WEAI/AERE 2009 - Individual Paper Abstract
Title: Environmental compliance through endogenous audit mechanisms
Author(s): Scott M. GILPATRIC, Dept. of Economics, 505A Stokely Management Center, U. of Tennessee, Knoxville, TN 37996; sgilpatr@utk.edu; Michael McKee, Dept. of Economics, Walker College of Business, Appalachian St. U., Boone, NC 28608; mckeemj@appstate.edu; Christian A. Vossler, Dept. of Economics, 505A Stokely Management Center, U. of Tennessee, Knoxville, TN 37996; cvossler@utk.edu
Abstract:
This paper adds to the environmental compliance literature through the development and experimental testing of two endogenous audit mechanisms designed to provide incentives for firms to accurately disclose their emissions. The theoretical analysis has disclosure as a continuous choice and allows the possibility that audits do not perfectly reveal actual emissions. Suppose an industry consists of N firms and the regulator can conduct k<N audits. In principle the regulator seeks to audit the k firms that disclose the least. However, in reality firms are not identical and comparing disclosure levels is likely to be problematic. At the very least firms may produce different levels of output and the disclosed emissions can then be converted to a "per unit of output" measure for comparison among firms.
The first audit mechanism is an endogenous "tournament" audit mechanism. If the regulator observes firm's output imperfectly this adds error to the process of determining which firms to audit. The regulator ranks the firms according to their error-adjusted reports and the k firms with the lowest error-adjusted reports are audited. A similar but distinct endogenous rule arises if the probability a firm is audited is an increasing function of the difference between its report and the average of the other firms reports. We refer to this as the endogenous "yardstick" mechanism.
Under either endogenous mechanism the Nash Equilibrium level of disclosure is strictly greater than with simple random auditing. One interesting result is that high levels of compliance may arise in the endogenous audit settings even when the equilibrium audit frequency is so low that if audits were purely random firms would not comply at all. Compliance occurs because otherwise firms stand out relative to the industry, and because of this face a high probability of an audit, which may be costly in itself even if the likely fines aren't that high.
The theoretical analysis is complemented with a series of laboratory experiments designed to test theoretical predictions and provide empirical evidence of the effectiveness of the endogenous audit mechanisms relative to random audits. The experiments vary the audit probability, the cost of disclosed emissions and the cost of revealed undisclosed emissions. Further, we look at the effect of varying the quality of the audit itself through treatments with perfectly revealing audits, imperfect but unbiased audits, and biased audits.