If 45% of adults approve of Trump, how many will vote for him?

Polls don’t provide a clear answer

Adam emails:

The NBC poll has Democrats up vs Trump but Trump approval at 47-50. What is up with that? Is this a polling issue or something else? He would be underperforming his approval by 1 to 3 points. I tweeted you about this but there are a lot of comments.

What he points out is super interesting and worth talking about. In the most recent NBC News/Wall Street Journal poll, Trump’s job approval rating is 47% and his disapproval is 50% (-3). On average versus his Democratic opponents, the president is underwater by six percentage points (44 v 50 percent). That means he’s underperforming his approval ratings by 3 percentage points.

This is indeed a departure from what I’ve observed in other data. It’s a bit hard to know what’s causing the pattern or what to make of it. In YouGov’s polling, for example, Trump has been hovering around a 5 percentage point deficit in the generic presidential ballot over the past month. All the while he has been about 10 percentage points underwater on net approval, a five-point over-performance. In swing-state polls released by Quinnipiac University today, the president had a net approval rating of -5 on average across Pennsylvania, Wisconsin and Michigan while polling about even with his Democratic opponents—another five-point over-performance.

As an exercise in handicapping, this is a bit perplexing, and that we have conflicting signals about the answer is important when we’re trying to infer the electoral environment from approval polling. When Trump’s at 48% approval in Michigan, is he winning, or losing? We should acknowledge the uncertainty in the heuristic.

But this is perhaps equally as interesting as a question of political psychology. Let’s think through what it means for Trump’s net approval ratings to be higher than his vote margin. Such an outcome would indicate that there is a sizable share of the electorate that thinks he’s a fine president, but would prefer a different one. The alternative scenario (Trump winning with a vote margin higher than his approval ratings) would indicate that there are people who would prefer the president to his competitor regardless of his shortcomings.

These are interesting questions. But frankly, I don’t know what the “truth” is. I do think the narrative suggested by the YouGov and Quinnipiac polls is more believable than the one by the NBC data. And I think I need some more before I really update my prior.

What are your thoughts?

---

Editor’s note:

Thanks for reading my thoughts on this subject. And thanks for subscribing! Your membership adds up and makes all this newslettering possible (reminder: I do all this work independently). Please consider sharing online or with a friend; the more readers, the merrier! 

As always, send me your tips about what you’d like to read about next, or your feedback otherwise. You can reach me via email at elliott@thecrosstab.com or @gelliottmorris on Twitter.

—Elliott