Chapter 3. Methods of Risk Regulation
An approach to risk regulation based on pragmatism seeks to accommodate widely shared but potentially conflicting values by striving to achieve the maximum level of protection consistent with reasonable cost. Despite the advantages of this system, there are bound to be particular situations that yield regulation with which particular regulated entities cannot afford to comply, regulation whose application in a specific instance will impose costs of a magnitude that appears to be excessive, regulation whose application to a particular entity or economic sector may result in unacceptable dislocation, or regulation that has unintended adverse social consequences. This kind of regulatory output is inevitable because the bounded rationality problem will preclude risk regulatory agencies from fully understanding all aspects of a problem before they take steps to address it. Risk regulation is necessarily based on the application of heuristics, whether they take the form of cost-benefit analysis or the more pragmatic approach to regulation we prefer. In short, even a pragmatic approach can yield regulation that reflects an unsuccessful effort to accommodate economic and noneconomic considerations.
One way to try to minimize instances of “excessive” or wrong-headed regulation would be to adjust the general statutory standards. One could weaken the standards, for example, so that few if any members of the regulated community lack the economic capacity to comply with them, or one could mandate more rigorous attention to and comparison of predicted regulatory costs and benefits in a futile attempt to overcome bounded rationality. The first option threatens to reduce the levels of protection against risk below those regarded as effective by most citizens. The second is likely to yield the kind of “paralysis by analysis”, which, in turn, will also decrease the output of risk regulatory protection to levels that may be deemed unacceptable. Moreover, the very bounded rationality responsible for unintended regulatory consequences is likely to prevent the newly elaborate process from yielding more “rational” regulation.
We explore two alternative and, we think, preferable mechanisms for adjusting regulatory programs to eliminate unintended or counterproductive results. In his book on eco-pragmatism, Daniel Farber argues that experience with risk regulation during the past thirty years has demonstrated “the centrality of learning to the enterprise of environmental protection.” According to Farber, this experience supports efforts to “raise our regulatory IQ,” or, more colorfully, to “teach the elephant to waltz.” We believe that the two mechanisms we will discuss in this chapter already make the current system of risk regulation to a considerable degree “responsive to additional information.” The first mechanism is to permit regulated entities to choose the method by which they comply with risk regulations. The second is to make greater use of incremental regulation in which regulators adjust general regulatory commitments in light of the specific circumstances of some regulated entities or the availability of new information. These two avenues for mitigating the unintended adverse economic and social consequences of risk regulation reflect a pragmatic approach to the implementation of that body of regulation.
Each of the flexibility devices we describe in this chapter has been tried, some with better effects than others. All are subject to abuse, but we believe that, on balance, they represent a more pragmatic method of accommodating divergent factors such as efficiency and fairness or other nonutilitarian values than does an attempt to incorporate in a rigid fashion all relevant factors into the risk reduction standard itself. To address the potential for abuse, we support the adoption of procedures that, on the one hand, enable agencies to dispense incremental relief in a relatively streamlined fashion but that, on the other hand, enhance accountability by requiring agencies to provide opportunities for public input and disclosing when and why they have afforded such relief. The adoption of such procedures is consistent with pragmatism’s preference for methods of accountability that promote public participation.
A regulatory process that includes and relies upon the flexibility devices discussed in this chapter may enable an agency to avoid committing so many resources to perfecting regulation at the front end of the process that it is precluded from effectively assessing the practical effect of regulations once they have gone into effect. Critics of current risk regulation approaches claim that these approaches often result in misallocation of regulatory expenditures when agencies pursue relatively small risks while leaving larger risks unaddressed. But a regulatory approach based on a quest for comprehensive rationality may have an additional unintended adverse effect if it diverts agencies from the kind of planning that is capable of avoiding such misallocations. At the final of this chapter therefore addresses how a pragmatic system of risk regulation can use planning to minimize misallocation of regulatory expenditures.
One method of reducing regulatory costs is to choose the least-cost method of achieving a regulatory goal. This goal is consistent with the risk reduction framework that Congress adopted, although critics of risk regulation sometimes claim otherwise. Risk regulation generally employs “performance” standards that allow regulated entities to choose the method by which they comply with the risk reduction the government has ordered. In other circumstances, Congress has authorized the use of “incentive-based” instruments, which give regulated entities even greater flexibility to select cost-effective compliance options. Some critics of risk regulation are so enamored with the use of “incentive-based” methods that they would create a presumption in favor of their use. Although incentive-based methods may be a less expensive way to regulate, the evidence does not support the conclusion that they are usually a better way.
This section starts with the failure of the critics of risk regulation to recognize that existing regulatory goals are consistent with cost-effective methods of regulation. We then consider the role of performance standards and incentive-based regulatory methods, and the potential of each to lower regulatory costs and achieve other regulatory values.
The critics have charged that command-and-control regulation of the kind that has dominated federal risk regulation for the last thirty years is incapable of producing efficient regulation because, among other things, it bypasses opportunities for regulated entities to select the most cost-effective means of regulation available. According to the critics, command-and-control regulation tends to take the form of uniform standards applicable across the board to entire classes of risk-creating activities. The difficulty with such uniform controls is that they fail to recognize that the cost of control may vary significantly within a class of risk-creating activities. One polluter may have the capability of reducing its discharges at a fraction of the cost of another polluter discharging the same material. A rational approach to controlling pollution would be to impose stringent controls on the polluter whose costs of control are low, while imposing lenient controls on those with high control costs. In that manner, the desirable level of pollution control can be achieved at the least cost.
The complaint that traditional risk regulation forfeits opportunities for achieving cost-effective solutions to environmental problems is often misleading, because those making it generally fail to distinguish carefully between the adoption of environmental standards and the selection of techniques (or tools or instruments) for achieving those standards. The two inquiries need not be – and often are not – governed by the same analytical frameworks.
