Inaccuracy in New York magazine report on election forecasting

David Freedlander reports on some election forecasters. It’s a silly article, and equating Allan Lichtman with actual political scientist isn’t so fair to Alan Abramowitz, but I get it—it’s a fun feature story, not meant to be an in-depth exploration of the topic.

I did want to correct this statement, though:

ABC hired the former Economist modeler G. Elliott Morris to replace [Nate Silver, originator of the Fivethirtyeight website]. To replace Morris, The Economist hired Andrew Gelman, a renowned statistician at Columbia University.

I appreciate the “renowned statistician” bit, but I don’t think it’s accurate to say that the Economist hired me. They have full-time employees who do their forecast as part of their jobs. They also donated $15,000 to Columbia to support the work of Ben Goodrich, Geonhee Han, and I in helping them to fix up their 2020 forecasting model for the new election (see some discussion here). And in any case I did not replace Morris; Morris was replaced by new people on the Economist’s data science team. I advised the Economist in 2020 and again in 2024. In 2020, Elliott developed the Economist model with the collaboration of Merlin Heidemanns and me. It’s pretty much the same model, although various tuning parameters have been set differently.

Freedlander also discusses differences between different forecasts earlier in the campaign, back when Biden was still running and Fivethirtyeight gave each party a 50% chance of winning, as compared to the Economist and other forecasts, where the Republicans were at 75%. He quotes some trash talking on the topic.

If you’re interested in the statistical background on that story, I recommend our posts at the time: Polling averages and political forecasts and what do you really think is gonna happen in November? My comments on Nate Silver’s comments on the Fivethirtyeight election forecast I don’t think the trash talking on twitter contributed anything useful to the discussion, then or now. There were some interesting forecasting issues, though.

Again, I can’t blame the New York magazine author here. He’s on deadline, he didn’t have time to check the facts with me, and he’s writing a feature story where the trash-talking is kind of the point—if all those forecasters were polite to each other, there’d be no story, and if you want statistics discussions, you’d go somewhere else anyway. Still, I didn’t want people out there thinking that the Economist had hired me or that I’d replaced Elliott. Hence this post.