Enforcement by Hearing: An Integrated Model of Evidence Production
Columbia Economics Department Discussion Paper No. 9596-03
43 Pages Posted: 18 Mar 1996 Last revised: 1 May 2012
Date Written: December 1995
[This paper was published in two parts as: 1) Relying on the Information of Interested - and Potentially Dishonest-Parties, 3 Am. L. & Econ. Rev. 320 (2001), and 2) Games, Information and Evidence Production: With Application to English Legal History, 2 Am. L. & Econ. Rev. 342 (2000). It also appeared as a working paper under the title: "Enforcement by Hearing: How the Civil Law Sets Incentives."]
The paper is a theoretical analysis of incentive setting via civil litigation, with a focus on incentives for care in activities that may be harmful to others (torts). It makes two main contributions: one directly policy-relevant, one conceptual.
In most existing law and economics research, litigation is modeled in reduced form, as a sort of costly audit, without explicit specification of the fact-finding process. In such models it is always more costly to implement higher levels of care, implying that the second best level of care is lower than the first. Here we explicitly consider the court's information problem (in a single agent model; Sanchirico (1997) considers multiple parties). We find that implementation costs tend to decline in care level and the second best tends to exceed the first. This result in hand, we suggest that the familiar claim that current tort systems overdeter may in fact be no indictment.
A second contribution of the paper is its integrated conception of evidence production, endogenous cost evidence. Evidence production in court is modeled as costly signaling, where signal costs are endogenous to unobserved choices made outside the court room. Care is inspired to the extent that it reduces signaling costs and so increases payoffs at the subsequent proceeding. These signaling/evidence costs are in turn costs of incentive setting via evidence production. In contrast to existing models and conventional wisdom, this view has the ironic implication that perfect (non-falsifiable) evidence may not be good enough: less perfect evidence which is sufficiently-but not infinitely-more costly for disobedient actions is likely to be cheaper all around and thus a more efficient means of setting incentives.
While the paper's focus is on tort law, the model it introduces may be useful in analyzing general situations in which a principal attempts to influence the hidden behavior of an agent based on information that the agent herself supplies.
JEL Classification: K40
Suggested Citation: Suggested Citation