities, and with the border still not under control, Republicans should be cleaning up,” Republican pollster Whit Ayres said. “Republicans still are overwhelming favorites to take the House and have at least a 50-50 chance to take the Senate, but it’s closer than it should be because of the Supreme Court’s decision on abortion energizing Democratic women and because Republicans have nominated a number of inexperienced first-time candidates. So that’s keeping it closer than it would otherwise be.”
A national poll like this Times one isn’t a dud. These polls do show a type of mood spread across the country. But statewide polls are generally more important to pay attention to, especially at this point in a campaign cycle, because they measure respondents’ reactions to specific candidates and events, not just a general mood.
“It’s a national poll, and there is no national election,” Brodnitz said. “Only 10 percent of the House races are competitive. This is a national poll. Ninety percent of those surveyed are not in a competitive congressional race. So I think it’s a very blunt instrument for trying to understand what is going on in campaigns. But to the extent that their conclusion is people care about economic issues and that they might be looking beyond Dobbs to figure out where the candidates are on other issues, that’s consistent on everything that I’ve seen.”
Interviewing a handful of pollsters, perhaps fittingly, results in a handful of distinct responses. On this Times poll, though, there were a few consensus points: This poll is just one data point; it’s going to be a close election; anybody who says they know what’s going to happen doesn’t. But a one-month, 32-point swing based on 95 women maybe shouldn’t top the list of things Democrats ought to be worrying about.