Before any of the debates. Before SuperStorm Sandy. Before the last-minute barrage of campaign advertising. He predicted the result perfectly over a month ago.
Over the past several months, political scientists have consistently forecasted an Obama victory–a relatively narrow one, but a victory nonetheless. And, in the end, that’s exactly what happened yesterday. Every state but Florida has finished counting its votes. Assuming that Florida goes Democratic, then the following forecasters will all have correctly predicted EVERY state’s presidential vote:
- Simon Jackman, a political scientist at Stanford University, writing for Pollster.com (now owned by HuffPost)
- Nate Silver at 538 (now owned by NY Times)
- Sam Wang of Princeton University, writing for the Princeton Election Consortium (almost–he put Florida for Romney)
- Drew Linzer, a political scientist at Emery University, writing for Votamatic (his blog)
- And, less publicized, Jay DeSart at Utah Valley University
Since this is a Utah-based blog, let’s look at Jay DeSart’s model for a minute. Over a month ago, Jay DeSart and his collaborator Thomas Holbrook posted this forecast:
Compare that forecast to the actual election result map. I took this one from the New York Times this morning:
Look carefully. You will not find a single difference. Working from here in Utah, Jay DeSart nailed it. Before any of the debates. Before SuperStorm Sandy. Before the last-minute barrage of campaign advertising. He predicted the result perfectly over a month ago.
Forecasts made far in advance of the election mostly relied on economic indicators, also called “the fundamentals” (which predicted a narrow victory for Obama). Forecasts made closer to election day mostly relied on polling data (which produced similar results). Either way, social science works. Consider what John Sides wrote today on every quantitative political scientist’s favorite blog, The Monkey Cage (the links are all his):
Barack Obama’s victory tonight is also a victory for the Moneyball approach to politics. It shows us that we can use systematic data–economic data, polling data–to separate momentum from no-mentum, to dispense with the gaseous emanations of pundits’ “guts” and ultimately to forecast the winner. The means and methods of political science, social science, and statistics, including polls, are not perfect, and Nate Silver is not our “algorithmic overlord” (a point I don’t think he would disagree with).
But 2012 has showed how useful and necessary these tools are for understanding how politics and elections work. Let’s hope that these tools become even more prominent in political journalism and punditry.