Human Behavior Dept: Clinton, Obama & the Narcissist's Tale

Going way back to the original linked article, Shankar Vedantam wrote:

The fault line in this dilemma -- the interests of a candidate pitted against the collective interest of his or her party -- shows up in many economic and political domains and is sometimes called the "tragedy of the commons." Individuals embroiled in similar dilemmas find them impossible to solve on their own, because they are confronted by a Hobson's Choice: Act selfishly and cause collective disaster, or act altruistically and aid someone [i]else[/i] who is acting selfishly. Either way, selfishness wins.

I take some issue with the use of the term "narcissistic" -- I don't see anything particularly "narcissistic" in acting in one's own self interests, even if those self interests conflict with the interests of others or even the interest of the community. Nevertheless, the article is correct that elementary game theory can explain why a candidate like Hillary Clinton would stay in a race long after it has been effectively decided (the author uses the "tragedy of the commons" as an explanation -- it can also be analyzed as a "Dollar Auction").

From a game theory perspective, the most interesting thing now is not that Clinton is still in the race, but that some superdelegates are still committing to her campaign. According to game theory, this shouldn't be happening. When there is an advantage to joining the winning side and a disadvantage to joining the losing side, there should be a rush toward the side that is perceived to be winning or which has already won. That's what's known as "bandwagoning."

Now, it's true that some superdelegates, like Reps. Heath Shuler of North Carolina and Chris Carney of Pennsylvania, pledged to support whichever candidate won their districts in the recent primaries, and in both cases it was Clinton. But then there are supers like Rep. Ciro Rodriguez, who made no such commitment and who endorsed Clinton on May 9, after the results of the NC and IN primaries.

According to game theory, that shouldn't be happening. If it's better to be with the winner than the loser, then Obama should be getting all of the late deciders -- and a lot of Clinton's supers should be switching camps. It's probable, then, that Clinton's late endorsers expect that there are more advantages to backing a loser this time than in backing the winner, or that the disadvantages of backing the loser aren't really all that bad. It's not clear to me what the advantages to endorsing Clinton at this late stage might be, but it's the only way to explain an otherwise inexplicable decision.

Well, Rodriguez isn't far off from the other two:



http://www.dallasnews.com/sharedcontent/APStories/stories/D90IEEQ00.html

He didn't promise ahead of time, no. But the popular sentiment has been that the superdelegates should go along with what the voters decide. That sentiment has been gaining traction (and is in Obama's favor, overall.)

I'm not so sure about that. At least in the most recent contests, Obama has been racking up large majorities in urban districts and losing rural districts. For instance, in the April 22 Pennsylvania primary, Obama won only five out of nineteen congressional districts but lost the statewide vote by only 220,000 votes. In Indiana, Obama won three out of nine districts but lost by less than 20,000 votes. His support, therefore, has been highly concentrated in a few districts. I don't know if Obama would necessarily gain an advantage by having representatives follow the will of their congressional districts.

I would also add that, in the case of Ciro Rodriguez, it took him more than two months to come to the realization that the voters in his district supported Clinton in the March 4 Texas primary. It seems rather strange that Rodriguez would use that as an explanation for his decision, on May 9, to endorse Clinton.

True, the last few contests have not been in his favor in that regard -- but in terms of the whole contest, the fact that he has the lead in pledged delegates means that he'd have the lead in superdelegates who hold elected office if they went with how their district voted, right? (I'm not certain, seems that way though...)

The main thing I'm referring to though is that Hillary's only possible remaining avenue to victory is a mass exodus of superdelegates -- the idea that they should hew to what their districts decided rather than just support whomever they want therefore is in Obama's favor.

I don't disagree -- that's part of why I mention that the "superdelegates follow voters" idea is gaining traction (and has since March 4th).

But he may have just dragged it out because the undecided superdels gets lots and lots of attention!

Or maybe his constituents finally convinced him...

Only if the presidential race is the only game these players are playing. It could be a show of commitment that they hope to leverage in other games in the future. Some kinds of senator-to-senator deals perhaps.

***

On the original question, I'd like to adapt a frequently-made remark about paranoia: Even if you're paranoid, they may still be out to get you. Likewise, even if you're narcist, it may still be you who should rule the free world.

As I mentioned:

It's probable, then, that Clinton's late endorsers expect that there are more advantages to backing a loser this time than in backing the winner, or that the disadvantages of backing the loser aren't really all that bad.