Monday, October 11, 2004
Economics Nobel announced
Kydland and Prescott won the Economics Nobel this year. Economists and political scientists reading this will be familiar with their work and its impact, but for my other readers, I'll quickly summarize their most famous contribution. Kydland and Prescott noticed that there are some choices we plan to make, and want to make, but when the time comes to make them, we find ourselves temporarily unable to. To choose a few everyday examples, imagine a small kid at the waterpark, who wants to ride the big water slide, and knows that he will enjoy it (say, from past experience), but when he gets to the top of the stairs leading to the slide, gets scared and backs out. Or imagine a dieter who really wants to lose weight, but faced with a donut breaks down for the momentary pleasure of it.
These are examples of what KP called the time inconsistency of optimal plans. The dieter plans to avoid donuts, to lose weight. This plan, if followed, will make the dieter happy. But in the moment of temptation, the dieter finds himself unable to follow the optimal plan.* KP's work has had a big impact because time consistency problems happen to policy makers too. The classic example is a politician who would like to promise not to try to stimulate the economy using monetary policy. Monetary policy can give a short-run jolt to the economy, but with lasting inflationary consequences---so you would want to use it when the economy is slipping into recession, but not on an everyday, "I wish the economy were even better" basis. The "optimal plan" is to resist inflationary monetary policy. But suppose an election is coming, and the politician is worried he might lose. An "even better" economy would be nice insurance. Maybe he'll inflate just this once... The catch is that while no one cares whether the dieter reaches for that donut, the whole economy is watching the politician, wondering if he will stimulate the economy. They know the politician will be tempted to inflate, and so they expect higher inflation---which is to say, they expect dollars to be worth less tomorrow than they are today. And that's a self-fulfilling prophesy. (Unless the politician somehow resists temptation, in which case the economy slows down, due to the expectations he wouldn't resist!) The politician is in a bind, because everyone knows he faces a time inconsistency problem, and they hedge against it. Everyone would be better of it the hedge were unnecessary---and that's only so if the time consistency problem is "solved".
The most popular "solution" to this problem is to find some "commitment mechanism" which reassures observers in advance that the politician must stick to his optimal plan. In the lingo, "he ties his hands", gives up discretion over the policy in the short-run, and thus the time consistency problem vanishes.# Well, if the commitment device is foolproof. If the politician has a backdoor method for subverting that, we're back where we started.
My own research, on central banks and monetary policy, follows in this tradition, but notes that the most popular commitment mechanism, delegation to an "independent" agent, is only a partial solution in practice, because the tempted party often retains various methods for tempting the agent---not least, selecting a potentially pliable agent in the first place. But there is little doubt that a desire to at least mitigate (if not fully solve) the time inconsistency problem underpins a lot of the delegation to independent agencies that happens in politics (the Supreme Court and the Federal Reserve are the key examples, but there are many others).
*Okay, so maybe dieting isn't the best example; behavioral economists might not agree this is a time consistency problem per se. But bear with me.
# The classic story here is from the Odyssey; Odysseus wishes to hear the Sirens, but knows he will be drawn to his death by their call if he does, so he binds himself to the mast of his ship.
These are examples of what KP called the time inconsistency of optimal plans. The dieter plans to avoid donuts, to lose weight. This plan, if followed, will make the dieter happy. But in the moment of temptation, the dieter finds himself unable to follow the optimal plan.* KP's work has had a big impact because time consistency problems happen to policy makers too. The classic example is a politician who would like to promise not to try to stimulate the economy using monetary policy. Monetary policy can give a short-run jolt to the economy, but with lasting inflationary consequences---so you would want to use it when the economy is slipping into recession, but not on an everyday, "I wish the economy were even better" basis. The "optimal plan" is to resist inflationary monetary policy. But suppose an election is coming, and the politician is worried he might lose. An "even better" economy would be nice insurance. Maybe he'll inflate just this once... The catch is that while no one cares whether the dieter reaches for that donut, the whole economy is watching the politician, wondering if he will stimulate the economy. They know the politician will be tempted to inflate, and so they expect higher inflation---which is to say, they expect dollars to be worth less tomorrow than they are today. And that's a self-fulfilling prophesy. (Unless the politician somehow resists temptation, in which case the economy slows down, due to the expectations he wouldn't resist!) The politician is in a bind, because everyone knows he faces a time inconsistency problem, and they hedge against it. Everyone would be better of it the hedge were unnecessary---and that's only so if the time consistency problem is "solved".
The most popular "solution" to this problem is to find some "commitment mechanism" which reassures observers in advance that the politician must stick to his optimal plan. In the lingo, "he ties his hands", gives up discretion over the policy in the short-run, and thus the time consistency problem vanishes.# Well, if the commitment device is foolproof. If the politician has a backdoor method for subverting that, we're back where we started.
My own research, on central banks and monetary policy, follows in this tradition, but notes that the most popular commitment mechanism, delegation to an "independent" agent, is only a partial solution in practice, because the tempted party often retains various methods for tempting the agent---not least, selecting a potentially pliable agent in the first place. But there is little doubt that a desire to at least mitigate (if not fully solve) the time inconsistency problem underpins a lot of the delegation to independent agencies that happens in politics (the Supreme Court and the Federal Reserve are the key examples, but there are many others).
*Okay, so maybe dieting isn't the best example; behavioral economists might not agree this is a time consistency problem per se. But bear with me.
# The classic story here is from the Odyssey; Odysseus wishes to hear the Sirens, but knows he will be drawn to his death by their call if he does, so he binds himself to the mast of his ship.
