SEARCH

SEARCH BY CITATION

Keywords:

  • flexible enforcement;
  • optimal penalty;
  • responsive environmental regulation

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices

This paper considers the level of, and changes in, optimal noncompliance penalties under the following conditions: (i) where the regulator responsible for setting policy parameters, such as a penalty, is different from (and thus may have a different objective from) the regulator responsible for enforcing existing regulations; and (ii) where enforcement behavior changes from one in which enforcers are unresponsive to overtures on the part of firms to increase compliance to one in which enforcers are responsive to such overtures. The model developed shows that when enforcers “switch” from unresponsive to responsive enforcement, the optimal penalties for noncompliance need to be reduced. The analysis also gives insights as to what variables dictate the degree of penalty reduction.


1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices

In recent years, economists, policy analysts, and other political and social scientists have developed a substantial body of research devoted to the study of corporate social responsibility (CSR). Vogel (2005) and Hay et al. (2005) synthesize and extend this research and serve as examples of the current prominence of CSR both in the social sciences and in policy analysis. This focus has, in large measure, been driven by an increasing number of instances whereby firms appear to be voluntarily choosing to behave in this type of “self-regulatory” fashion as well as a proliferation of both private and government-sponsored certification programs that seek to induce firms to act in a socially responsible manner.

Corporate social responsibility can come in many forms, from limiting effluent releases of unregulated pollutants beyond what is required by law, to voluntarily investing in and implementing worker safety devices and practices, to voluntarily participating in government-sponsored regulatory compliance programs (e.g. Vogel 2005, pp. 131–132).

Existing empirical work has primarily focused on ascertaining the characteristics of firms choosing to self-regulate or participating in voluntary programs (e.g. Prakash & Potoski’s [2006] in-depth analysis of firm decisions to seek ISO 14001 certification), while the theoretical work has focused on more normative questions. Several studies that are discussed here have concluded that many programs designed to induce voluntary compliance are not necessarily socially beneficial. Perhaps this is somewhat surprising, given such programs’ emphasis on reduced governmental intervention generally considered to be inefficient, and the reliance on enhancing internal incentives of firms to engage in socially responsible activities.

Deviating from the broader empirical direction, this paper seeks to contribute to the normative literature on government-sponsored voluntary compliance programs through the construction of a formal principal–agent model that focuses attention on firm behavior and regulatory structure. Specifically at issue is the optimal level for noncompliance penalties under two conditions. The first involves a regulatory governance structure whereby the regulator responsible for setting penalty levels and the regulator responsible for enforcing existing law are different actors with different objectives. The second condition involves the enforcement regulator altering his traditional policy of periodically inspecting firms for noncompliance to a more “responsive” type of enforcement behavior whereby the level of enforcement intensity is a direct function of a given firm’s compliance efforts. Hence, improved firm-level compliance will now be met with reduced regulatory scrutiny. In language more consistent with the criminology literature, this switch can be thought of as a change from a “deterrence model,” where the regulator seeks to punish firms that violate law, to a type of “bargaining model,” where the regulator seeks to persuade firms to voluntarily improve compliance (Gormley 1998). Here, the enforcement regulator’s goal is to persuade a given firm to improve compliance so as to receive a subsequent “reduced scrutiny” reward.

As will be discussed in more detail, both of these conditions are of some importance. With respect to the first, much social regulation, in particular environmental and worker safety regulation in the US, shows this type of multiple-principals governance structure, but has generally received little scholarly attention. In so doing, key policy insights are likely being overlooked. With respect to the second, many government-sponsored voluntary compliance programs do indeed focus specifically on enforcement. An effective understanding of how policy parameters, that is, penalties, need to be adjusted, if at all, is thus necessary if socially desirable outcomes are to be achieved.

With this basic structure, the main outcome, or condition, that the model produces indicates that a switch on the part of the enforcement regulator to responsive enforcement should be met with a reduction in the optimal penalty. This condition, indicating monetary limits on fines for noncompliance, speaks directly to a much broader social science literature on the economics of crime and punishment sparked by Becker (1968), whose primary insight was that as deterrence can be achieved by either increasing the penalty (generally considered to be a costly transfer) or increasing enforcement effort (generally considered costly), penalties should be set at the highest level possible.

The remainder of this paper is organized as follows. In Section 2, a brief review of voluntary compliance programs in the context of responsive regulation is discussed. Specific attention is directed toward emphasizing the need for the governance structure assumed here and the benefit of focusing on enforcement behavior specifically. In addition, a brief review of the existing theoretical work on optimal penalties is offered. In Section 3, the basic structure of the model is presented and some of the key assumptions are discussed. In Section 4, the model is solved assuming the enforcement regulator is not responsive to the firm’s observed investment in improved compliance. In Section 5, the model is resolved under the assumption that the enforcement regulator does indeed respond to the firm’s investment in compliance and the optimal fines under the two cases are compared. Section 6 concludes and offers avenues for future research.

2. Voluntary compliance programs and optimal penalties

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices

2.1. Voluntary compliance and responsive regulation

The idea of responsive regulation has been around for some time and is rather extensive. The basic notion behind responsive regulation is that effective regulation needs to be attuned to the diversity of characteristics and motivations of the regulated community. Moreover, effective regulation should be prepared to respond to industry conduct: “The very behavior of an industry or the firms therein should channel the regulatory strategy to greater or lesser degrees of government enforcement” (Ayres & Braithwaite 1992, p. 4).

