Earlier today, I was talking to a friend who’s a political junkie. He doesn’t like Trump but he’s against impeachment. (Aside: my friend’s argument was: (1) if you impeach Trump and fail, you strengthen him in the election, and (2) if you impeach Trump and succeed, you get President Pence.)

I was telling my friend that I was fresh off watching Mitt Romney’s very moving 8-minute speech in which he stated why he would vote for the first article of impeachment.

As I talked about it, though, I remembered how much Mitt Romney has shown his disdain for Trump and I said that out loud. Then I pointed out that Mitt is not up for reelection until 4.5 years from now and a lot of things can happen in between that blunt the negative effect of his vote on his reelection chances. Then I pointed out that of all the red states to be from, Utah was probably the safest state from which to vote for impeachment because a lot of Republicans there are not crazy about Trump.

So within 5 minutes, my point completely changed from “Isn’t Mitt a principled guy?” to “Basic political science (public choice) predicts his vote. He doesn’t like Trump, his seat is safe for 4.5 years, and a lot of the Republican voters he needs don’t like Trump either.”

By the way, Romney could well be principled and that might explain his vote. My point is that basic incentives predict his vote even without principle.

My friend agreed with my reasoning and then said, “I think political science doesn’t predict the votes as well as it would have 20 or 30 years ago. I might, for example, have expected Susan Collins to vote for impeachment.”

I answered, “I don’t think so. She’s up for reelection this year, she took a lot of heat from the left for voting for Kavanaugh, and she badly needs the Trump base.”

He agreed.