Reading the Tea Leaves on Bernie

by Hamilton E. Davis

   The whole State of Vermont has gone bonkers over the presidential candidacy of Sen. Bernie Sanders, as has a sizeable chunk of the country. Bernie’s win in the Michigan primary took Vermonters into hyperventilation territory and they are likely to stay there until next Tuesday when Illinois, Florida, North Carolina and Ohio hold their primaries. So, what is going on? Can the grumpy old guy with the bad hair from the north end of Burlington actually get into the finals?
   The hard-nosed guys who never hyperventilate don’t think so, and I think they are right. There are two essential sources, in my view, that can get you to see past the emotional responses to political questions like this. The first is Nate Silver, who runs a web site called 538 and who has made himself the nation’s leading prognosticator by virtue of his skill as a statistician. The second is Nate Cohn, who writes under the Upshot heading in The New York Times. Read these guys and you still might get something wrong—they both got Michigan wrong—but you will be as grounded as it's possible to get.
   No one knows whether Michigan was a fluke, or not. I think Bernie himself was shocked. During the run-up to the Michigan primary itself, Tad Devine, Bernie’s campaign manager and the architect of his stunning run so far, gave an interview to the Politico site in which he suggested that Hillary Clinton might choose Bernie as her running mate in the general election.
   Maybe they’re going to put him on the ticket then, Devine said. He isn’t kidding as far as I can tell, the interviewer wrote.
   That looks to me like a message that you might send if you thought your campaign had peaked. Keep in mind that the guy saying it wasn’t some indiscreet volunteer playing hooky from college; it was the top campaign guy.
   In any event, the fact is that Bernie did win Michigan and if it didn’t shock Bernie, it shocked everybody else, on both sides of the equation. It certainly knocked Nate Silver sideways: he had estimated Clinton’s chance to win Michigan at greater than 99 percent and Bernie’s chances to win at less than one percent. The question now is whether Michigan portends a Bernie surge that can actually get him the nomination.
   Here is what Cohn has to say in the Upshot:
   Imagine…a brutal stretch for Mrs. Clinton, one where she underperforms the demographic projections by as much as she did in Michigan for the rest of the year. She would lose in Ohio and Missouri on Tuesday. States where Mrs. Clinton was thought to have an advantage, like Arizona, New Mexico, Pennsylvania, Illinois, Indiana, California and Connecticut, would become tossups. Mrs. Clinton would win New York, but by just eight percentage points.
   She would be swept in the West, including 40-point losses in places like Alaska, Idaho, North Dakota, Utah and Montana and 30 point losses in Washington and Oregon. She would lose by 20 points in Wisconsin and Rhode Island, by 30 in West Virginia and Kentucky.
   She would still win comfortably.
    The reason, Cohn writes, is that she has too big a lead to be overtaken in any realistic scenario. Keep in mind that these aren’t simply finger-in-the-wind guesses. They also have nothing to do with super delegates. They are what you get when you apply the Michigan results to all those states. So, is a Bernie nomination it possible? Sure, but it is very remote. My guess is that the only Vermonter who fully understands that is Bernie himself.
   As I noted above, Nate Silver has a firmer grip on the hard statistics than anyone else. He made his bones in national elections beginning in 2008 and he had a striking performance in the 2012 election, when President Obama defeated Mitt Romney. Silver saw it coming all the way, and his work badly embarrassed some of the old political guard. He made Peggy Noonan, the intelligent person’s Republican who writes in the Wall Street Journal, look like a fatuous ninny when she predicted a Romney landslide after talking to a cab driver, or some damn thing. Another victim was Karl Rove, the very smart but nasty brains behind former President Bush; Rove got run over on national television when he refused to believe the results in Ohio after they had come in.
   Silver doesn’t seem to have fully come to grip on Michigan. He described it right after the votes came in as the worst screw-up in polling history, since his political readings are based primarily on polls by others, to which he applies various correction factors. It certainly was the greatest screw-up in Nate’s history. Saying that Clinton has a higher than 99 percent chance of winning and Bernie less than one percent really isn’t statistics at all—it just says, forget about it. If he had said that Clinton had a 98 percent chance to win Michigan and Bernie had a two percent chance, he would have been fine. Two percent is not no percent in statistics land.
   It is interesting, however, that Silver hasn’t backed off an inch in the approach to next Tuesday’s tests. He gives Bernie no chance (less than one percent) in Illinois, Florida and North Carolina and just two percent in Ohio.
   So, Tuesday will be big. If Silver is right, the Clinton lead in delegates will grow significantly and the path for Bernie will be pretty much closed off. If he exceeds expectations, even by a lot, then the path will look like it’s there, but Bernie, and his Vermont chorus will then have to face the Nate Cohn thesis—that the path is already gone.

N.B. While I agree with Silver and Cohn, I also think that Bernie’s campaign has been unique in American political history, and has had a marvelously salutary effect on the country’s political life. I’ll make that case next.