It is perhaps this idea that is behind the establishment of certain voluntary compliance programs that focus particular attention on enforcement. For instance, under the US Environmental Protection Agency’s (EPA) National Environmental Performance Track program, firms choosing to participate must establish in-house, self-auditing compliance programs, conduct periodic self-audits, institute plans and corrective actions should compliance problems be found, and submit progress reports to the EPA office and the public. In return for these efforts, participating firms are rewarded in several ways, one of which is that they will be considered “a lower inspection priority.”1 In a similar light are recent changes to the US EPA’s general audit practices designed to encourage regulated firms to voluntarily discover, disclose, and correct any violations of existing environmental statutes. In exchange for these efforts, the EPA states that it will, among other things, “refrain from additional audit requests.”2

This enforcement behavior appears in worker safety regulation in the US as well. In 1982, the Occupational Safety and Health Administration (OSHA) established the Voluntary Protection Program (piloted in California in 1979) whereby firms who choose to participate must undertake efforts, in conjunction with OSHA, to improve worker safety conditions on site, not only to better meet current mandatory requirements, but also to exceed such requirements. In response, participants benefit in several ways. Prominent among such benefits is that participants are “removed from OSHA programmed inspection lists for the duration of their participation.”

Note that in these cases, responsive enforcement, offered on a firm-by-firm basis, must be earned. That is, the firm must do something to improve compliance, such as invest in cleaner pollution control technology, to receive the regulatory benefit. This feature of responsive enforcement figures prominently in the model developed here. It should also be emphasized that these programs are principally enforcement focused.

Moreover, in many regulatory arenas in the US and other countries, there tends to be a disconnection between policy construction, generally set at the broader federal level, and implementation (e.g. enforcement) of the law, generally delegated to state governments (Gerber & Teske 2000). This is true, for instance, for both worker safety and environmental regulations (Bowman & Kearney 1986). Maximal penalties are generally set by environmental statutes in the US. However, it is the purview of the individual states and the EPA’s regional offices to inspect for compliance and bring enforcement actions.3 Similar in spirit to environmental regulation, federal regulatory restrictions and maximum penalties are initially established in the Occupational Safety and Health Act of 1970. Enforcement is then delegated to state and regional offices as well. Despite this widespread structure, scholarship has paid little attention to these intergovernmental dynamics.4 Hence, as a matter of regulatory governance, it seems logical, perhaps essential, for any model attempting to suggest policy directives to consider the potential for differing policy and enforcement objectives.

With Ayres and Braithwaite’s (1992) characterization of responsive regulation as regulation designed to be firm-specific and flexible, rewarding those firms that go the extra step to achieve and maintain compliance and punishing aggressively those that do not, the resulting implication then is that such programs are welfare-enhancing.5 In many instances, this seems reasonable. However, in a recent survey of voluntary activities and responsive regulation promoted in the environmental regulation realm, Lyon and Maxwell (2002) identify studies suggesting that this is not always so. For instance, the authors cite DuPont’s voluntary acceleration of the phase-out of chlorofluorocarbons which may have prompted regulators to adopt more stringent future regulations, creating entry barriers within the industry, thus raising prices and reducing consumer welfare. Moreover, Maxwell and Decker (2006) show that responsive enforcement may result in too many environmental investments. In addition, Pfaff and Sanchirico (2000) point out that compliance programs designed to induce firms to self-audit may be ineffective, as firms may fear that their own discovery of noncompliance could be used against them by regulatory officials.6

Hence, the efficacy of regulation that is responsive to voluntary behavior is still subject to substantial debate. Given the potential for voluntary actions and responsive enforcement behavior to detract from socially desirable outcomes, a change to other policy variables, that is, penalties, may be necessary. Thus, there is a need to address optimal penalties in the context of responsive regulation. It is therefore beneficial to briefly revisit the determinants of optimal penalties.

2.2. Optimal penalties and responsive regulation

As has been stated, the literature on optimal penalties has its roots in Becker’s (1968) seminal paper on the economics of crime and punishment. In that paper, he suggests that if the policy goal is to achieve an optimal level of criminal deterrence at least cost, fines should be set at the highest level possible so as to (i) maximize compliance and (ii) minimize enforcement costs.

As a matter of empirical observation, however, penalties for noncompliance are not set at such maximal levels and vary from violation to violation. Many authors have thus responded by developing mathematically based models that offer rationales for why optimal penalties might be limited in nature (e.g. Polinsky & Shavell 1979, 1991; Bose 1995; Heyes 1996; Arguedas 2005). This literature is substantial and growing, and several recent and substantial survey articles exist, among them Cohen (1999), Heyes (2000), and Polinsky and Shavell (2000). Given space limitations, only a few key papers are highlighted here, largely because they feature elements similar to the model developed in this paper, notably, firm investments in compliance and the timing of when regulatory actions are taken.

Malik (1990), for instance, shows that if a potential violator invests in some detection-avoidance capital, then a welfare-maximizing penalty setter will set the optimal penalty below the potential violator’s wealth level. As regulatory parameters (penalties as well as monitoring and enforcement efforts) are set to maximize social welfare, the penalty setter will not set the penalty too high as doing so may result in too much (costly) investment in avoidance capital.

