Jon Healey of the LA Times asked:
I was wondering if you’d done any research into whether vice presidential choices make a measurable difference in a presidential candidate’s ability to win. Have you? And if so, what did you find?
A few years ago we did an analysis that estimates the VP effect to be about 2 or 3 percentage points in the VP’s home state. This can be hard to notice given the year-to-year swings in relative state votes. Also, we never did a careful study of the VP effect, it was just a regression that we ran. Still, if I had to stake a claim, I’d say that you won’t see much upside at the national level . . . but it could help in a key state.
I was happy to see that he quoted my statement of variability (“This can be hard to notice given the year-to-year swings”) and also my disclaimer (“we never did a careful study . . . it was just a regression that we ran”). I’m always worried that reporters will simplify and understate uncertainty, so this was a pleasant surprise.