"In the Christian faith," a friend likes to say, "we have Matthew, Mark, Luke and John to guide us. In politics, we have Nate Silver." True enough. After switching from baseball to politics and predicting the outcome of the 2008 election, Silver's prognostications are considered so canonical that others are relegated to the ranks of Nostradamus-like-also-rans. It takes a special kind of genius to be so wildly known while still self-identifying as a statistician. Yesterday, Silver released his first analysis of the 2016 general election.
Silver's methodology is part of the genius. He is not himself a pollster -- he builds on the work of others, taking their data and running it through a series of 10,000 simulations to test it and see how it holds up with that many modulations.
I think the bigger part of Silver's genius comes from something other than his algorithms. I think it is his modesty. Unlike many pollsters, Silver takes the time to explain not only the methodology he uses, but the limits of that methodology, that he is not, repeat not, explaining what will happen on election day but is, instead, explaining what people tell pollsters about their behavior on election day, which is a long way off and a million things could happen between now and then, and no two candidates are alike and. ... In short, Silver is that rare statistician who recognizes the limits of his craft and of his data. He knows what he knows and he knows, too, the many variables that cannot be known.
This may be especially important this year. As I noted in my links column yesterday, there may be people who are not inclined to tell a pollster they will be voting for Donald Trump because they know that he is considered a buffoon and a racist, but once the curtains close, they might be so convinced the nation's political class needs a shake up, they will vote for him yesterday. The column to which I linked yesterday was entitled, "I hate Donald Trump, but I might vote for him," and it is a sentiment that will keep both the Clinton campaign, and the pollsters, up at night until November.
We also do not have any idea how this campaign will play out in part because it is far from clear what kind of candidate Trump will be. The departure of his campaign manager, Corey Lewandowski, removes the bad devil of Trump's candidate conscience, whispering into his ear, "You got this far by not listening to the pros. You are smarter than all of them. Follow your gut." But, that does not mean Trump will learn overnight how to acquiesce to the advice of the professionals he has mocked, and not only mocked, but defeated, to date.
I also worry that stories that the political class think are important will have a different effect from that which they anticipate. Above the fold in this morning's Post front page is the headline "List of no-shows growing for GOP." Beneath the headline, there are pictures of former Gov. Mitt Romney, Sen. John McCain, and former President George W. Bush, as well as former Gov. Jeb Bush, none of whom will be attending the GOP convention in Cleveland next month. Think about that: It is unprecedented so far as I can recall for a party convention to be avoided by that same party's three previous nominees. But, I also expect that story doesn't hurt Trump with independent voters who, by definition, choose not to belong to either political party. And, I also suspect the absence of these men will not hurt Trump with core Republicans as much as people might think: They still want to defeat Hillary. And, finally, I suspect Trump will play up their absence, not ignore it, and turn it to his advantage.
Nate Silver is quite clear that he is not telling you what will happen between now an election day. What is especially useful about this year's analysis is that he differentiates between analysis that relies solely on current polling data and that which relies on polling data, economic data and historical performances. In the polls-only prognostication, Clinton wins big, with an 80 percent chance of winning the presidency, and a predicted 353.8 electoral votes, flipping both North Carolina and Arizona into the blue column. When you factor in historical analysis and economic data, those two states flip back; Clinton still wins, but the margin is smaller.
The map Silver provides is completely interactive, a quality in which I usually do not place much value, but in this case, it is enormously helpful to be able to click on the different states, and the different models in each state, and then back to the national model.
The net takeaway from Silver's first prognostication? This race is Clinton's to lose. She is leading in all the swing states. If the Latinos who are eligible to vote in Arizona register and show up on election day, a state that the GOP has won by 9 points the last two elections could flip. Obama won North Carolina and Indiana in 2008 and lost them both in 2012: Clinton could bring back North Carolina but Indiana appears to be a more solidly red state. Missouri and Ohio will be battleground states as far as the eye can see. And, perhaps most importantly, Silver's analysis shows no real evidence that Trump can flip the Rust Belt states like Michigan and Wisconsin, without which, he has not real path to the White House.
Those of us who eat, drink, and sleep politics, have been waiting for Silver's first analysis. It did not disappoint. He may not deserve a mosaic of him in St. Matthew's Cathedral, but he is as canonical as a statistician can be.
[Michael Sean Winters is a Visiting Fellow at Catholic University's Institute for Policy Research and Catholic Studies.]
