The pollster J Ann Selzer wrote a very interesting article over at The Des Moines Register, Selzer Score Helps Bring Order to GOP Herd. It tries to get a handle on what will happen in the Iowa caucus. The issue is that caucuses work in a more “hands on” way and so there is jockeying for support during the caucus. The Democratic process is even more interesting in that only candidates who get at least 15% of any given caucus’ vote are allowed; so if someone votes for, say, Lincoln Chafee and less than 15% of the people there vote for him, they will have to switch to their second favorite candidate. The Republican caucus is more straightforward, but the same kind of thing will happen. And over time, this is what happens in the overall election as the weaker candidates drop out.
So Selzer put together a system where she looks at voters secondary choices. She uses three components. First, she gives double weight to people’s first choice. Second, she gives regular weight to people’s second choice. And third, she gives half weight to whether they would ever consider voting for the candidate. It’s a clever scheme. But I don’t see any reason to believe it. The results say far more about the model than about reality. Still, there is some explanatory value in the exercise.
What I think is most interesting are just the numbers of people who would consider even voting for different candidates. That to me seems like a reasonable way of looking at the long term potential of each. Obviously, these opinions can change. But it is hard to take serious a candidate that 58% say they would never vote for. I’ve taken Selzer’s data and put it together in the following chart:
Let’s start by discussing the second column: Consider. This is the percentage of people who say they would consider voting for the candidate. Donald Trump is a great example. He clearly is a non-starter as a viable candidate, but he still has 4% of the vote in terms of first choices. As Selzer noted: “enough to rank him in the top 10.” And that means it is enough to put him in the first presidential debate. Yet only 34% of those polled say they would ever vote for him. Similarly, Jeb Bush just makes it into the top ten by this accounting.
The third column provides an important caveat. It is the total “Consider” column plus the number of people who said they didn’t know. This provides some insight into people who might get some traction as they make their way through the campaign: Bobby Jindal, Carly Fiorina, and John Kasich. But that speaks really poorly of Jeb Bush. By this accounting, he comes in 12th place — only a bit ahead of George Pataki! There are also people like Ben Carson who lots of Republicans think they know because he’s gotten so much attention on Fox News. But I think in an actual election, he’s going to show himself to be the political amateur that he is, and his numbers will come down accordingly.
The Republican Party would be smart to look at something like this for their debates. It has the advantage of weeding out two of the most embarrassing candidates: Donald Trump and Chris Christie. Just the same, there is plenty of crazy getting high scores: Carson, Huckabee, Cruz. I don’t understand why Republicans think so lowly of Trump (42%) and so highly of Cruz (80%). But given that there isn’t much in terms of ideological difference between “establishment” candidate Scott Walker and “crazy” candidate Ben Carson, maybe there isn’t anything to understand. Maybe it’s just random.