Optimal Fines under Announced and Surprise Inspections

• Published date: 01 October 2016
• DOI: http://doi.org/10.1111/jpet.12166
• Date: 01 October 2016
• Author: EMMANUEL DECHENAUX,ANDREW SAMUEL

OPTIMAL FINES UNDER ANNOUNCED AND SURPRISE INSPECTIONS

EMMANUEL DECHENAUX

Kent State University

ANDREW SAMUEL

Loyola University in Maryland

Abstract

This paper examines optimal fines in a regulatory framework where

the regulator can choose either surprise or announced inspections to

monitor a firm for compliance. The firm can invest in detection avoid-

ance, but it receives a fine if the regulator discovers that it is noncom-

pliant. In the welfare maximization problem, we focus on the trade-off

between the frequency of inspections and the magnitude of the fine.

We find that when inspections are unannounced, the optimal fine is

maximal, but when they are announced, the optimal fine may be less

than maximal.

1. Introduction

The question regarding whether fines should be maximal has received considerable

attention in the law and economics literature. Becker’s (1968) early work in this area

showed that fines should be maximal when raising the probability of detection is costly.

Specifically, his argument is that if violators are risk-neutral, then by lowering the de-

tection probability while raising the fines, a regulator can maintain compliance while

lowering costs. Thus, a welfare-maximizing regulator will raise fines to their maximal

level (see also Garoupa 2001).

Extending Becker’s work, Malik (1990) allows for the possibility that the agent can

invest in avoidance activities to reduce the probability of detection. In his model, the

probability of detection is increasing in the regulator’s level of monitoring and decreas-

ing in the violator’s level of avoidance. He shows that if detection avoidance activities

are costly, the optimal (welfare-maximizing) fine is not maximal because raising the fine

can increase the agent’s incentive to hide. Malik’s key insight is to recognize that raising

Emmanuel Dechenaux, Department of Economics, Kent State University, Kent, OH 44242

(edechena@kent.edu). Andrew Samuel, Department of Economics, Loyola University in Maryland,

4501 N. Charles Street, Baltimore, MD 21210 (asamuel@loyola.edu).

We are grateful to Andrew Daughety, Hyunseok Kim, Jennifer Reinganum, two anonymous refer-

ees, and an associate editor for helpful comments. We also thank seminar participants at Cleveland

State University and the conferences of the Association for Public Economic Theory (July 2013), the

Canadian Law and Economics Association (September 2013), the French Economic Association (June

2014), and the Western Economic Association (July 2014). Financial support from Loyola University

Maryland’s summer research grant program is also gratefully acknowledged.

Received March 2, 2015; Accepted March 3, 2015.

C2016 Wiley Periodicals, Inc.

Journal of Public Economic Theory, 18 (5), 2016, pp. 786–801.

786

Optimal Fines 787

the fine is not a costless transfer from agent to regulator, because raising the fine in-

creases the equilibrium level and cost of avoidance. Thus, raising the fine and lowering

the monitoring level is not necessarily a cost-reducing policy maneuver.

Langlais (2008) extends Malik’s work to study the effect of the regulator’s level of

monitoring on the marginal effectiveness of avoidance activities. He focuses on the case

where monitoring enhances the marginal effectiveness of avoidance, a case we refer to

as substitutes. He shows that raising the fine while lowering the level of monitoring leads

to two opposing effects. First, raising the fine increases avoidance activity and its associ-

ated costs, as in Malik (1990). However, lowering the level of monitoring now lowers the

marginal effectiveness of avoidance, thereby reducing the agent’s incentive to hide. If

the second effect dominates the first, then the fine should be raised to its maximal level

and Becker’s result may once again apply. The role played by avoidance activities in low-

ering the effectiveness of enforcement policies has received much attention recently, as

surveyed by Sanchirico (2010).

An issue that the literature on the optimal fine has not yet addressed concerns the

nature of the detection or inspection regime. Regulators typically can choose between

inspecting their agent’s continuously or occasionally. For example, while the United

States Department of Agriculture continuously monitors meat packing plants, it moni-

tors other types of food-processing plants only occasionally. When inspections are con-

ducted occasionally, a regulator faces a further choice of inspecting agents by surprise

(unannounced inspections) or to announce the inspections (Chin 1999; Lazear 2006;

Dechenaux and Samuel 2014). For example, the Transportation Security Agency in-

spects every traveler’s baggage (continuous inspections) whereas the New York subway

subjects passengers’ bags to random searches (occasional, unannounced inspections).

Intuition suggests that when the agent can engage in avoidance activities that reduce the

probability of being detected, surprise inspections may be particularly effective because

they allow the regulator to catch the agent off-guard.

In order to study this issue, we develop a model of inspections that builds on

Dechenaux and Samuel (2014). In this model, a regulator inspects firms to detect

whether they are compliant with some regulation. The regulator specifies the inspection

regime; that is, whether inspections are announced or unannounced. It also chooses

the frequency of inspections, the monitoring intensity, and the fine that a noncompli-

ant firm must pay if it is detected. Taking this regulatory environment as given, the firm

chooses its level of avoidance, which is not observed by the regulator. Following Malik

(1990), the probability of detecting a firm’s noncompliance is increasing in inspection

effort and decreasing in the firm’s level of avoidance activity. As in Langlais (2008),

monitoring intensity enhances the marginal effect of avoidance activity on the probabil-

ity of detection. Under reasonable assumptions on the probability of detection, we find

that whether the fine is maximal depends critically on the inspection regime chosen.

In an unannounced regime, the optimal fine is always maximal even when avoidance is

costly. Whereas in an announced regime the optimal fine is not always maximal.

The analysis in this paper differs in several ways from Dechenaux and Samuel

(2014), Langlais (2008), or Malik (1990). First, Dechenaux and Samuel (2014) study

the compliance-maximizing inspection regime within the context of a principal–

supervisor–agent model where the supervisor is subject to moral hazard. Thus, the

principal, or regulator, needs to use incentives, such as a reward for a successful in-

spection, to encourage the supervisor to monitor the firm. Furthermore, in that paper

the regulator chooses the inspection frequency (probability), while the fine and the

supervisor’s reward are exogenous. In contrast, this paper focuses on welfare maximiza-

tion in a simpler inspection framework, where the regulator can directly choose the