What is the purpose of the Nobel Peace Prize?

I watched Obama's Nobel Peace Prize speech earlier today, and I have to say it was funny to see a room of peeved Europeans watch him defend 1. the last few decades of American foreign policy, and 2. the need for war.

I also like that he said his accomplishments are slight compared to those of Albert Schweitzer.

Anyhow, I'm wondering what exactly the point of the Nobel Peace Prize ought to be? It's rather unlike the other Nobel Prizes in that it doesn't reward intellectual accomplishment - the other Nobel prizes are each sort of grand prize for an entire field of endeavor, and it plainly makes sense to award them on the basis of pure merit. For the peace prize, however, it seems to me that the good reasons for awarding it have little to do with the objective merit of the recipient's accomplishments but (in contrast to the other Nobel prizes) are entirely instrumental:

1. it gives an incentive to current and former world leaders to do good things
2. it helps to lend international prestige to movements against various forms of oppression
3. it gives those movements an incentive to remain nonviolent

two of these justifications are ex ante (1 and 3 - they influence people looking to win the prize) and the other is ex post (it directly helps the cause of the people who win the prize). Can you think of other good reasons to give someone the prize?

It seems to me that if we're clear that the main value of the prize is instrumental, it makes little sense to award it to someone at the beginning of their presidency, before they have done much to help world peace. It also seems to me that it makes little sense to award it to NGOs, like the International Committee of the Red Cross, that (while they may be deserving) are unlikely to change their behavior in light of the prize.

Possibly the greatest case name ever

United States of America v. Approximately 64,695 Pounds of Shark Fins (9th Cir 2008) 05-56294

Principal-Agent Problems in the National Football League

I was watching the Greenbay Packers lose to the Tampa Bay Buccaneers the other day. The Buccaneers had just caught an interception with something like 30 seconds to go, and the defender ran it back for a touchdown. My roommate pointed out that (since the Buccaneers had been up by 3) by running the ball back for a touchdown, that defender had actually increased the likelihood of losing the game. Had he just been tackled where he intercepted the ball, the Buccaneers could have knelt it and delayed the game until the clock ran out. By running it back, he ran the risk that the Packers could get a quick touchdown then retrieve an onside kick to get a field goal, tying the game.

The thing is though, interceptions returned for a touchdown are like manna from heaven for receivers. Having a few more of them per season on your stats can make a big difference for your desirability to coaches. So we have a classic principal-agent problem: the individual player's incentives don't line up with the team's.

Greenbay's quarterback quickly demonstrated the principle again. They had 30 seconds left to play, and rather than heaving a hail mary into the end zone he tried to work the ball gradually up the field, basically guaranteeing defeat. The thing is if he had gone for the hail mary the chances of an interception would have gone up, and interceptions are deadly for QB stats. So rather than taking a risk and trying to win the game, he played it conservatively and guaranteed defeat.