Pushin' Primaries in Iowa and NH

The Associated Press review voter frustrations over the influence of New Hampshire and Iowa in the presidential primaries. The background on the issue is that these two states hold their primaries before all the other states, and candidates who perform badly often drop out before even appearing on another ballot.

Both states have been criticized as unrepresentative of the country given their size and lack of racial diversity. Iowa — population 3 million — is 95 percent white; New Hampshire — population 1.3 million — is 96 percent white. Democrats tried to inject more diversity into the process by adding early contests in Nevada and South Carolina, but Iowa and New Hampshire moved even earlier.

The system became so scrambled last year that New Hampshire Secretary of State Bill Gardner was prepared to move the primary into December to keep ahead of other states that scheduled their own early primaries and caucuses. If anything, the front-loaded calendar made Iowa and New Hampshire more important.

Put your game theory to work now: 1) Use the prisoners' dilemma to explain why a simultaneous game would lead to all states converging to a single point in time further from the general presidential election; 2) Using a sequential game, explain why the New Hampshire SOS Bill Gardner's (credible) threat to continue to move earlier may prevent this convergence outcome. For extra credit, think about the economic benefits of hosting the initial primaries that Iowa and New Hampshire have capitalized into their tax revenues and their constituents' economic profit make their threat to move earlier so credible.

Note that the latter part of the article discusses the argument that New Hampshire and Iowa more closely examine the candidates, weeding out the candidates without enough "substance." How does the median voter model undermine this service?

Game Theory and the War Funding Bill

For those of you who have been watching the news, you are aware of the Democrats' War Funding Bill stipulating an end date for the pull-out of troops from Iraq. This whole topic is loaded with Game Theory applications. First, the pull-out date that the Bush administration opposes is based on game theory. There is a big difference in strategic games outcomes when the game has an ending point versus ones that the players perceive to have no end-date. If you take game theory here later in your schooling you will learn about that. Secondly, Bush promised to issue a veto if the Democrats passed the bill through congress. You can easily imagine a sequential game where this changed the strategy the Democrats pursued. Suppose you are a congressman who wants the political benefit of being in favor of a troop pull-out, but you also don't want to be remembered as the politician who prevented any form of success in Iraq. Now because of the veto you can have both, by supporting the bill in Congress knowing that the policy won't reach that outcome. I think the bill would have been much more likely to fail, or would hve received fewer votes anyway if Bush had not promised a veto. No way to know for sure though.

The Winner's Curse

This post is inspired by Google's aquisition of DoubleClick, in which they outbid Microsoft and others to obtain. There is a well known result of auctions known to economists as "The Winner's Curse," that you would see in an advanced game theory course. This is the idea that we would expect whoever wins the auction probably got a bad deal. Assume that the bidders have similar budget constraints. Prior to going into the auction, each bidder will make estimations about the value of the aquisition (assuming that it's the same for both groups), from which this value would represent the maximum bid. Statistically as a group, we know that on average the bidders would come up with the correct estimate. That is, the bidders will have an equal amount of over- and under- estimation of the true value so that the errors will cancel each other out. However, the winner of the auction will be the bidder who overestimated the true value by the largest amount. It's an interesting idea you should keep in mind with auctions that might fit the assumptions like the Google example above. Another that might work is the Boston Red Sox's bid for the rights to negotiate with Daisuke Matsuzaka.

Update (April 14, 2007): It appears that others believe Google drastically overpaid, supporting my earlier post regarding the Winner's Curse.