Jul.18

electoral-vote.com

Dem 51

   

GOP 49

Today's Presidential Polls

Another day of not very good polls for Joe Biden. When one or two polls say he's down in, say, Georgia, those could be outliers. But when a whole bunch of polls say the same thing, it's much more likely to be correct.

At the moment, supporters of Biden—including people in the White House—are pointing to the FiveThirtyEight prediction model, which still has Biden as a slight favorite to win the White House. We've had a bunch of questions about the model, and whether it should be taken seriously. We were going to address the matter this weekend, but how about we just do it now?

Besides the FiveThirtyEight model, the other predictive tool that still favors Biden is Allan Lichtman's "Keys to the White House." What these two models have in common is that they very heavily weight fundamentals—things like unemployment, inflation, the stock market, etc. Lichtman's model does not incorporate polling at all, while polls are a relatively small part of FiveThirtyEight's algorithm, at least at this point in the cycle. The various models that are primarily poll-driven all have Trump as a small-to-moderate favorite.

So, which approach is the correct one? We wish we could tell you. In most elections, focusing on fundamentals is an appropriate choice through the end of August or so, because so many voters aren't paying attention and don't really know the candidates. However, Biden and Trump are two of the best-known candidates in American political history. So, a fundamentals-first approach may not be called for in this particular cycle.

Another thing that is not a great sign for Biden: People like Nancy Pelosi and Chuck Schumer have been interpreting polls for decades, and have access to fine-grained information that the general public does not. And, as we note above, Pelosi and Schumer are scared witless.

And to pile on a bit more, if you go back to 1936, there were 15 races featuring incumbent presidents. Only two incumbents who were behind in mid-July went on to win, George W. Bush in 2004 and Harry Truman in 1948. All the others who were trailing, lost.

All of this said, most polls still have 10%-20% of the electorate who are not backing a major-party candidate. As we have written many times, the final percentage is almost never that high, and usually checks in at something less than 5%. If we presume that 8%-12% of those voters will eventually cast ballots for one of the two major-party candidates, then Joe Biden would be in a position to win if he can take a majority of those votes. So, could he? Yes, he could. In particular, in an era where ticket-splitting is rare, it's hard to believe that huge numbers of voters would vote for a Democratic U.S. Senate candidate but not for the Democratic presidential candidate. Put another way, when you see that, say, Sen. Tammy Baldwin (D-WI) is at 50% in Wisconsin and Biden is at 43%, it's entirely plausible that 7% of voters are strong for Baldwin, and are hoping for a different presidential candidate, but will fall in line behind Biden if they end up with no other choice.

We're just trying to give readers all the various things to consider. Which of these things is most instructive, however, we just don't know. (Z)