Lawyers as Agents of the Devil in a Prisoner's Dilemma Game

• Date: 01 September 2013
• Published date: 01 September 2013
• DOI: http://doi.org/10.1111/jels.12014
• Author: Gordon B. Dahl,David E. Bloom,Orley Ashenfelter

Lawyers as Agents of the Devil in a

Prisoner’s Dilemma Game

Orley Ashenfelter, David E. Bloom, and Gordon B. Dahl*

Do the parties in a typical dispute face incentives similar to those in the classic prisoner’s

dilemma game? In this article, we explore whether the costs and benefits of legal represen-

tation are such that each party seeks legal representation in the hope of exploiting the other

party, while knowing full well that failing to do so will open up the possibility of being

exploited. The article first shows how it is possible to test for the presence of such an

incentive structure in a typical dispute resolution system. It then reports estimates of the

incentives for the parties to obtain legal representation in wage disputes that were settled by

final-offer arbitration in New Jersey. The article also reports briefly on similar studies of data

from discharge grievances, court-annexed disputes in Pittsburgh, and child custody disputes

in California. In each case, the data provide evidence that the parties face strong individual

incentives to obtain legal representation that make the parties jointly worse off. Using our

New Jersey data, we find that expert agents may well have played a productive role in

moderating the biases of their clients, but only early on in the history of the system. Over

time, the parties slowly evolved to a noncooperative equilibrium where the use of lawyers

becomes nearly universal, despite the fact that agreeing not to hire lawyers is cheaper and

does not appear to alter arbitration outcomes.

The nickname Prisoner’s Dilemma, attributed to A.W. Tucker, derives from the original anecdote used to

illustrate the game. Two prisoners, held incommunicado, are charged with the same crime. They can be

convicted only if either confesses. Further, if only one confesses, he is set free for having turned state’s evidence

and is given a reward to boot. The prisoner who has held out is convicted on the strength of the other’s testimony

and is given a more severe sentence than if he had also confessed. It is in the interest of each to confess whatever

the other does. But it is in their collective interest to hold out.

Rapoport and Chammah (1965)

I. Introduction

This article explores the possibility that the parties to a typical dispute face incentives very

similar to those faced by the prisoners in the classic dilemma described above. The goal is

to explore the possibility that the costs and benefits of legal representation are structured

*Address correspondence to Gordon B. Dahl, Professor of Economics, University of California, San Diego, 9500

Gilman Dr. #0508, La Jolla, CA 92093; email: gdahl@ucsd.edu. Ashenfelter is Joseph Douglas Green 1895 Professor

of Economics, Princeton University; Bloom is Clarence James Gamble Professor of Economics and Demography,

Harvard University.

The authors are grateful to the Russell Sage and Alfred P. Sloan Foundations for financial support. An early

version of this article, with limited data on arbitration cases in New Jersey, appeared as a 1993 NBER working paper.

bs_bs_banner

Journal of Empirical Legal Studies

Volume 10, Issue 3, 399–423, September 2013

399

so that each individual party seeks representation in the hope of exploiting the other party,

while knowing full well that failing to do so will open up the possibility of being exploited.

In the prisoner’s dilemma, the players are faced with incentives that may lead to

behavior that is not in the best interests of either party. The incentive structures embedded

in such institutions provide a payoff to each party for engaging in noncooperative behavior.

The mere existence of such incentives is not enough to produce the prisoner’s dilemma,

however. To produce inefficiency, the payoffs to noncooperative behavior must be larger

than the costs. Determining whether a particular institution embodies the incentives asso-

ciated with the prisoner’s dilemma thus requires determining whether the individual

benefits from engaging in noncooperative behavior are likely to exceed the costs.

The first part of this article shows how it is possible to test for the presence of an

incentive structure like the one in a prisoner’s dilemma game in a typical dispute resolution

system. The goal is to measure the payoff matrix faced by the parties. Although the analysis

is framed in the context of an arbitration system, it could equally as well be applied to any

dispute resolution system, including the court system. The second part of the article reports

estimates of the incentives for the parties to obtain legal representation in wage disputes

that were settled by final-offer arbitration in New Jersey.1The article also reports briefly on

similar studies of data from (1) the arbitration of discharge grievances, (2) the arbitration

of court-annexed disputes in Pittsburgh, and (3) the settlement of child custody disputes in

California. All of these data provide evidence that the parties face strong individual incen-

tives to obtain legal representation. Whether these incentives lead to excessive legal repre-

sentation that has no social benefit is discussed in the final section of the article. Using our

New Jersey data, we find some evidence that lawyers may well have improved outcomes by

moderating client biases early on in the history of the system. However, over time, the

parties slowly evolved to a noncooperative equilibrium where the use of lawyers becomes

nearly universal, despite the fact that agreeing not to hire lawyers is cheaper and does not

appear to alter arbitration outcomes.

II. Incentives for Noncooperation

The theoretical literature on the economics of litigation, particularly game theory applied

to law, is extensive. In contrast, there has been less empirical work, consisting of a few

experimental papers, on game theory and litigation and some empirical work in field

settings.2The main contribution of this article is to lay out a simple model of the role

lawyers play in dispute resolution, and empirically test whether the incentives present in

such systems generate a prisoner’s dilemma.

1A related, but more technical, analysis of New Jersey’s arbitration system can be found in Ashenfelter and Dahl

(2012).

2For a summary of theoretical models of negotiation and legal services, see Spier (2007). Prominent papers in this

literature include Katz (1988), Kerkmeester (1995), P’ng (1983), Reinganum and Wilde (1986), Rong (2012), and

Spier (1994). Experimental papers on game theory and litigation include Ashenfelter et al. (1992), Kritikos (2006),

and Plott and Coughlan (1997). Greiner and Pattanayak (2012) provide an example of empirical work in a field

setting.

400 Ashenfelter et al.