David Friedman has authored a penetrating article on how game theory can help escape the Hobbesian jungle. In it he uses the notion of Schelling points to make his case.

A Schelling point is (in a game) a unique choice among many which is chosen by all people playing the game who cannot talk with one another precisely because of its uniqueness. It makes possible coordination without communication. Thus, if two people are given a series of numbers 2, 5, 9, 25, 69, 73, 82, 96, 100, 126, 150 (Friedman’s example) and can win a prize if both select the same number, they can succeed if they know which number the other person interprets as uniquely special, even though the concurrence in picking any number will suffice. Thus, 2 might be picked by both, because it is the only smallest prime number. 100 might be picked, because it is a nice round number. Since there are three perfect squares in the series, it is less reasonable to try to pick one of them, because even if both for some reason like squares, they will still have only 1/3 chance of selecting the same one. Thus, a Schelling point is “each person’s expectation of what the other expects him to expect to be expected to do.”

Does Friedman’s theory apply to “social” contracts? For example, is the US Constitution a Schelling point as an agreement between the federal government and the people / states? The federal government can be either limited or unlimited; but if it is unlimited, then no one knows exactly which form it will take or how big it will be. This uncertainty will please no one. But if it limited, then we can specify or enumerate all of its legitimate functions. This can be a small list or this can be a big list, but it will limit the government. Such a government is a Schelling point; it is a unique solution to politics. It so happened that the point settled initially on a very small government. As the ideology has changed from the days of the founding of the US into something far more statist, it would still have been better to obtain the vastly bigger government that we have now by amending the Constitution than by retiring it altogether, which is the de facto situation today.

The recommendation to the people then is not to wait until “a long train of abuses and usurpations” has left its mark on the country but rather to revolt at the smallest sign of corruption of which they should be eternally vigilant. This is because any exceeding of government authority, if not checked right away, will embolden the government to try to do the same again and again, thus moving the agreement further away from the Schelling point and encouraging the government to make unlimited demands.

For if the first abuse is tolerated, then what argument remains for not tolerating all the future abuses? Give the government an inch, and it’ll take a mile. So, punishing unconstitutional acts immediately and ruthlessly (e.g., by tarring and feathering the President) is a good policy even if it seems initially out of proportion to the crime committed by a public figure. This strategy will install essentially a revolution in permanence, because the government will always try to evade its restrictions.

Why shouldn’t a Mafioso (as in The Sopranos) eat pussy? Because if he eats pussy, he’ll eat anything.

Similarly, the people ought not to allow government crimes, no matter how small, because tolerating them means that the government will be motivated to rape the people with ever increasing contempt and boldness.

So, it seems that enforcing Schelling points is deterrence with a vengeance.

Jefferson thought there ought to be rebellions at least every 20 years or so, so that the rulers are “warned from time to time that this people preserve the spirit of resistance.” He underestimated the folly of the American people. From a shining city on the hill America has turned into an ant hill. And this has been partly due to disregarding game theory.

Categories: From the Vault Redux