In the run-up to the 2020 presidential election, the polls showed something strange: Joe Biden was faring far better than expected among voters over age 65. Some polls showed him ahead by 10 points or more.
It was a little hard to explain — and believe. Yes, the pandemic hit seniors hardest. Yes, Mr. Biden was old himself. Yes, the baby boomer generation was aging into the 65-and-older group, replacing somewhat more conservative voters. But could Mr. Biden really be winning older voters? When the final overall results came in far better for Donald J. Trump than the polls suggested, it appeared to offer an obvious answer: no.
Three years later, I’m wondering whether there was more to Mr. Biden’s strength among older voters than it seemed. Maybe he didn’t win older voters by 10 points, but maybe he actually did come close to winning older voters or outright did so.
Our own Times/Siena polls, for instance, were highly accurate. They did not overestimate Democrats. And yet the Times/Siena polls found the generic congressional ballot tied among seniors, at 45 percent support for each party. In a question asking how they voted in the 2020 presidential election, the polls still found Mr. Biden leading Mr. Trump, 53 percent to 47 percent, among older voters.
Could Mr. Biden really have done so well? Unfortunately, it’s very hard to be sure. The various post-election studies — like the exit polls or the data from the Democratic firm Catalist — still show Republicans winning the group in 2022. Worse, the hard election results don’t offer much additional evidence to help clarify the matter. Voters aren’t nearly as segregated by age as they are by race or education, making it difficult to find additional evidence in voting results to confirm whether the trends evident in the polls are ultimately borne out on Election Day.
But there is one additional data point worth considering: our high-incentive mail study of Wisconsin. As you may recall, we promised Wisconsin voters up to $25 dollars in an effort to reach the kinds of people who don’t usually take political surveys. In the end, it achieved a response rate surpassing 20 percent (by contrast, only about 1 percent of our attempted phone calls yield a completed interview in a typical poll). The response rate among older Wisconsinites appeared to be much, much higher.
Democrats fared better among older voters in the Wisconsin mail survey than in any other major election study. The mail survey found the Democrat Mandela Barnes beating the Republican incumbent senator, Ron Johnson, by 52-40 among older registered voters. In comparison, the concurrent Times/Siena poll — using our traditional live-interview methods — found Mr. Barnes up by 46-43 among that group, while the other election studies were even farther to the right. The exit polls found Mr. Johnson ahead by seven points with that group while AP/VoteCast found Mr. Johnson up by four points.
The findings were just as extreme when voters were asked to recall how they voted in the 2020 presidential election. In the high-incentive mail survey, voters over 67 in 2022 (meaning over 65 in 2020) said they backed Mr. Biden by 55-38 over Mr. Trump. In contrast, the Times/Siena poll found Mr. Biden ahead, 48-43, among the same group. The exit polls and VoteCast data both found Mr. Trump winning seniors by a comfortable margin in 2020.
To reiterate: There’s not much additional evidence to help corroborate these very different versions of what happened among older voters. But the mail survey in Wisconsin is intriguing evidence. It’s renewed my curiosity in the possibility that maybe, just maybe, Democrats are doing better among older voters than is commonly thought.
If they are, it would help make sense of the party’s new strength in special elections — which tend to have very old electorates — and perhaps in last November’s midterm elections as well.