Game theory and judgment

Everyone knows that judgment is the most important quality a lawyer can have.  There is some dispute over whether it can be taught or only learned.  I am inclined to the latter view.  I do not know whether clinical training does more for judgment than classroom teaching, as some have suggested (see Luban & Millemann, GJLE 1995), but for those of us who do not teach clinical courses the problem remains: the most important thing is something we cannot teach but must be learned by students who are reasonably but short-sightedly fixated on the MPRE. 

Worse yet, judgment is notoriously hard to pin down.  How can we teach it if we cannot say what its elements are?  We can show examples of good judgment and bad, but unless we can specify some generalizable aspect of the examples, some facility the exercise of which the examples reveal, there is no reason to expect them to be of much benefit in learning a generalizable skill. 

I don't claim to know what judgment is, but I do think two important parts of it can be specified.  The facility of judgment is dynamic and interdependent.  That is to say, to exercise good judgment one must think ahead, several steps down the road, and one must think in terms of how those affected by some decision will react to it.  I decide, you react, I react to your reaction, you react.... and so on.  A lot of mistakes are made by people who never think in terms of an endgame, and who fail to think about how others will see them and react to their actions.

Game theory is a branch of economics that focuses on dynamic and interdependent decision-making.  I think its principles are important to the PR curriculum.  I don't want to suggest it is a perfect substitute for other sources of judgment, but it is very valuable in demonstrating the importance of thinking dynamically and interdependently. (With the Nobel Memorial awards this year, it also is the subject matter of two prizes, including John Nash's.)

The most familiar example is, of course, the prisoner's dilemma.  I suppose everyone does this, but I use it to introduce criminal conflicts and joint defense agreements.  Once you show students the coordination problem, you can present joint representation as a solution (the Edward Bennett Williams old-school "everyone inside the tent pointing out (I bowdlerize), no one outside the tent pointing in) approach. Conflict rules can disrupt coordination (through counsel) that solves the game, thus creating a risk of defection (confession in the classic game) that joint representation minimizes.  Joint defense agreements can be seen as another move in the structure of the game, contractually re-creating the confidentiality joint representation previously provided, and which might facilitate cooperation, but which might not be available because of conflicts, absent the JDA.

Another example, also involving the prisoner's dilemma, can be found in Robert Axelrod's wonderful book The Evolution of Cooperation (retrospective here), and its  endorsement of tit-for-tat as an optimal strategy for repeated games.  Tit-for-tat is a strategy in which you cooperate in round one of a game and then repeat the other player's move in each subsequent round.  I.e., you cooperate if they do, but you defect if they do, too.  Axelrod ran computer simulations showing that this strategy best induced cooperation.

How on earth could that apply to real lawyers?  Discovery conduct is one example, on which I can offer an anecdote.  I had got so fed up with the posturing and time-wasting of discovery spats, that once I tried explaining to opposing counsel that I would play discovery by cooperating in our first encounter and then by following whatever they did.  If they cooperated rather than pounded the table and made unreasonable demands, I would cooperate too.  If not, I would respond in kind. 

I cannot vouch for the effect of the statement--talk is cheap--but I can vouch for the effectiveness the strategy. Very early on you learn that you if you ever want a break yourself, you have to be willing to give one, but if you always give breaks as a strict rule, you may never get one.  You have to be willing to respond in kind.  That doesn't mean breaking any rules (I am talking here about things like agreeing to extensions and scheduling things to accommodate the other side), but it does mean exercising discretion in a way that makes cooperation more likely.  It would be nice if things were different, but they aren't.

There are problems with applying game theory to law.  If you add players and variables to approximate real-world decision-making, things can quickly get so complicated that results become indeterminate, which one might describe as "multiple equilibria are possible," re-translatable as "it depends."  But its not as if moral philosophy is conclusive and determinate, either.  The principles of dynamic and interdependent thinking are central to judgment, which is central to law, and game theory helps us think about them.

DM