One take-home message from the wake of this year's elections is fairly simple: a data-driven analysis of voter attitudes and intentions is far preferable to that of relying on gut feelings or wishful thinking. I know that Nate Silver, whose 538 blog has been a must-read for me for several years (it is one of the few I have considered to be essentially reliable, and whose author makes sure his posts are well-researched prior to publication), had taken a lot of heat this particular cycle, and guess what: once it was all over but the crying (or celebrating depending on one's perspective) it was the poll aggregators such as Nate Silver who for all intents and purposes got it right. Those who know me personally know that I was not the least bit surprised that Obama won his re-election bid, or that the Democratic Party maintained its majority in the US Senate. That's what the data were suggesting for several months. I could also tell folks from my current home state that it was going to be a nail-biter in the state legislative races, but it would end up at least mildly disappointing for anyone who considered himself or herself a partisan Democrat. Sure enough, the elections broke about as the polls would have suggested. I'm an advocate of following the evidence, the data that have been collected via well-designed research. Although the numbers may not exactly "speak for themselves," they at least point us to the most valid range of possible interpretations.
Here's
Brad DeLong:
The best pollsters, Silver said, “let the sample tell you what it is
by themselves.” The ones who didn’t “put their finger on the scale” and
didn’t make assumptions about the voting public were most successful.
Just going with what the data said instead of making assumptions
is usually the best practice whenever you're doing any kind of
scientific survey and that worked again this year…. I’m a pro-horserace
guy. I’m more interested in diagnosing 2016 now than Benghazi, for
example, because that’s where my bread is buttered. But if you’re going
to do horserace, then do it the right way because it can be more data
driven.
