Who Will Be America's (Next?) Next President?

Today the Republicans are, regrettably, partying on in Tampa. I don't want to write about it, because I don't want to pay attention to it beyond reading things that other people write about it, because I can't stand listening to Republican politicians. So, instead, let's gossip about 2016! Because, you see, the British betting site Ladbrokes is already offering odds on the identity of the winner of the 2016 U.S. Presidential election. They're really expansive odds, going down to Karl Rove at 200/1. (There really needs to be a way to short the odds given by these bookmakers.) No one else is listed at worse than 100/1, where we find Julian Castro (D), the Mayor of San Antonio; Jon Huntsman (R), who lost badly in the 2012 primaries; Donald Trump (R?) who, well, you know who he is; Condoleeza Rice (R), the former Secretary of State; Rahm Emanuel (D), the Mayor of Chicago; and Ron Paul (R-ish), who'll be 81. Cory Booker (D), the Mayor of Newark and hopefully the Governor of New Jersey by that point, is the only person listed at 80/1. The 66/1 crowd includes Mike Huckabee (R), David Petraeus (R?), Sarah Palin (R), and Rand Paul (R). At 50/1 we find Deval Patrick (D), Eric Cantor (R), Michael Bloomberg (I), and Rick Santorum (R). Elizabeth Warren (D) clocks in at 40/1, and Bobby Jindal (R) and Chris Christie (R) are at 33/1. Martin O'Malley (D) and Andrew Cuomo (D) are at 25/1, while Jeb Bush (R), Marco Rubio (R), and Joe Biden (D) are at 20/1. Paul Ryan's got 14/1 odds. Then we get Hillary Clinton (D) at 6/1.

And then there's Mitt Romney. He's the favorite, of course, and the odds on him are 5/1. Aside from the existence of these odds in the first place, that's really the point of my post, because that strikes me as an incredibly bearish number for Romney. Let's do the math, shall we? 5/1 odds mean that for every time the event occurs, it would not occur 5 times, so it equates to a prediction that the event in question will occur 1 in 6 times, or about 16.6% of the time. Now, in order to win the 2016 election, Mitt Romney basically needs to win the 2012 election first, and then win re-election four years later. If you assume that the odds of Romney winning each election are the same as each other, if he gets to them, that's an implicit 40.8% chance of winning each election. Now, I don't know about you, but I feel like Romney's a lot more likely to win the 2016 election conditional on his having won in 2012 than he is to win in 2012 right now. Among other things, incumbents win most of the time, and we're sort of slated for a decent patch of economic growth over the next four years, although I'd be the first to tell you that four years of Republican governance isn't the best thing for the trajectory of the economy. So I'd say the implicit odds here are a bit worse than 40% that Romney wins this election.

Okay, that's not particularly bearish, is it? I suppose the FiveThirtyEight numbers give him worse odds right now. But consider the following: bookmakers set their odds so that they get to keep a bit of money, on average. That means overstating the probability of everything. In other words, if they actually though Romney had a 16.6% chance of winning the 2016 election, they wouldn't give him 5/1 odds, they'd give him 4/1, or maybe 7/2 or something. So the 5/1 odds actually give us a slightly lower estimate of that percentage, which in turn gives us a lower estimate of Romney's odds of winning each election separately. Which in turn gives an estimate of the state of play for 2012 that's very similar to the 538 numbers, which are pretty lousy for Romney. So, yet another sign that, as of today, Obama's winning.