August 14, 2005

Dangerous Games

A strong challenge to responsible business behavior comes from recent work in game theory. Modern game theory is scary stuff - researchers run complex simulations in which altruistic and selfish actors compete with one another. The normative has no role here - the best strategy wins, ethical or unethical, socially responsible or not. The results are not encouraging. The good guys usually don't win. Here's one example: For 20 years researchers have held a competition to identify the best strategy for winning a form of the Iterated Prisoner's Dilemma. For many years the champion was the decidedly amoral "Tit-for-Tat". Under this strategy the actor cooperates on the first time, then reviews the behavior of its competitor. If the competitor cooperates, the actor continues to cooperate. But if the competitor cheats, the actor switches strategies and starts cheating, too. (What happens to actors who always cooperate? They lose.) In 2004 "Tit-for-Tat" was finally de-throned, by Southampton, a strategy that introduces collaborative cheating. Crime was, in effect, replaced with organized crime. This is just one example in a complex field, but the more of this stuff I look at, the more I fear that ethics is in danger of becoming a quaint artifact, a relic of a simpler time. This article on business school students suggests that this is not an idle concern. We can't all be cheaters, can we? Maybe not, but perhaps there is some equilibrium level of cheating in our society. Colman and Wilson (1997) did a very interesting game theory analysis showing that amoral behavior (sociopathy) could evolve as a survival strategy, but also suggesting that that there are limits on how many sociopaths are likely to be in a population. The Lotka-Volterra equation, also known as the predator-prey equations, may also have something to say about this. This mathematical analysis describes the dynamics of the populations of a predator species (e.g., lynx), and their prey (e.g., snowshoe hare). But it needn't describe just animals - with a little tweaking it could describe thieves and honest citizens as well. If you think of it that way, you see a couple of interesting things right away. First, an equilibrium tends to set in, and the populations oscillate around equilibrium levels unless something disrupts the equation. Second, it is possible, in some versions of the problem, for the predators to wipe out the prey and then die off, resulting in total extinction for both species. That may seem apocalyptic, but there are certainly human societies where criminal behavior becomes so rampant that there is little incentive to pursue honest labor. Interestingly, markets seem to know when a society has crossed this threshold. In a study that won the 2003 Moskowitz Prize, Charles Lee and David Ng of Cornell University showed that companies in more corrupt countries receive lower valuations from the market. So, people and markets are a little smarter than the models would suggest. Think of what sets humans apart from other species: are we stronger, faster, or gifted with superior senses than other species? No. But we are intelligent, social, and able to communicate with one another in great detail. We can use these capabilities to collaborate to our mutual benefit. Businesses and financiers who forget this do so at their peril.


Post a Comment

<< Home