Saha and Poole (2000) present a model in which a penalty-setting regulator determines the optimal fine under two scenarios. First, the penalty-setting regulator selects the optimal fine at the start of the regulatory game, to which enforcement and compliance decisions subsequently respond. This leads to an optimal penalty level consistent with a traditional, Becker-type finding. In their second scenario, which the authors define as “endogenous penalty” setting, the fine is determined simultaneously with the enforcement and compliance decisions. They find that the optimal fine is lower in the second scenario. Their model is relevant here for at least two reasons. First, they show that the timing of when the optimal penalty is set is of critical importance. Second, they show that non-maximal optimal fines are possible when the monitoring and penalty-setting agents’ objectives differ, without appealing to regulatory errors. However, the simultaneous determination of enforcement effort and penalties for noncompliance seem inconsistent with existing regulatory structures as well as with the type of responsive enforcement policy described here.

To date, no research has explicitly linked an analysis of voluntary compliance efforts on the part of regulators and optimal policy parameters set to meet socially desirable outcomes. The following model attempts to link these two literatures.

3. Model structure

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices

The mathematical model presented here extends Maxwell and Decker’s (2006) model and involves three players: a penalty-setting regulator, an enforcement regulator, referred to as the “enforcer,” responsible for monitoring and enforcement, and a representative firm subject to regulatory constraints. In the multiple principal–agent framework mentioned, the enforcer and the penalty-setting regulator can be thought of as principals in that both their objective functions (presented here) are affected by the behavior of the firm, and the firm as the agent whose objective differs from that of the principals’. The model is predicated on several strong assumptions common in the economics literature regarding the behavior of enforcers and firms, although not universally accepted in the broader social science realm. Hence, this broader literature will be addressed where appropriate.

In conventional economic models of the type explored here a few critical, and in some instances debatable, assumptions are made. First, such models presume individuals modeled are rational and can achieve their desired objective and second, private sector motivations for compliance involve compliance cost minimization. This approach, including compliance efforts and expected noncompliance penalties, belies the potential for firms to behave altruistically, choosing socially responsible strategies simply because management deems it the correct thing to do. There is literature that explores these notions of firm behavior (see Ayres & Braithwaite 1992, pp. 19–53 for a summary of this work). In this paper, the economic convention will be followed.

Therefore, consistent with existing environmental compliance models, it is assumed that a representative firm chooses a level of environmental investment that minimizes its total expected compliance costs. These costs comprise two elements. The first, modeled as z (discussed here), comprises the capital expenditures, maintenance efforts, and so forth that a firm undertakes to improve compliance. These costs are defined generally as g(p), where p ∈ [0,1] is the probability that the firm will achieve compliance with existing environmental regulations. As a matter of modeling convention, cost functions are assumed to be convex in nature. For the function g(p), this implies the following properties. First, the first derivative, or “slope” of the g function, is positive, meaning that a higher level of p comes at a higher cost. Second, the second derivative, which measures how the slope of g “changes,” is positive, meaning that costs accelerate with higher p. This convexity of the cost function reflects the fact that additional inputs directed toward compliance, such as additional environmental capital investments, may require additional maintenance. This requires that available manpower be diverted away from other tasks toward compliance issues. Moreover, it is likely that these environmental investments crowd out other types of investment in a firm’s capital budget. The opportunity cost of these compliance efforts will likely increase as more of such efforts increase. To find a closed-form solution for the optimal penalty structure, it will be convenient to adopt, without loss of generality, a specific function form for g. As such, the simplest convex functional form that still captures all the essential characteristics delineated, including both the acquisition and implementation of z, is g(p) = p2.

The second element of the firm’s compliance cost function is the expected penalty the firm faces for noncompliance. This cost depends on the probability of noncompliance, the probability of being found in noncompliance, and the penalty levied in the event of discovered noncompliance. Mathematically, this can be represented as (1 − p)mf, where m ∈ [0,1] is the probability that the enforcement agent will discover and enforce noncompliance, and f is a strictly positive penalty assessed for noncompliance. This penalty can be thought of as embodying both a monetary component as well as cleanup costs that the firm must pay to affect remediation. This penalty structure, whereby the “polluter pays,” appears consistent with many enforcement actions that ultimately require that violating firms pay for remediation and restoration in addition to monetary payments.7

As has been discussed, compliance involves an investment in z. This variable is explicitly introduced into the model as an observable and costly investment that is difficult for the firm to either resell or use for purposes other than improved environmental compliance. This, then, commits the firm to a particular level of p that the enforcer (and the penalty-setting regulator) treats as credible, thus making the enforcer inclined to actually follow through on the reduction in monitoring. Credible commitments are important in stage games such as this one.8 Endogenizing p by making it a function of z gives more credence to the firm’s compliance decisions in that it is easier to conceptualize a costly investment in z as a credible commitment to a higher degree of compliance rather than simply adjusting the compliance probability p directly. In short, z is likely to have more “commitment power” than does p alone. Another beneficial reason for endogenizing p is that a number of the compliance programs described appear to require costly undertakings by participating firms specifically directed toward improving compliance in order to receive program benefits offered by the enforcement agency.9

It is assumed that increases in z increases p. With no universally accepted means of modeling p, in the interest of model simplicity, the following linear functional form is assumed:

  • image(1)

where inline image represents a level of costly environmental capital investment that ensures compliance.10 The parameter αcan be thought of as a technological efficiency parameter. It indicates the marginal effect an increase in z will have on the firm’s compliance probability. With this structure, the firm’s objective becomes

  • image(2)

where p = αz.

