So, I’ve talked before about the prisoners’ dilemma:
The interesting feature of this game setup is that the players have an incentive to rat each other out even though this results in a worse outcome for both parties than if they kept quiet. However, this incentive changes when the players are asked to make this choice with each other over and over, since in this setup players might test the waters of cooperation and decide that the one-time gain from screwing over the other person doesn’t outweigh the subsequent losses incurred because the other person won’t cooperate with you anymore.
What most teachers don’t point out is that, in order to sustain cooperation, the game has to NOT have a defined end point. If there is a defined end point, people will want to try to screw each other in the last iteration, since the incentives are like they were in the one-shot game. But then if your opponent is going to screw you in the last game, there is no incentive to cooperate in order to keep his trust, so the players end up screwing each other in the next-to-last iteration. But then there is no incentive to cooperate in the iteration before that, and so on. If people are logical, then they will likely stop cooperating in this game once they learn when the game is going to end.
Obviously, not everything in life is a prisoners’ dilemma, but similar cooperation problems arise in a number of differnet contexts, and in many cases the outcome is highly inefficient. For example, consider the process of a house being foreclosed on. Once the house is sold and the original owner knows he has to move out, there is little incentive to keep the house in a nice condition. In fact, there are many stories of vandalism in these cases, since the original owner is pretty pissed off in addition to having no incentive to do any upkeep. Now, a smart buyer would take into account the potential vandalism/neglect problem when determining how much they are willing to pay for the property, since it’s not like the foreclosed on owner can commit to not destroying the place. This scenario results in both less of a payout to the owner and a crappier property for the buyer, which is certainly not ideal.
To give another example, have you ever wondered why people who get fired from a job often get immediately ushered out of the building, even when, logically speaking, it would be better for them to stay on for a bit to transfer their knowledge and such to other people? It’s to avoid behavior like this:
“So, for the rest of the week, we are going to introduce comedy bits that aren’t so much funny as they are crazy expensive.” Hee. Conan is essentially vandalising his foreclosed on home. You can read more about it here, among other places. Now, I do somehow doubt that NBC actually shelled out $1.5 million for this, but I would also argue that the above bit wasn’t spending in lieu of comedy, it was spending as a complement to comedy, since in context it was pretty damn funny.