Statutes of limitations for accident cases: Theory and evidence

• Pages: 47-67
• DOI: https://doi.org/10.1016/S0193-5895(00)19002-5
• Published date: 06 September 2000
• Date: 06 September 2000
• Author: Matthew Baker,Thomas J Miceli

STATUTES OF LIMITATIONS FOR

ACCIDENT CASES: THEORY AND

EVIDENCE

Matthew Baker and Thomas J. Miceli

ABSTRACT

This paper provides a theoretical and empirical analysis of statutes of

limitations for accident cases. The theoretical model formalizes the trade-

off underlying a finite statute of limitations: while a shorter statute limits

injurers' exposure to liability, thereby curbing incentives for care, it also

limits costly litigation associated with legal error. The model yields

several comparative static results that are tested using cross-state

variation in statutes of limitations for personal injury cases and accidental

damages to property. In order to minimize the impact of historical inertia

in the statutes, we used census data from 1910 and statute lengths from

1916. This represents the earliest period when there were both adequate

census data to construct the explanatory variables, and a convenient

survey of the statutes for the forty-seven states. Despite difficulties in

constructing the data, the empirical model performs reasonably well in

explaining variation in the statute lengths.

I. INTRODUCTION

This paper provides a theoretical and empirical analysis of statutes of

limitations for accidental damage cases. From an economic perspective, the

Research in Law and Economics, Volume 19, pages 47-67.

Copyright © 2000 by JAI/Elsevier Inc.

AH rights of reproduction in any form reserved.

ISBN: 0-7623-0308-5

47

48 MATI'HEW BAKER & THOMAS J. MICELI

trade-off underlying a statute of limitations in general is well-known. 1 On the

one hand, a statute limits injurers' exposure to liability, thereby reducing

incentives for care, but on the other hand, it reduces legal error resulting from

stale evidence and saves on litigation costs. The optimal period of limitation

balances these two effects. The theoretical contribution of this paper is to

formalize this trade-off in the context of a simple accident model with legal

error and litigation costs. 2 The benefit from developing a formal model is that

it yields specific comparative static results that can form the basis for empirical

analysis.

To our knowledge, the only existing empirical analysis of statutes of

limitations is by Netter, Hersch & Manson (1986), in the context of claims for

title under adverse possession law. As in their analysis, we examine the

determinants of cross-sectional variation in statute lengths across states, but we

apply the analysis to statutes of limitations for property damage cases and

personal injury cases. The empirical model performs reasonably well given that

data and practical limitations prevented us from keying the analysis to the years

when individual statutes were passed. In general, the model performed better

for the property damage statutes than for the personal injury statutes. This may

reflect greater ease in measuring property losses in economic terms.

This paper is organized as follows. Section II develops the basic accident

model and shows that, in the absence of legal error, there is no basis for setting

a finite statute length. Section HI, therefore, introduces legal error and derives

the optimal statute length based on the trade-off described above. It goes on to

derive several comparative static results that form the basis for the empirical

analysis. Section IV describes the data and presents the results of the empirical

tests of the model. Finally, Section V concludes.

II. THE BASIC MODEL

To examine the impact of a statute of limitations, it is necessary to assume that

not all victims file suit immediately following an injury. One reason for delay

by victims in filing suit may be that the injury is not immediately evident.

Another is that the injurer might avoid immediate detection. (Although we will

not explicitly model the latter factor, we will take it into account in specifying

the empirical model in Section IV.) For purposes of the model, we abstract

from the specific reason for delay and simply treat the length of time after an

accident that a plaintiff files suit, denoted t, as a random variable with

distribution function G(t), where G'------g>0, G(0)=0, and G(oo)= 1. Thus,

under a statute of limitations of length L, a fraction G(L)< 1 of all possible

suits will have been filed by time L.