Turning attention toward regulatory enforcement, the enforcer is also assumed to minimize the cost of inspections and enforcement. Unfortunately, there is no universal agreement on the objective of regulators, much less enforcers. Cohen (1999), for instance, describes no fewer than nine different potential objectives for environmental regulators with little empirical support favoring any one over another.11Gormley (1998) identifies several inspector characteristics that hint at regulatory objectives as well. Still, there is little consensus as to the motivation for enforcers.

What can be reasonably asserted, however, is that (i) conducting inspections is costly and (ii) no inspections would likely be undertaken if there were not some sort of cost for failing to monitor and enforce. Because, in fact, inspection and enforcement activity is observed, some means of modeling these costs are necessary. Because the ultimate goal here is to develop a model that offers insight into how policy parameters (i.e. fines) should respond to observed changes in enforcement policies, it is assumed in this model that the enforcement agency suffers a cost, c, for failing on its own to discover and apprehend a noncompliant firm.12 For instance, c may represent a type of reputation cost, or “bad press” for the agency. For example, voters, community leaders, environmental interest groups, or state legislators may levy tacit penalties on enforcers for failing to meet their obligations as stewards of the environment.13 Mathematically, the enforcer runs the risk of incurring c if it fails to catch, with probability 1 − m, a violating firm, with probability 1 − p, or: (1 − m) (1 − p)c.

The cost of conducting inspections and enforcement is defined as ϒ(m), which is also assumed to be convex. That is, monitoring costs increase (meaning that ϒ(m) is positively sloped in m) at an increasing rate (the slope is increasing) with more monitoring. Again, a simple and convenient convex functional form to utilize is ϒ(m) =γm2. The parameter γ can be thought of as a monitoring and enforcement efficiency parameter. A higher γ indicates a more difficult, and thus more costly, monitoring and enforcement environment to the enforcer. For instance, conducting inspections of more complex plants, or plants located in remote locations, can result in more costly monitoring and thus a higher γ. Taken together, then, the goal of the enforcer becomes

  • image(3)

Treating the expected penalty as a pure transfer from the firm to the penalty-setting regulator, as is common practice with much of the existing literature, the goal of the penalty-setting regulator is to set f such that the sum of enforcement and investment costs is minimized.14 Formally:

  • image(4)

As will be shown, the level of the penalty chosen depends critically on the enforcer’s policies toward the firm. In what follows, the optimal penalty selected by the penalty setter under two different enforcement regimes is analyzed. The first is where the regulator develops its enforcement strategy taking the firm’s investment strategy as given and constant. The second regime treats the enforcer as being responsive to observed investments in z by reducing monitoring activity as is the case with, for example, the EPA’s National Environmental Performance Track, where the enforcer responds to the firm’s increased self-compliance efforts with reduced inspection activity.

4. Unresponsive enforcement

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices

In this section, the following stage game is presented. In stage 1, the penalty-setting regulator sets a penalty for noncompliance, f, and in stage 2 of the game, the enforcer develops his enforcement strategy while simultaneously the firm chooses its environmental investment level. Following convention, to ensure subgame perfection of the resulting equilibrium, the game is first solved via backward induction.

4.1. Stage 2: Nash equilibrium for z and m

Minimizing the firm’s expected costs of compliance (2) and the enforcer’s expected enforcement costs (3) with respect to z and m, respectively, and assuming an interior solution generates the following first-order conditions for the firm and the enforcer, respectively:

  • image((5a))
  • image((5b))

These conditions yield the following “best response” functions for the firm and enforcer:15

  • image(6)

Thus far, it should be noted that there is nothing in the parameters of the model currently defined that ensures that the probability of monitoring defined in (6) does not exceed 1. To ensure this will be the case, it is sufficient to assume that 2γ − c ≥ 0.

The behavior suggested by these functions, illustrated in Figure 1, appears to be logical. The firm invests in more z with more aggressive monitoring and enforcement, and higher penalties. Moreover, the better z is in achieving compliance (i.e. a higher α), the less z is required. The enforcer monitors and enforces more when the cost it faces for failure to catch a violator is higher and the probability of noncompliance is greater. Moreover, the enforcement intensity declines with higher enforcement costs.

image

Figure 1. Best-response functions and equilibrium outcomes under unresponsive and responsive regulations.

Download figure to PowerPoint

Solving conditions (5) simultaneously yields a system of stage 2 Nash equilibrium levels of z.16 These equilibriums are labeled zur and mur and are illustrated in Figure 1:

  • image(7)

4.2. Stage 1: Optimal penalty decision

To find the optimal penalty to assess the firm for noncompliance, conditions (7) are substituted into the penalty setter’s objective function, (4). Doing so and solving for f yields

  • image(8)

(see Appendix I for this derivation). As penalties are not negative, it is necessary to further assume that 2 − c > 0.

While condition (8) provides the basis for comparison with the optimal penalty under responsive enforcement, it is interesting to note before continuing the following condition:

  • image((8a))

This is a condition very much in the spirit of Becker (1968). That is, the more costly it is to undertake monitoring and enforcement, the higher the penalty.

