As I write this early on what used to be called, back in the old America, election day (now it’s simply Nov. 6 of election month or two), it seems to me that the large horde of election pundits and prognosticators has divided itself into two distinct schools: those who think the polls are accurate and those who don’t.
Modern polls, their defenders say, are scientific, objective. They believe that the pro-Democratic sampling bias of most of them is the product of pure science, not stupidity or, well, bias. Typical of this view is the Washington Post‘s young avatar of the conventional wisdom, Ezra Klein:
The polls, taken together, are typically pretty accurate. Systemic problems, while possible, aren’t likely. There are a lot of pollsters producing a lot of polls and each and every one of them has every incentive to try and get it right. When they converge, it’s typically with good reason. And right now, they have converged. The 3-4 percentage point error necessary for Romney to be the real favorite in this race is extremely unlikely.
On the other side are the old pros — Peggy Noonan, Ed Rollins, Karl Rove, Michael Barone, Hugh Hewitt — who believe the polls have failed to capture the enormous wave of enthusiasm, not only anti-Obama but also in recent weeks pro-Romney, that gives them what amounts to a gut feeling that Romney will win. Here’s how that point was made recently by a young pro, The Weekly Standard‘s Jay Cost:
When I started making election predictions eight years ago, I had a very different perspective than I do today. I knew relatively little about the history of presidential elections or the geography of American politics. I had a good background in political science and statistics. So, unsurprisingly in retrospect, I focused on drawing confidence intervals from poll averages.
Since then, I have learned substantially more history, soured somewhat on political science as an academic discipline, and have become much more skeptical of public opinion polls. Both political science and the political polls too often imply a scientific precision that I no longer think actually exists in American politics. I have slowly learned that politics is a lot more art than science than I once believed.
We shall see pretty soon whether the artists or the scientists got this one right.
UPDATE II [8 November]
From the Chronicle of Higher Education, the quants (the poll quantifiers and number crunchers) decisively defeated the gut-consulting pundits:
So the quants and their statistical models were right, while the pundits and their guts were wrong. This was also true in 2008: Silver and Wang were both nearly perfect. But in 2012 the victory of mathematics over bloviation was even more resounding somehow, perhaps because the battle lines were more clearly drawn.
Before the election, Peggy Noonan of The Wall Street Journal asserted that “nobody knows anything” about the potential outcome. David Brooks of The New York Timessaid forecasters lived in “sillyland.” Jennifer Rubin of The Washington Post tweeted that “averaging polls is junk.”
Can anyone seriously argue that now?