My colleaque, political scientist Tim Feddersen, says he does not understand recent U.S. elections. He understood Bill Clinton, Dick Morris and triangulation. He understood George W Bush and “compassionate conservatism”. These are all consistent with the logic suggested by the median voter theorem: politicians should adopt policies that will win the support of the median voter (forget the fact that the politicians will renege on their promises as soon as they get in office!). But George W Bush was not a compassionate conservative in 2004. And for some reason, Tim can’t rationalize how President Obama got elected via the median voter theorem.
So there is a competing theory which is less well worked out, the ideological voter theory. In this theory, it is the extremes that determine elections so the candidate must play to them. One version of the theory would include turnout as a second dimension of voting strategy as well as how the citizen actually votes once she gets in the voting booth. Extremists have lower costs of turning out so the payoff to motivating them at the cost of alienating independents is greater than the benefit to motivating independents at the cost of not motivating extremists. This theory would say Santorum is the biggest threat to Obama. Watching Santorum’s speech the other night, I was moved by his story about his grandfather’s life as a coal-miner while I found Romney’s speech formulaic and artificial. So, I have some sympathy for this theory. (Of course, Jon Stewart has yet to fully explain to me what Santorum’s policy positions are If they are as extreme as my preliminary google search suggests, I might switch back to the median voter theory.)
So, the ideological voter theory suggests that Santorum is Obama’s most potent adversary. This implies Obama supporters should hope that Romney comes out alive from the Republican primaries. (I have yet to incorporate Larry Kotlikoff’s candidacy in my world view though I completely endorse Jeff’s fulsome praise. I visited BU Econ a couple of years ago and it has a great “corporate culture” that I too attribute to Kotlikoff.) But Romney has plenty of money so no need to bankroll him. Just give it to Obama.
Cheap Talk 
4 comments
Comments feed for this article
January 5, 2012 at 11:35 am
David Fibush
This supports my view that, where possible, Obama supporters should vote for Romney in open primaries. It would be a minor disaster if Romney became President. Santorum would be a major disaster.
January 5, 2012 at 12:12 pm
Sandeep Baliga
Your view is based on risk aversion alone. It leads to the same conclusion as the theory above which is not based on risk aversion but a model of turnout.
January 5, 2012 at 12:53 pm
kerokan
Your model of turnout should probably include a “fear” factor. The median voter who would not bother to vote when candidates are appealing “moderately” to the extremes may in fact become more likely to turnout if a candidate becomes too extreme.
What I mean is, suppose policy space is [0,1] and MV is at 0.5. MV does not turnout as long as candidates are in [0.3, 0.7]. But if a candidate moves to 0.2 or 0.9, then MV bothers to vote against him.
January 9, 2012 at 10:51 am
Rick Almeida
I believe this article refers to what’s called the “directional” version of the spatial theory of choice (as opposed to the proximity formulation).