Fast, rigorous or a bit of both?
This was the question many political pollsters were asked to weigh on Sunday as they vaulted into action, measuring the impact of President Biden’s withdrawal from the presidential race and the prospects of Vice President Kamala Harris, his likely replacement as the Democratic nominee.
Since Sunday, pollsters have met the moment in a variety of ways. Some have focused on speed; others have focused on thoroughness. Each approach has its advantages and potential pitfalls, grounded in the inherent tension between getting timely results and measuring public opinion in a broader and deeper fashion. (The New York Times began fielding its own multiday poll, with the Siena College Research Institute, on Monday.)
The first polling data taking into account Mr. Biden’s announcement began trickling in early in the week. On Tuesday, Ipsos/Reuters published a poll conducted on Monday and Tuesday indicating that Ms. Harris led former President Donald J. Trump slightly, at 44 percent to 42 percent among registered voters in a head-to-head contest.
Polls like this one can offer insight into the most urgent questions voters might have about Mr. Biden’s momentous exit. “We do think it’s important in moments like this, where the public is potentially shifting, that it’s not just speculation about what’s happening, that there is actually some data, some information, even if it may change” said Chris Jackson, vice president of Ipsos Public Affairs.
But polls with short fielding periods — the amounts of time spent collecting respondent data — also come with caveats, he said. “None of them is a perfect picture of reality because reality is just too big, too complex.” A shorter period in the field can exaggerate a problem common to all polls: an inadvertent bias toward voters who are highly engaged and more likely to participate in polling. This could decrease accuracy in a race that could be decided by disengaged voters.
“The quick polls tend to overrepresent the strong partisans, who are more stable, and, if anything, underrepresent the more fluid, independent-type voters,” said Courtney Kennedy, vice president of Methods and Innovation at the Pew Research Center, whose polls typically have a weeklong field period.
On Tuesday, the Marist Institute for Public Opinion, with NPR and PBS News, also released a poll fielded the day after Mr. Biden’s announcement. The results suggest that Mr. Trump leads Ms. Harris by 46 to 45 percent among registered voters with Mr. Biden out of the race. Marist conducted the poll using exclusively online panels, rather than a combination of online panels and phone calls as it usually does, said Lee Miringoff, the director of the institute. “It’s an effort to get a quick snapshot, a reaction, to this major event, and then obviously go back in the field with our hybrid methodologies,” he said.
Some pollsters, like Quinnipiac University, were already in the field when the news hit. Doug Schwartz, the director of Quinnipiac’s poll, said the organization had anticipated that Mr. Biden might drop out over the weekend and tailored questions in its poll that began Friday to be answerable under the alternate script. “It was a challenge,” he said. “Could we come up with a wording that would work under both situations?”
“Do you think that Joe Biden dropping out of the presidential race is the right thing or the wrong thing to do?” was one of the formulations the Quinnipiac team settled on. And while Quinnipiac had originally planned to be in the field through Monday, he added, given that the poll had amassed a satisfactory sample size, Quinnipiac chose to end it on Sunday, before the news of Mr. Biden’s exit was fully metabolized. “I thought it prudent before there are more changes in the political environment,” Dr. Schwartz said. “This is the moment before there’s any other changes. Let’s see where the race stands.”
The Quinnipiac poll found that Mr. Trump led Mr. Biden 48 to 45 percent among registered voters, whereas Mr. Trump led Ms. Harris 49 to 47 percent. (Dr. Schwartz said the pool of people sampled by Quinnipiac after Mr. Biden’s announcement was too small to evaluate the announcement’s effects.)
Dr. Schwartz acknowledged that voters assessing a Harris-Trump matchup before Mr. Biden’s announcement could well feel differently about the race after it. In responding to any hypothetical poll question, “people aren’t just telling you how they feel,” said Mr. Jackson, of Ipsos. “They’re telling you about how they might feel if something happens, predicting their own behavior,” he added. “People aren’t always the best at doing that.” This also explains some pollsters’ eagerness to begin polling right away after Mr. Biden’s exit announcement, and his endorsement of Ms. Harris, made her candidacy an imminent possibility.
Now that Ms. Harris’s nomination has effectively been locked up, things could change.
“We’ll be beginning to get an idea of how does she perform on her own, in her own right, as the candidate,” said Charles Franklin, the director of the Marquette Law School Poll.
Dr. Franklin said he was relieved not to have started polling on Friday, as originally planned.
Instead, Marquette will be releasing one poll of Wisconsin voters, and one of voters nationwide, the first week of August. Both are currently in the field, meaning that by then they will have been conducted over a period of seven to 10 days. That’s longer than many other polls.
Giving people a quick look at the state of the race when they want it most, as the Reuters/Ipsos poll does, is valuable, said Dr. Franklin.
But one or two-day polls have their own trade-offs. “They don’t have time to work the sample to try to get a higher completion rate, to go back to people who don’t immediately respond, to give them another chance or two to respond,” he said. “What we do, with our longer field period, is deliberately sacrifice the immediacy of getting the results for more time to work the sample.”
Neil Newhouse, a pollster with the Republican firm Public Opinion Strategies, said he had advised his client candidates to pause their private polling until after the dust from Mr. Biden’s announcement had settled. “I think this early polling right now could reflect a kind of a sugar high, or maybe a honeymoon, for a new candidate coming into a situation like this,” he said, since voters’ views of Ms. Harris may evolve as they become more acquainted with her.
Corrie Hunt, a pollster for the Democratic firm Hart Research, said that the Republican National Convention, which wrapped up last week, could conceivably cause a simultaneous surge in support for Republican candidates in the polls. And while a convention bump might be expected, it will be hard to disentangle it from voters’ reactions to the assassination attempt against Mr. Trump. “It’s honestly really hard to say when there are multiple things happening at once,” Dr. Hunt said.
Dr. Hunt also said that an event as significant as Mr. Biden’s departure from the race was new in her polling career. “I’ve never polled in something like this, where the entire dynamic of the race and the candidates and the way people are thinking about the candidate is changing.”
The post Biden Dropped Out 3 Days Ago. Ready or Not, Here Come the Polls. appeared first on New York Times.
