Trump’s job approval rating won’t save him 📊 October 11, 2020

A popular argument about the polls being biased has one critical flaw

Thanks as ever for reading my weekly data-driven newsletter on politics, polling and the news. As always, I invite you to drop me a line (or just respond to this email). Please click/tap the ❤️ under the headline if you like what you’re reading; it’s our little trick to sway Substack’s curation algorithm in our favor. If you want more content, I publish subscriber-only posts 1-2x a week.


Dear reader,

It is hard to know where to begin. On the one hand, the presidential race has been quieter this week than what has seemed normal over the past month And yet, the torrent of news still seems unending. Just this week we had news cycles about court-packing and judicial nominations, Amy Coney Barrett’s religious roots, cancelled debates and a plot to kidnap the governor of Michigan.

I will not attempt to cover all of these stories. Alas, only pundits pontificate about every subject in the news. Instead, I want to hone in on something entirely different: an arcane point about a popular and misleading argument about why polls might be underestimating support for the president.


Trump’s job approval rating won’t save him

A popular argument about the polls being biased has one critical flaw

On one hand, it is hard to reckon with the idea that Joe Biden could win the election by 12 or 13 percentage points—yet is is also surprising that a president with so low of favorability ratings, and with as storied a presidency, as Donald Trump could even come anywhere close that. The tension between these two suppositions has been cause for quite a bit of handwringing on the modeling front this year. What estimates are more believable than others? What do we need to adjust for? Are there shy Trump voters—or, indeed, shy Biden voters?

Evidently I am not the only one who has faced these problems, and more. In The New Yorker this weekend, journalist Isaac Chotiner has published an interview with Sean Trende, the polling analyst at Real Clear Politics, that ponders similar questions.

I think the interview is quite insightful and I recommend that you read it, but Trende also falls into the trap of repeating a common argument about why polls might be underestimating support for Trump. Here it is in his words:

There’s also this question of Trump’s job approval, which is running two or three points ahead of his vote share. And, going back in time, Presidential job approval is one of the strongest indicators we have of how the President is going to fare. Who are these people that approve of the job he’s doing but aren’t going to vote for him? Again, you can tell yourself a story. There are people who like his policies, but hate his persona. But it could also be that these are voters who are saying they’re undecided, but they really aren’t. But, again, this is just kind of conjecture, it’s not really good evidence.

The logic goes that since Donald Trump’s approval rating is higher than his vote share, there might be some group of voters who approve of him but won’t say they’ll vote for him. Maybe, the implication is, these are the shy Trump voters we have been hearing about. Maybe they are even the reason polls were wrong in 2016!

This is an appealing argument, but it omits a few crucial points that I fear make the argument more about unsubstantiated “conjecture”—to use Trende’s own words about the angle—than a real concern.

The first is that the difference between Trump’s approval rating and his vote share is actually smaller than two or three points. Based on FiveThirtyEight’s average — which is better than Trende’s not least because it adjust for the fact that Rasmussen, a prominent approval pollster, almost always gives Trump a 10-point boost net boost — the residual is actually closer to one point. Today, 538 has Trump’s vote share at 42%, and his approval rating at 43%.

But the second point is about a more severe lack of context. It’s that there is also an identical difference between Donald Trump’s disapproval rating and Joe Biden’s vote share. Again according to 538, 53% of voters currently disapprove of the president, while Biden is winning 52% of the vote for president. The explanation? A higher rate of people answering that they “don’t know” or are undecided about their opinion on the ballot test compared to whether or not they approve of the president.

This calls the implication that Donald Trump has a potential upside with undecided voters (at least on these grounds) very much into question. If we are expecting Trump’s approval to be a guide for gaming out how undecideds will lean, then we have to conclude that there are at least as many people out there who will vote for Joe Biden than will vote for Donald Trump.

And that all sets aside that we have some actual data on how undecided voters—who would not be enough to save Trump even if he won them all—lean:

There are much better arguments for why polls might be underestimating support for Trump — the best of them ranging from from bad weighting to bias caused by differential partisan non-response. I wouldn’t put much stock in this one if I were you.

Posts for subscribers

What I'm Reading and Working On

I read an old copy of Hemingway’s In Our Time this weekend to unwind from a very busy week.

But this upcoming week won’t provide much relief, I’m afraid. I’ll be writing on undecided voters and vote-switchers this week, as well as preparing a big piece on racial disparities in absentee ballot rejection rates.

Thanks for reading!

Thanks for reading. I’ll be back in your inbox next Sunday. In the meantime, follow me online or reach out via email if you’d like to engage. I’d love to hear from you.

If you want more content, I publish subscriber-only posts on Substack 1-3 times each week. Sign up today for $5/month (or $50/year) by clicking on the following button. Even if you don't want the extra posts, the funds go toward supporting the time spent writing this free, weekly letter. Your support makes this all possible!

Photo contest

Here is a picture that Brandon sent in a while ago. He says “My parrot, Butch, who I've had since I was 7 years old (I am now 60!) Butch is a 54 year old Amazon parrot.”

I think that’s pretty freaking cool. (PS I’m running low on a stock of pet photos for next week, so do us all a favor and send in a snap of your cute pets for next week and the weeks beyond.)

For next week’s contest, send me a photo of your pet(s) to elliott[AT]gelliottmorris[DOT]com!