Making Sense of Polling
Times Insider explains who we are and what we do and delivers behind-the-scenes insights into how our journalism comes together.
The lead-up to the 2022 midterm elections is packed with dizzying statistics. Voters are regularly inundated with polling numbers: There are conflicting data points, varying sources of information and different lenses to interpret it all. The Tilt, a subscriber-only newsletter from The New York Times that started this month, tries to make sense of the electoral whirlwind.
The goal is to do “the best job we can of honestly appraising what we know and don’t know,” said Nate Cohn, The Times’s chief political analyst, who writes the newsletter. Using polling data collected by The Times and other outlets, as well as surveys and electoral trends, The Tilt examines the historical context behind the numbers.
The Tilt is an evolution of Mr. Cohn’s “polling diary,” a 2020 analysis of daily election polls. In an interview, he discussed what readers can expect from the newsletter. This conversation has been edited.
What is the difference between The Tilt and the “polling diary”?
I thought that the 2020 polling diary was a success, but it really tried to cover every base possible: It tried to tell readers about almost every poll that came out on a given day and how to make sense of all of it.
In a presidential election, I do think there is a demand from a certain segment of readers to have every last data point interpreted. I don’t think that’s necessarily true in a midterm election when there’s a little less interest, and I think it’s a little harder to justify, given how poorly the polls have done over the last few cycles. We need to step back away from the data as often as we need to be in it.
The State of the 2022 Midterm Elections
With the primaries over, both parties are shifting their focus to the general election on Nov. 8.
- Sensing a Shift: As November approaches, there are a few signs that the political winds may have begun to blow in a different direction — one that might help Republicans over the final stretch.
- Focusing on Crime: Across the country, Republicans are attacking Democrats as soft on crime to rally midterm voters. Pennsylvania’s Senate contest offers an especially pointed example of this strategy.
- Arizona Senate Race: Blake Masters, a Republican, appears to be struggling to win over independent voters, who make up about a third of the state’s electorate.
- Pennsylvania Governor’s Race: Doug Mastriano, the Trump-backed G.O.P. nominee, is being heavily outspent and trails badly in polling. National Republicans are showing little desire to help him.
So The Tilt moves beyond the numbers?
We’re going to talk a lot about our own work here at The Times, whether it’s in terms of our own polling or election night modeling. We’re also going to have the opportunity to talk about historical precedence and the other factors that underlie the way that we analyze elections.
Take this midterm election, for instance: One of the biggest reasons people believe Republicans will regain control of the house is because there’s a long history of the party that’s out of power doing well in midterm elections. Diving into that history, and whether what we’re seeing this year lines up with the past or something different, is the thing we can do that’s different from just adding and subtracting the latest polls.
How will The Tilt pull back the curtain on the process?
I sometimes think of cooking shows as an analogy. If you step into a kitchen, or you watch a cooking show, you get to see how it all gets made. You come away with a different level of respect for the chef, even if it’s not the food you would have made yourself. We can be a little bit more of a cooking show for polling.
There are plenty of reasons to look at the recent track record of polling and be a little skeptical of just how accurate it can be. But to the extent that readers can at least understand all the work that goes into it and why it is what it is, and where it can go wrong, they will be better prepared tomake sense of it.
Why should voters be invested in polling?
Polling is sort of the lifeblood of our democracy. You may not like it that way, but in a democratic system, the responsibility of our elected officials and the political candidatesis to represent the public, and the role ofcitizens is to engage in the democratic process. Polling is one of the major ways that all of the political actors I just mentioned make sense of public opinion, and then choose how they want to engage with that system.
When the polling is bad, that’s really bad for the system in important ways. If the polls aren’t accurately representing the public’s attitudes, then elected officials don’t make decisions that reflect the will of the public.
We don’t have very many other ways to measure the attitudes of the population without this. Otherwise, we would assume that our own prejudices about what ideas sound good or don’t sound good are probably held by other people. Or we might assume that the rest of the country looks like the places where we live. We’ve learned over the last decade that that’s almost certainly not true, even if we had deceived ourselves into that view beforehand.