We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 13 February 2009
In his most recent article, “The Appeals Process as a Means of Error Correction,” Steven Shavell asks a very important question: Why do we use a hierarchical court structure? The flip side of this inquiry is whether we might not be better off simply making our trial courts more efficient (in the sense of making them less error-prone). Although I certainly applaud the recent efforts of Shavell and other law and economics scholars to examine issues of institutional design, this particular attempt suffers from two major flaws. The first involves the asymmetric treatment of litigants before trial courts and appeals courts: Appellants can choose to either appeal cases or not, but the population of cases before the trial courts is assumed to be exogenous. This severely limits the ability to improve the efficiency of trial courts, leading to an overestimation of the value added from an appeals process. The second problem involves the behavior of judges, who are assumed to behave as automatons. Allowing judges to behave rationally dramatically changes the nature of the model's results, most fatally leading to the elimination of the separation equilibrium that Shavell wishes to achieve.
1. JLS 24:379–426 (1995).Google Scholar
2. “[O]ne must explain … why society may find die appeals process superior to die alternative of enhancing die quality of die trial process.” Shavell at 380–81.
3. This is not the case within Shavell's model. I turn to his assumption/result that only legitimate appeals are brought shortly.
4. An assumption Shavell makes, as well. If they were perfect, appeals would be unnecessary.
5. Shavell at 384.
6. “Under these circumstances, not only may the appeals process result in error correction, it may do so cheaply, for the legal system will be burdened with reconsidering only the subset of cases in which errors were more probably made.” Shavell at 381.
7. This renders false the statement, “This may render society's investment in the appeals process economical by comparison to its improving die accuracy of the trial process—an approach that… would require extra expenditure in every case.” Shavell at 381–82.
8. This problem was first brought to my attention by David Kreps, with reference to a paper of mine about the Medieval law merchant, Trade in a World Most Uncertain: Error and Incentives in the System of the Medieval Law Merchant, under review for IRLE. I am uncertain whether there exists an equilibrium to this game at all. All players, legitimate claimants, illegitimate claimants, and the judge would have to play mixed strategies. If there exists such an equilibrium, it would require the application of a fixed-point theorem to find it I doubt that one exists.
9. Spence, , M., Job Market Signaling, OJE 87:355–74 (1973).Google Scholar
10. Since the state cannot differentiate legitimate appeals from illegitimate ones, it is not possible to charge illegitimate appellants more to bring appeals. The only separation mechanism I can think of is that attorneys may charge clients with illegitimate appeals more to prepare them because it requires more effort to “design” a legal argument for an illegitimate appeal. Contingency fees are likely to eliminate such a separation effect It seems unlikely that there exists much of a cost differential for legitimate and illegitimate appeals.
11. Shavell does recognize this problem in a brief section immediately preceding the discussion. His response is to declare that appeals conn judges should simply choose not to make the inference about appeals separation. “[I]t is best for the rules of procedure to specify that such information should not be considered in judging an appeal” (p. 412). This is, of course, not subgame perfect