5. Responsive enforcement

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices

This section analyzes the model assuming responsive enforcement. In this game, the penalty setter sets an optimal fine for noncompliance consistent with its objective (4) in stage 1. In stage 2, however, the firm chooses its environmental investment level contingent upon the enforcer’s offer to subsequently respond to this investment when considering its monitoring and enforcement strategy. Hence, the enforcer moves in stage 3 of the game by setting an optimal enforcement strategy, m, consistent with its objective (3). Again, the game is solved via backward induction.

5.1. Stage 3: The enforcement decision, m

As was the case under unresponsive enforcement, minimizing (4) with respect to m yields the following condition (note that this is identical to inline image in conditions (6)):

  • image(9)

Again, assuming 2γ − c ≥ 0 ensures the probability of monitoring will not exceed 1. From (9) the following expression is derived:

  • image(10)

which highlights the regulator’s willingness to respond favorably to higher environmental investment levels with reduced monitoring and enforcement. It seems reasonable then, that the firm may find it beneficial to increase its investment in z to reap the cost advantages associated with reduced enforcement scrutiny.

5.2. Stage 2: The investment decision, z

Substituting (9) directly into the firm’s objective function (2) and minimizing the resulting cost function with respect to z yields the following optimal condition:

  • image(11)

This mirrors Maxwell and Decker’s (2006) condition that responsive enforcement induces additional environmental investment. When comparing (11) with condition (5b), it can be seen that marginal benefits to investment in z now has two components. The first component is the “internal” benefit the firm realizes, which is associated with improving the probability of compliance and thereby avoiding the penalty. The second component is a “strategic” benefit. Note that as additional environmental investments on the part of the firm reduces the incentive for the enforcer to monitor as frequently, that is, inline image, then the expression inline image is positive. Hence, this strategic benefit to making additional environmental capital investments (i.e. additional z) comes from the reduction in monitoring and enforcement intensity directed toward the firm. Hence, the firm has an incentive to increase z. Indeed, this can be showed explicitly by substituting (9) and (10) into (11) and solving for z, yielding a stage 2 Nash equilibrium condition, z*:

  • image(12)

Comparing (12) with the expression for zur in (7), one finds that there is indeed more investment in z when enforcement is responsive. To see this more concretely, notice that

  • image(13)

This is illustrated in Figure 1.17 Clearly, then, environmental investments increase when enforcement is responsive.

Notice the consequence of this additional investment for the enforcer. Substituting (12) into (9) yields the following optimal level of monitoring:

  • image(14)

Comparing (14) with the corresponding condition (7), it can be shown that this level of monitoring is less than the level of monitoring, mur, under unresponsive regulation, as illustrated in Figure 1. Hence, responsive enforcement, not too surprisingly, results in a savings of monitoring activity for the enforcer.

5.3. Stage 1: The penalty decision, f

What, then, should be the response of the penalty setter to responsive enforcement? Should the penalties for noncompliance be adjusted and if so, in which direction and to what degree?

In this stage of the game, the penalty setter sets a penalty taking into consideration the stage 2 behavior of the firm, and the stage 3 behavior of the enforcer. Incorporating this information into the penalty setter’s objective (4) and solving for the new optimal fine under responsive enforcement, f*yields

  • image(15)

(see Appendix II). Comparing this expression with the optimal fine, fur, defined in (8) yields

  • image(16)

or, stated differently, f* is half the size of fur. Hence, if enforcers shift to a responsive style of enforcement, that is, establishing a monitoring and enforcement strategy contingent upon the firm’s investment decision, the penalty setter should reduce the optimal penalties for noncompliance. This is illustrated in Figure 2. The rationale for this comes from the fact that the penalty setter is concerned about social costs, which includes the cost of the investment, z. As the enforcer offers to reduce his monitoring and enforcement intensity in response to environmental investments, the firm recognizes that this reduces its own expected fine. However, if the enforcement response is, in a sense, too large, this could prompt an overspending in environmental investments (Maxwell & Decker 2006). If the optimal fine were kept at fur under the unresponsive enforcement regime, then the introduction of responsive enforcement will result in too much z, thereby increasing social cost (again, illustrated in Fig. 2). To counter the stimulus to investment that responsive enforcement causes, the penalty setter will have to reduce the fine in the initial stage of the game, thereby reducing the incentive for the firm to invest in stage 2.18

image

Figure 2. Expected social costs under unresponsive and responsive regulations.

Download figure to PowerPoint

If this mathematical outcome is considered in light of Becker (1968), the adoption of responsive enforcement in a regulatory structure where the penalty-setting regulator responsible for the design of the regulation does not necessarily have the same objective as the agent responsible for enforcement of the regulation may suggest another reason for optimal penalties not to be maximal. From Becker’s model, recall that a regulator motivated to minimize the social cost of enforcement would be inclined to maximize the penalty (perhaps even setting f =+∞) since enforcement is costly as potential violators respond to the expected fine (mf) when making their compliance decisions. In this model, a switch to responsive enforcement promotes more limited fines.

It is worth pointing out that, while this downward adjustment to the optimal fine prompted by responsive enforcement is relatively intuitive given the structural conditions of the model, from a broader perspective, the condition itself seems to suggest a rather nontraditional regulatory direction. As a matter of more traditional analysis, the promotion of compliance, whatever the means, is generally considered a good thing for society. Yet, responsive enforcement behavior that induces additional environmental investments may not be a good thing for society as a whole after all. Indeed, a policy correction in the form of a lower penalty for noncompliance may be required.

