September 2019

Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30          

« Libertine Customs | Main | Worst bank in the neighborhood »

Oct 20, 2011


Feed You can follow this conversation by subscribing to the comment feed for this post.

David Boyd

I was just reading something about Charlie Munger. He was saying that we misjudge what incentivizes people constantly. (I'd add that not only do we misjudge what incentivizes people initially, but that those incentives will change over time.) It's the law of unintended consequences.

Take the Greenspan quote for example - “I made a mistake in presuming that the self interest of organizations…was such that they were best capable of protecting their own shareholders.”

I can think of tons of reasons for why this is so, but one would be the responses to the corporate raiders in the 80s. We empowered management at the expense of Gordon Gekko-types because we wanted to stop companies from being broken up and people laid off. One of the things that then manifested itself was that boards and managers were less accountable to the people who owned the company.

This is why designing efficient regulation is so difficult. People do not always react to new rules in ways that we expect them to.

Andrew Brod

What Benkler says makes sense, but he erects a number of straw men. The social-scientific world, even in economics, isn't so monolithically arrayed against common sense.

For example, Benkler claims that in the standard Prisoner's Dilemma game, "everyone should defect," i.e. refuse to cooperate. This is troubling, because collective well-being is maximized when everyone cooperates. But Benkler is correct only if it's a static game. If the game is played only once, then yes, defection is the dominant strategy.

The detectives on "Law and Order" make use of this insight all the time. In the non-TV world, the insight is what site-selection consultants exploit when they try to finagle incentives for their clients. The competition among local governments in these instances is often a one-off. Consequently, when reality matches a static Prisoner's Dilemma game, policy is needed to combat the mutually destructive strategic behavior.

But economists and game theorists have long understood that players of the PD game use cooperative strategies when the game is played sequentially. Reputation effects come into play and sanctions have time to work. In long-ago computer simulations, the winning strategy was one in which cooperation is the default, with brief spurts of non-cooperation used only as punishment for the other player's defection. The result is long spells of cooperation, which matches what we see in real-world examples, for example with product cartels like OPEC.

Even when cooperation happens, it does so because because individuals see it as being in their self-interest. Sure, we're pretty orderly when we stand in line, but it's also clear how self-interest leads any one person to want to enforce the social order when someone jumps the line in front of him.

Benkler refers to political economists Elinor Ostrom, who won the Nobel for her work on common-property resources. Well, she didn't win the award by ignoring the interplay between group cooperation and individual self-interest.

It's certainly important to talk about cooperation. But it's hardly a rebellious new thought in the social sciences.

The comments to this entry are closed.