I’ve been leaning on the electoral logic of late, but Harry Enten suggests that this might not be such a good idea. Enten has three main points:
- (a) The chance of an Electoral College and popular vote split is small.
- (b) Without looking at all the states, we don’t know how to interpret state polls.
- (c) We don’t have a lot of state polls.
Well, Rutherford B. Hayes (1876), Benjamin Harrison (1888), and George W. Bush (2000) each won the Electoral College while losing the popular vote. With historical data, there’s always the question of framing. One way of framing it is that we have had 3 splits during all 57 US presidential elections = 0.05. Another way of framing it is that we have had 1 split during the 4 US presidential elections during the 21st century = 0.25. Which one should we prefer? Enten has a more sophisticated approach to framing:
Research by Nicholas R. Miller, a political scientist at the University of Maryland Baltimore County, has found that — all else being equal — there’s about a 25 percent chance of a split if the national popular vote is decided by about 1 percentage point, and that the chance is cut in half when the margin is 2 percentage points. We all remember the razor-thin margin in the 2000 election, when George W. Bush won the presidency even though Al Gore won the popular vote. But we forget that the 1960 and 1968 elections were about equally close but didn’t produce a split. Jimmy Carter in 1976 and George W. Bush in 2004 won relatively clear victories in the Electoral College while winning the popular vote by a little more than 2 points. If the national popular vote margin is greater than 4 percentage points, Miller found, the chance of a split is about zero.
Okay, there’s enough here to make me want to look harder at the national polls.
So what do we see? As a rough first try, I took the unweighted average of the national Clinton vs. Trump polls done this month, and the results were 53.43% for HRC and 46.57% for Mr. Trump, with a standard deviation of 1.8%. That translates (with all the usual qualifications) into a 0.972 probability of victory for HRC. Interestingly, that almost precisely what the Monte Carlo simulation I ran told me: a 0.974 probability of victory for HRC. I realize that I’m cherry picking a bit, but the fact that two reasonable but unrelated models predict in overwhelming likelihood of a D victory seems at least a little significant.
Also, there’s about a 9% chance according to that the difference between the candidates will be 2 percentage points or less. If Professor Miller’s take on things is more-or-less right, that leaves about a 2.5% chance of a split decision (so to speak) between the Electoral College and the popular vote.
All that said, it’s still too early to take these polls too seriously. Remember Gov. Dukakis’ 17-point lead in the polls during the 1988 campaign? Good times. I realize that there’s some reason to think the polls are likely to be more reliable this year because both HRC and Mr. Trump are better known than presidential candidates in the days of yore. Maybe. Right now there are 178 days, 14 days, 26 minutes, and 17 seconds until the election. Historically, polls are not very predictive at this point. The average adjusted R squared is about 0.5 this far out, but it rises to higher than 0.9 immediately before the election.
So I’m taking it slow.