As a corollary to this model outcome, from (16) it interesting to note that

  • image((16a))

which mirrors (8a) in that, from a social cost perspective, higher monitoring and enforcement costs can be offset by higher penalties. However, upon closer inspection, the condition that obtains here may not be immediately obvious. Note that as monitoring costs increase, the upward adjustment to the optimal penalty under responsive enforcement is less than the upward adjustment to the penalty under unresponsive enforcement. This can be seen readily by comparing:

  • image((16b))

This might seem counterintuitive at first, as responsive regulation results in lower monitoring and enforcement under responsive enforcement, thus requiring a higher counterbalancing penalty. However, in this model, the penalty is being adjusted to counter the social-cost effects of too much environmental investment. In light of this observation, if difficulties in monitoring increase (i.e. γ increases) under responsive enforcement, then both the optimal fine under responsive enforcement, f*, will be increased and the optimal level of monitoring under responsive enforcement, m*, will be scaled back. However, the increase in f* is mitigated (relative to unresponsive enforcement) so as to prevent too much environmental investment from occurring.

6. Conclusion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices

This paper considers the level of, and changes in, optimal noncompliance penalties (i) where the regulator responsible for setting policy parameters, such as fines, is different from (and thus may have a different objective from) the regulator responsible for enforcing existing regulations and (ii) where enforcement behavior changes from one in which enforcers are unresponsive to overtures on the part of firms to increase compliance to one in which enforcers are responsive to such overtures. The model illustrates two key outcomes that arise when enforcers “switch” from unresponsive to responsive enforcement. First, the firm has an incentive to increase spending on environmental capital investments. Second, this investment may represent too much additional spending from a social cost perspective, requiring a reduction in optimal penalties for noncompliance to reduce the incentive to invest.

Both points are noteworthy. It has become fashionable to attribute overtures on the part of firms to improve their environmental records (such as investing in pollution control technologies or other environmental investments) as purely voluntary, driven by altruistic motives. While this is certainly reasonable in some cases, in others such investments may be ultimately driven by desires to maximize profits. The model presented here emphasizes that enforcement policies may indeed induce additional environmental investments by firms motivated to maximize profit (minimize cost) that might mistakenly be characterized as purely voluntary.19 If true, then, it might be the case that enforcement policies induce too much “green” and costly investment, and thus there may be a need on the part of other policy officials to curb some of this.

The model may be fruitfully modified in several ways. First, the model assumes that each regulator is able to observe (predict) the actions of the other as well as the investment and compliance decisions of the firm. This may not be viable in many real-world instances where uncertainty abounds and outcomes are not so perfectly predictable. Building uncertainty into the model (e.g. Pfaff & Sanchirico 2000) and exploring optimal adjustments to penalties when responsive enforcement is offered may have serious social welfare implications. Second, the assumption that players in this game are rational and will achieve their desired objective may be worth further theoretical investigation. Perhaps exploring “trembling-hand” game-theoretic models may offer additional insights (see Rasmusen 2001 for further discussion). Finally, it may be of interest, following Bose (1995), to introduce regulatory errors into the mix. This may have significant implications for the direction of optimal penalties under responsive enforcement. These considerations are left for future research.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices

The author wishes to thank John Braithwaite, Cary Coglianese, David Levi-Faur, John W. Maxwell, and three anonymous referees for their helpful comments on this paper. The author is responsible for any and all remaining errors.

Notes
  • 1

    This national program, launched in 2000, is based on New England EPA’s StarTrack program. See Maxwell and Decker (2006) for additional details on StarTrack. For further information on the National Environmental Performance Track program, visit http://www.epa.gov/region1/pr/2000/071000.html.

  • 2

    It is the case that for firms to realize these benefits, they must document completely their compliance efforts. Firms must establish a systematic management plan to achieve and maintain compliance, annually prepare and publicize a comprehensive compliance report, and annually engage in third-party verification and review to certify compliance efforts. Effective management along these lines is likely to be quite costly for firms seeking enforcement relief under these audit policy stipulations. That said, however, there is evidence that the EPA does follow through on its claim toward reduced enforcement. For instance, in 1997, the EPA’s Region V office in Chicago, Illinois, encouraged a number of small steel mills to participate in this new audit policy plan. According to EPA reports, approximately half of those so encouraged undertook the necessary efforts to improve compliance. Those that did not choose to participate were inspected with much greater frequency.

  • 3

    For instance, under the US Clean Air Act, criminal penalties for egregious and repeated violations of clean air restrictions can reach $250,000 per day per violation and up to 5 years in jail (Brownell 1995, p. 132). However, enforcement of environmental regulations, through inspections (i.e. monitoring) and injunctions (i.e. enforcement actions), is largely conducted through the enforcement arm of the US EPA’s Office of Environmental Compliance and Assurance (OECA), regional EPA offices, and, perhaps most importantly, state environmental protection agencies. This seems particularly true with respect to inspections. According to OECA data, between 1991 and 2000, there were 318,744 inspections conducted to check for compliance with three major US environmental regulations: the Clean Air Act, the Clean Water Act, and the Resource Conservation and Recovery Act. Fully 95.5% of these inspections were conducted by state environmental authorities.

  • 4

    That said, there is a rich literature on the determinants of state-level governmental behavior and their influence on regulatory outcomes. For a survey of this literature, see Gerber and Teske (2000).

