Opinion | The Healthy and Unhealthy Ways to Digest Polls in the Final Two Weeks

“Keep calm and throw it in the average” is the best way for a voracious consumer of polling data to stay sane this election season. The ups and downs that can come from seeing your preferred candidate pingpong back and forth, from day to day, become less stressful when placed into context — even when this presidential election is remarkable for just how little bouncing there has been in the polls on average.

To put this year’s post-Labor Day race into some historical polling context, I looked at how much one frequently cited polling average, that of RealClearPolitics, has moved going back through the 2008 election. (I’m relying on the RealClearPolitics average for this exercise, despite the excellent polling average published by The Times, because it’s the only one I found that goes back for that many presidential cycles. It’s worth noting that, as I write this, both averages show the head-to-head race as Harris 49 percent, Trump 48 percent.)

In 2008, shortly after Labor Day, John McCain actually held close to a three-point lead in national polls, but by Election Day, things had swung heavily toward Barack Obama, with the RealClearPolitics average showing Obama leading by around seven points — a 10-point range of results over the course of about two months. In both 2012 and 2016 you can see the average shift by a few points during the final two months. In 2012, it goes from a substantial lead for the Democratic candidate, Mr. Obama, to a near tie with his opponent, Mitt Romney. In 2020, a race in which Joe Biden consistently led in the national polls, the polling average still moved in a range from around Mr. Biden +6 to Mr. Biden +10.

This year? Since Labor Day, the polls have moved from a high-water mark for Kamala Harris of +2.2 to a recent slim lead of just +0.9.

So when we say the race is barely budging, it is barely budging.

Why might this be the case? As I suggested two weeks ago, Trump’s very well-defined brand image anchors the election, leaving a very narrow band in which the race can trade. I also think there’s the possibility that something methodological is behind the stability we see in the averages. Because some pollsters are weighting their surveys to a fairly fixed assumption about what the electorate will look like, including with regard to partisan makeup, it prevents numbers from moving as much as they might otherwise if the partisan makeup of their samples could float freely.

One thing that doesn’t explain this? Herding. Herding is a phenomenon wherein pollsters conceal results that look like outliers, possibly to avoid public backlash if their numbers prove incorrect. For instance, one analysis found that in the final week of Britain’s 2015 general election, “the decrease in the variance” on the estimate of the Conservative Party’s lead was “consistent with herding.”

But more important here, I’m not suggesting that the 2024 polls, individually, contain no variation. Just since the start of October, respected polls have produced results ranging from around Trump +2 to Harris +5. We could still see the polls herd in the final two weeks, but the reason the averages aren’t moving isn’t because all the polls are saying the same thing.

People want to know what is going to happen in this election. Assuming the polls on average are accurate, the race will be extremely close. And assuming that some won’t quite hit the mark is understandable, but doesn’t provide much solace, considering that whether the polls “miss” similarly to 2022 or more like 2020 yields a wildly different result in the contest. The race is close, it isn’t moving much and we might have to learn to live with that uncertainty for the next two weeks. Keep calm and throw it in the average, indeed.

Kristen Soltis Anderson, a contributing Opinion writer and Republican pollster, is the author of “The Selfie Vote: Where Millennials Are Leading America (and How Republicans Can Keep Up).”

The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips. And here’s our email: letters@nytimes.com.

Follow the New York Times Opinion section on Facebook, Instagram, TikTok, WhatsApp, X and Threads.