« Sincere Thanks | Main | The important thing is to keep them pledging »

August 15, 2002

Playing Legal Games

Politics is strategy -- and what better word than politics can describe the scheme of Federal clerk hiring?

It’s hardly a surprise that chinks are starting to appear in federal appellate hiring “consensus.” (Thanks to Howard for pointing out Judge Smith's defection.) It's time for lawyers and legal theorists to wake up and smell the strategy!

A Game of Legal Chess
A Brief Lesson before We Play the Game:
Rational choice theory assumes actors have goals and freedom of choice; the actors attempt to achieve those goals by making choices. We assume an actor pursues his goals by selecting the very best method available to him. A strictly dominant outcome is one that is the best choice for an actor, no matter what anyone else may do.

Strategic behavior arises when this first actor (Player One) interacts with another (Player Two). Player Two knows that Player One will only choose strictly dominant strategies. Player One knows that Player Two will only choose strictly dominant strategies and Player One also knows that Player Two knows that Player One will only choose strictly dominant strategies, &c. &c. This kind of behavior is represented (after a terrible amount of abstraction) in a game theoretic model.

So, let us assume (as we must), that judges are strategic. This strategy is not normally apparent by simple observation -- only by sophisticated modeling techniques. Here, because judges are unconstrained by institutional norms (stare decisis, the value of political capital in sustaining majority coalitions on panels or en banc courts), it is easy to observe.

We will suppose that judges want the very best law students they can possibly hire because a) it will help them write better opinions and b) they will become known as Supreme Court feeder judges. Let us also assume that judges do not care about professional goodwill beyond that, because a) they have life tenure and b) the federal judicial conference really can't tell them who or how to hire.

The classic two-player non-zero sum (i.e., there are possibilities for cooperation) game is the Prisoner's Dilemma. A nice little representation of this game is below.

		Judge Two
		Wait	Defect
Judge One	Wait	(8,8)	(5,10)
	Defect	(10,5)	(6,6)

Key: (Value to Judge One, Value to Judge Two)

As you can see, collectively, the judges are better off if they both choose to wait on hiring. If one judge chooses to defect, the other will choose to defect. Because both judges are rational, they will both preemptively defect, rather than risk waiting.

There is an incentive to take advantage of the other player, even in the face of an agreement, even moreso due to the fact that these "moves" happen sequentially, rather than simultaneously. Here, the game is abstracted to only two players. What happens when there are hundreds of judges champing at the bit to get the very best law students? Whoever jumps ship first gets "more" than the rest. Judges that strike first are able to snag the very top students from the very top schools and do not need the extra information available to them after an additional year of waiting.

What about in the long term? Normally, in a repeated game -- this one is repeated yearly -- that has an indefinite end time, the judges would wait until it is probable {much omission of mathematical abstraction} that it is in their best interest to defect. In price collusion schemes, this time period could be rather lengthy. In fact, colluding companies could raise and lower prices as they collude, take advantage of each other, and collude again (think airlines!).

However, in this game, it makes no sense at all for a judge to wait, then defect, then return to the correct timing. In fact, such a nonsensical judge might find himself clerkless! Moreover, because of the assumptions I have listed above, we find that the payoff in such a repeated game is highly discounted, especially over time (the value of waiting is discounted by the probability that everyone else will defect).

Here are the possible outcomes:
1) Nobody actually adheres to the scheme because everyone knows that nobody will play along.
2) If everyone cooperates, one judge defects to maximize her outcome. Then, everyone else follows her lead. When this judge defects depends on the discount factor - if she thinks the rest will defect soon, she wants to be the first.

There are other outcomes that aren't applicable to the federal hiring scheme because it is illogical for a judge to cycle wait/defect/cycle and judges cannot retaliate against one another.*

It appears that the Federal Judicial Conference will probably follow outcome two -- payoffs aren't considerably lessened by only a few judges abstaining from the hiring freeze, especially since the law students competing for the very best positions are more or less the same. Once the situation reaches a critical mass, everyone will eventually defect from the scheme as the pool of quality applicants declines.

Then, of course, the cycle begins again!

I do apologize to any economists aghast at my explanation. But the question remains, is there any incentive short of legislation we can produce to induce the early defectors to adhere to the hiring freeze? And would legislating a solution even be constitutional?

*It is possible that the judges could publicly ridicule each other for defecting, but it is unlikely that early defectors would actually care (especially if there is intrinsic value for them in bucking the system) and it is certain that the general public would care. Thus, I wouldn't expect this outcome to happen.