  • 5

    To avoid any confusion, it should be noted that Ayres and Braithwaite’s (1992) definition of responsive regulation is much broader than the type of responsive enforcement discussed earlier in this paper and that characterizes the model developed here. Responsive regulation can in effect involve a whole host of regulatory responses to firm behavior, such as statute modification, regulatory delay, penalty leniency, or even direct subsidization. The focus in this paper is specifically on enforcement whereby enforcement agencies respond to a firm’s compliance behavior with adjustments to site inspection frequency.

  • 6

    While it is the case that much of the empirical work does not necessarily support the potential benefits of voluntary programs that many theories would predict, there are a couple of reasons for this. The first is measurement difficulties. It may be, for instance, that in the DuPont case, the company’s voluntary action did deter entry. However, the isolated effect of reducing chlorofluorocarbons may have dominated the monopoly effect. At best, both effects are difficult to precisely measure. The second reason for the empirical work not matching theory could be that the existing theory is not refined well enough to offer accurate testable hypotheses. That said, the ultimate goal of this paper is not to refute these empirical studies but to refine some existing theory; rather than attempt to test it empirically, offer some suggestions that could, as part of a larger analysis, offer insights as to potential policy directions.

  • 7

    For instance, in 1997 Pacific Gas & Electric was fined substantially for violations of the US Clean Water Act. In addition, the company was required to pay an additional $6 million in environmental restoration (EPA 1998).

  • 8

    See Dixit and Nalebuff (1991, pp. 142–167) for a discussion of credible commitments and their importance in stage games.

  • 9

    For example, participants in the EPA’s National Environmental Performance Track program are required to show “continuous improvement.” This involves, among other things, on-site improvements to reduce emissions and other wastes in production, distribution improvements that reduce plastic packaging, and an overall attempt to limit energy use in the production and distribution processes (see http://www.epa.gov/performancetrack/program/index.htm for additional information on this program). To make these improvements and modifications to production and distribution will in all likelihood require costly investments that probably would not have been undertaken without participation in the program. It should be pointed out, however, that mathematically the model presented here could have been constructed treating p as a model parameter directly. Qualitatively, all the resulting model derivations presented here would be the same as what is presented here. However, based on the discussion here, p was left as a function of z.

  • 10

    This simplified function assumes that increases in z increase p by a constant proportion α. While this is a significant simplification, econometric analysis was conducted that lends empirical support to this functional form. These results are available from the author upon request.

  • 11

    Some authors suggest that benefits are derived from reducing the expected cost to the environment from noncompliance, or from forcing the noncompliant firm to remediate the environmental damage created (Maxwell & Decker 2006). Bose (1995) postulates that the benefits from monitoring and enforcement are the penalty revenues extracted from violating firms.

  • 12

    Obviously, this would require that noncompliance was discovered by some other group, such as a group of private citizens suffering ill-effects of increased exposure to toxic pollutants, or environmental interest groups such as Greenpeace or other “watchdog” groups. Such private enforcement activities, however, are not explicitly considered in this model.

  • 13

    This line of reasoning follows Becker’s (1983)“interest group pressure” model where regulators attempt to maximize net positive political support from relevant stakeholders, such as voters, elected officials, industry, etc. by engaging in regulatory activities.

  • 14

    Notice that, in effect, the policy-setting regulator benefits from the penalty being levied against the firm as these funds are then used to rectify any damages as a consequence of noncompliance. This is reasonable in that these penalties often support state and federal general funds the can be drawn upon to cover cleanup operations and natural resource reclamation projects (Kleit et al. 1998).

  • 15

    “Best response” functions illustrate how the optimal values of a player’s strategic choice variable vary with different values of the rival player’s strategic choice variable. For instance, from condition (6), the firm’s optimal value of z increases as the regulator’s monitoring effort (m) increases.

  • 16

    Formally, the Nash equilibrium is a condition describing a set of strategies selected by each player of a game in which no player can improve his/her payoff by unilaterally changing his/her strategy, given the other players’ strategic choice. For instance, the firm’s selection of zur represents its best choice from all other possible values z, given the enforcer’s choice of mur.

  • 17

    Notice from the figure that there is no “best response” function to draw for z here as the optimal level of m that the enforcer will select in stage 3 of this game (i.e. condition [9]) is already being taken into consideration by the firm when selecting z in stage 2.

  • 18

    As a point of clarification, note from Figure 2 that the adjustment of the fine downward when responsive enforcement is adopted by the enforcer is required to bring expected minimum social costs back to their original level achieved under unresponsive enforcement. Graphically then, given the functional forms assumed in this model, a switch from unresponsive to responsive enforcement causes a parallel shift leftward in the expected social cost function. As a matter of future theoretical research, it may be of interest to investigate what kind of behavior would cause the social cost function to not only shift leftward but also shift downward so that a lower minimum social cost level is achievable.

