Was New Hampshire a Truly Historic Result?

Justin Wolfers is unapologetic in the wake of Hillary Clinton winning the New Hampshire primary. Clinton’s victory, he says, was "one of the most surprising upsets in U.S. political history":

Election-eve trading had suggested that Sen. Obama had a 92% chance to win in New Hampshire, while Sen. Clinton rated only a 7% chance.

Against this background, it is no exaggeration to term the result truly historic… In terms of unpredictability, or at least the failure of everyone to predict it, it may have no modern match.

Except. What Wolfers fails to mention is that Clinton was actually the favorite to win in New Hampshire more or less all along. Here’s the price history of the InTrade contract, which expired at 100 when she won:


As you can see, Clinton was a very strong favorite through the fall, and was given a greater than 50% chance of winning right up until the point at which she came third in the Iowa caucus. Then there was a very brief dip before her eventual victory.

With hindsight, it’s perhaps not surprising that New Hampshire’s famously-independent voters turned out not to be as susceptible to Iowa caucus results as the InTrade trading might have indicated. Yes, the final polls in New Hampshire all gave Obama a substantial lead – so substantial, in fact, that one theory has it that independent voters, confident that "their" candidate, Obama, would win the Democratic primary, decided instead to vote for the most independent-minded Republican candidate, John McCain, instead.

It seems to me (again, with hindsight) that this fast-moving primary, with an enormous amount of news and campaigning bombarding a relatively small number of New Hampshire citizens on a 24-hour basis, is probably the last place you would ever want to place your trust in a static opinion poll.

In other words, given the choice between (a) the last-minute polls and the markets got it wrong, and (b) the last-minute polls and the markets got it right, and the chances of a Hillary victory were very small, and in this case it just so happens that the improbable came to pass – given that choice, I’m going to disagree with Wolfers and plump with (a).

This entry was posted in prediction markets. Bookmark the permalink.