Archive for the ‘game theory’ Category
We constantly make judgements, usually subconsciously, about how much to trust others. On average, we seem to be pretty decent at making these assessments- at first. But someone’s trustworthiness is not a fixed quantity. When someone knows that others trust them, they know that people are less likely to look for their possible misbehavior, and more likely to accept their version of events if evidence of misbehavior arises. Some people will take advantage of this trust to commit misdeeds.
The same analysis applies to professions, with the additional mechanism that people with bad intentions may seek out a place in trusted professions in order to commit their misdeeds with a reduced chance of being caught. The very fact that so many people trusted priests not to sexually assault children is what allowed it to happen so often. The very fact that most people trust police not to kill unjustifiably is what allows some to kill unjustifiably.
This seems like a bit of a paradox to me. Is it ever possible for a trusted profession to remain trustworthy for long? How? By trying as hard as possible to select for people of good character? This seems like a hard problem.
The paradox seems like a problem that others must have thought about a lot in many fields- literature, philosophy, and economics at the very least seem like they would be fruitful here. I think a game-theoretic analysis would be interesting, and may have no stable equilibrium. But nothing much comes to mind when I try to think of what others have said about this, besides “who guards the guards?”
What am I missing?
Legal Business: My liquor store is robbed. I call the cops.
Illegal Business: I am robbed for my cocaine.
If I call the cops, I am laughed at or arrested.
If I do nothing, they rob me again every week.
I can only save the business through vigilante justice.
Maybe a kid gets caught in the crossfire.
I have heard otherwise-intelligent people insist that the Federal Reserve is incapable of making believable commitments to take certain actions in the future. For instance, Angus keeps saying things like
“Please repeat after me:
THE FED HAS NO MECHANISM TO BIND ITSELF TO LIVE UP TO ANY ARBITRARY PROMISES IT MAY MAKE TODAY ABOUT THE FUTURE!!!
Promises to act against one’s preferences in the future that are made without any commitment mechanism are simply cheap talk and are extremely unlikely to shape agent’s expectations or actions.”
It is obviously wrong to say that the Fed has no commitment mechanism. The Fed makes believable commitments every day. It does not need to resort to Schelling-style special tricks like having its employees make contracts with 3rd-parties to lose lots of money if the inflation target is missed (though of course these tricks are available and potentially useful). The real commitment mechanism is that your reputation is itself a valuable thing, and the Fed’s leaders know this and have less-than-infinite discount rates. Ben Bernanke has carefully developed a reputation for keeping his word; he is not going to throw this durable good away in 2012, because he will still find it useful in 2013, and he values his 2013-utility a significant positive amount. Because Bernanke knows the importance of the expectations channel, he knows his reputation is more important than just about any one-time action. Ben will not sell his birthright for a mess of pottage.
We don’t even need to delve into the complexities of monetary policy to establish the obvious truth that the Fed can make commitments. Just consider the fact that thousands of Fed employees feel almost certain that the Fed will give them their next paycheck as promised. Surely the Fed would benefit in the short term by not paying them; it could get them to keep working for a few weeks at least by insisting there was some technical problem with the payroll system. Why doesn’t the Fed try this? Because the leaders of the Fed know that their reputation for doing what they say [eg paying people, even if they promised too high a salary] is far more valuable than any short-term savings. Their employees know this too, making it much easier for the Fed to manage their own labor market; for instance, no one is demanding payment up front, because they trust they will be paid later. Similarly, the FOMC announces their meeting dates months ahead of time. Somehow they always manage to meet when they said they would, even if everyone attending realizes at the last minute it is not a great time; they value their reputation more than a one-time inconvenience. That’s some magical stuff right there people!