  • 19

    There is some empirical evidence that does indeed suggest that enforcers do respond to firm reductions in toxic pollutant releases with reduced inspection activity (e.g. Decker 2005). It seems reasonable, then, to presume that firms’ incentives to improve their environmental records are in part driven by this desire to influence enforcement behavior.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices
  • Arguedas C (2005) Bargaining in Environmental Regulation Revisited. Journal of Environmental Economics and Management 50, 422433.
  • Ayres I, Braithwaite J (1992) Responsive Regulation. Oxford University Press, New York.
  • Becker RG (1968) The Economics of Crime and Punishment. Journal of Political Economy 76, 169217.
  • Becker RG (1983) A Theory of Competition Among Pressure Groups for Political Influence. Quarterly Journal of Economics 98, 371400.
  • Bose P (1995) Regulatory Errors, Optimal Fines and the Level of Compliance. Journal of Public Economics 56, 475484.
  • Bowman A, Kearney R (1986) The Resurgence of the States. Prentice-Hall, Englewood Cliffs, NJ.
  • Brownell FW (1995) Clean Air Act. In: SullivanTFP, AndersonRC (eds) Environmental Law Handbook, 13th edn, pp. 101134. Government Institutes, Rockville, MD.
  • Cohen MA (1999) Monitoring and Enforcement of Environmental Policy. In: FolmerH, TeitenbergTH (eds) International Yearbook of Environmental and Resource Economics, Vol. 3, pp. 44106. Edward Elger, Lyme, NH.
  • Decker CS (2005) Do Regulators Respond to Voluntary Pollution Control Efforts? A Count Data Analysis. Contemporary Economic Policy 23, 180194.
  • Dixit AK, Nalebuff BJ (1991) Thinking Strategically: The Competitive Edge in Business, Politics, and Everyday Life. W.W. Norton & Company, New York.
  • EPA (Environmental Protection Agency) (1998) Enforcement and Compliance Assurance Accomplishments Report. FY 1997, Jul EPA 300-R-98–003. Office of Enforcement and Compliance Assurance, EPA.
  • Gerber B, Teske P (2000) Regulatory Policy-Making in the American States: A Review of Theories and Evidence. Political Research Quarterly 53, 849886.
  • Gormley WT (1998) Regulatory Enforcement Styles. Political Research Quarterly 51, 363383.
  • Hay BL, Stavins RN, Vietor RHK (2005) Environmental Protection and the Social Responsibility of Firms: Perspectives from Law, Economics, and Business. Resources for the Future Press, Washington, DC.
  • Heyes AG (1996) Cutting Environmental Penalties to Protect the Environment. Journal of Public Economics 60, 251265.
  • Heyes AG (2000) Implementing Environmental Regulation: Enforcement and Compliance. Journal of Regulatory Economics 17, 107119.
  • Kleit AN, Pierce MA, Hill CR (1998) Environmental Protection, Agency Motivations, and Rent Extraction: The Regulation of Water Pollution in Louisiana. Journal of Regulatory Economics 31, 121137.
  • Lyon T, Maxwell JW (2002) Voluntary Agreements in Environmental Regulation. In: FranziniM, NicitaA (eds) Economics, Institutions, and Environmental Policy, pp. 75120. Ashgate Publishing, Brookfield, VT.
  • Malik AS (1990) Avoidance, Screening, and Optimum Enforcement. RAND Journal of Economics 21, 341353.
  • Maxwell JW, Decker CS (2006) Voluntary Environmental Investment and Responsive Regulation. Environmental and Resource Economics 33, 429435.
  • Pfaff ASP, Sanchirico CW (2000) Environmental Self-Auditing: Setting the Proper Incentives for Discovery and Correction of Environmental Harm. Journal of Law, Economics, and Organization 16, 189208.
  • Polinsky AM, Shavell S (1979) The Optimal Tradeoff Between the Probability and Magnitude of Fines. American Economic Review 81, 618621.
  • Polinsky AM, Shavell S (1991) A Note on Optimal Fines When Wealth Varies Among Individuals. American Economic Review 69, 880891.
  • Polinsky AM, Shavell S (2000) The Economic Theory of Public Enforcement of Law. Journal of Economic Literature 38, 4576.
  • Prakash A, Potoski M (2006) The Voluntary Environmentalists. Cambridge University Press, New York.
  • Rasmusen E (2001) Games and Information, 3rd edn. Blackwell Publishing, Malden, MA.
  • Saha A, Poole G (2000) The Economics of Crime and Punishment: An Analysis of Optimal Penalty. Economics Letters 68, 191196.
  • Vogel D (2005) The Market for Virtue: The Potential and Limits of Corporate Social Responsibility. Brookings Institution Press, Washington, DC.

Appendices

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Voluntary compliance programs and optimal penalties
  5. 3. Model structure
  6. 4. Unresponsive enforcement
  7. 5. Responsive enforcement
  8. 6. Conclusion
  9. Acknowledgments
  10. References
  11. Appendices

Appendix I

The optimal fine under unresponsive regulation

To find the optimal fine under unresponsive enforcement, functions (7) are substituted into the penalty setter’s objective function (4) as follows:

  • image((A1))

Minimizing with respect to f yields the following implicit first-order condition:

  • image((A2))

The term in the first set of square brackets is zero by condition (5b). As ∂zur/∂f ≠ 0, the penalty setter sets f such that the condition in the second set of square brackets in (A2) is 0. Substituting conditions (7) into this bracketed term and solving for f yields the optimal penalty condition (8).

Appendix II

The optimal fine under responsive regulation

To find the optimal fine under responsive enforcement, functions (12) and (14) are substituted into the penalty setter’s objective function (4) as follows:

  • image((B1))

Minimizing with respect to f yields the following implicit first-order condition:

  • image((B2))

As the expression in the first set of square brackets in (B2) is again zero by (5b), the optimal fine will be set such that the expression in the second set of square brackets is zero. Thus, solving for f yields the optimal penalty condition (15).