The formulation of environmental policy requires that policy-makers address two “central questions: (1) what is the desired level of environmental protection?; and (2) what policy instruments should be used to achieve this level of protection?” The literature on environmental policy, however, does not always carefully distinguish between these two issues. This blurring of conceptually distinct questions tends to mask the possibility that an environmental policy goal might be set using criteria other than economic efficiency, but that, once the goal has been established, regulators may select the policy instrument likely to achieve that goal at the least cost. There is no reason why, for example, a regulatory system whose goal is set without reference to cost – such as an ambient quality-based regulatory scheme – could not be coupled with incentive-based policy instruments, such as marketable permits or pollution taxes. Indeed, the Environmental Protection Agency (EPA) has long taken the position that although cost considerations play no legitimate role in the adoption of the Clean Air Act’s national ambient air quality standards, those considerations are critically important when policy-makers are designing appropriate means of achieving the standards.
The potential for incentive-based techniques to promote cost-effective means of compliance with regulatory standards is widely acknowledged, as well it should be. If incentive-based techniques such as emissions trading can achieve regulatory standards at lower costs than other regulatory instruments, it would be irrational not at least to consider their use. Allowing regulated firms to resort to emissions trading as a means of complying with their regulatory obligations undoubtedly creates the possibility that those firms will manipulate the system by engaging in “paper trades” that do little if anything to improve environmental quality. The fact that regulated entities may abuse the discretion to use emissions trading as a means of regulatory compliance is not a sufficient basis for condemning the entire technique, however. Rather, the criticism argues in favor of more careful creation of prerequisites for trading and more rigorous agency oversight.
Environmentalists also have objected to the symbolism of emission trading. Some view public resources such as clean air “as a basic inalienable right which is not for sale at any price. ... Allowing firms to trade emission rights sends a message that decisions about tradeoffs between economics and environmental quality can be left to the polluters.” In light of the intrinsic value of the environment, this objection is not easily answered. Pragmatism, however, eschews essentialism, which argues against investing resource protection with the status of an inalienable right. Further, the negative symbolic aspects of an endorsement of trading may be an acceptable tradeoff for the efficiency gains that are available in some situations through resort to the practice. А pragmatic approach to risk regulation attempts the difficult task of finding solutions that accommodate conflicting values to the greatest extent possible. In particular, pragmatism favors mechanisms that reduce the cost of regulation because such approaches reduce opposition to nonutilitarian standards of risk reduction that pragmatism strongly supports.
Even among those, including some of the major national environmental groups, which endorse incentive-based techniques such as emissions trading as a valuable method of regulatory compliance, the appropriate mix of more traditional regulatory devices (such as performance standards) and incentive-based techniques nevertheless remains a matter of ongoing debate. Cass Sunstein has recommended the adoption of economic incentives “as a presumptive substitute for command-and-control” regulation. Under this modified approach to instrument selection, EPA would be required, “wherever feasible, to use economic incentives rather than a ‘command-and control’ approach.”
Although we agree that incentive-based instruments have a useful role to play in the implementation of risk regulation standards, we cannot agree that such a presumptive approach is desirable or consistent with pragmatic regulation. A pragmatist would argue that, before displacing traditional regulatory instruments with incentive-based techniques on a much larger scale, the proponents of such a change should bear the burden of demonstrating that practical experience with incentive-based instruments justifies such a shift. Although Professor Sunstein is convinced that “significant cost savings can be achieved by using more flexible, market-oriented instruments, such as tradable pollution permits rather than uniform national requirements,” and that such cost savings support adoption of a presumption in favor of requiring a market-based approach, empirical support for his belief simply does not exist. According to Daniel Cole and Peter Grossman, “the existing ‘empirical’ studies do not demonstrate either that command-and-control regulations are inherently inefficient or that they are invariably less efficient than market-based alternatives.” Indeed, they assert further that in certain cases, command-and-control regulation can be and has been more efficient than alternative, incentive-based approaches. Similarly, Daniel Farber has warned:
Although these incentive systems are intriguing, we should not be too confident about translating their theoretical advantages into practice. There are good reasons for caution. Real-world implementation may raise significant enforcement problems, create barriers to entry by new firms, unduly favor some firms in the initial allocation of permits, or conflict with other goals like equity. Moreover, the actual legal entitlements are likely to differ considerably from the elegant theoretical models, if only for political reasons. ... The only way to see if these market solutions will work is to try them provisionally and carefully monitor the results.
Given the difficulty of ascertaining the costs of control a neutral baseline, in which policy-makers choose from available traditional performance standards and incentive-based instruments on a case-by-case basis with no preconceived bias, is preferable to any presumption. A second reason for rejecting a presumption is that the relative desirability of market incentive instruments is likely to be context-specific. As Professor Kenneth Richards recognizes, the “optimal choice of policy instrument to implement a particular pollution abatement goal depends on the nature of the pollutant, the kind of harm the pollutant causes, the available control technologies, the number and type of polluting entities, and the type of market failure.” Similarly, Daniel Cole and Peter Grossman have argued:
“There are institutional settings in which markets are not only less efficient than command-and-control regulations but are in fact completely ineffective in reducing pollution. In the real world, the relative efficiency with which a particular regulatory regime maximizes a social welfare function depends on institutional and technological circumstances. ... There may be important practical reasons for favouring one among alternative planning instruments. These reasons might involve ideological, political, legal, social, historical, administrative, motivational, informational, monitoring, enforcing, or other considerations”. Moreover, these considerations are not static, because “efficiencies can shift in response to institutional and technological evolution.”
As Professor Richards has pointed out, the total cost of applying a particular tool for achieving a specific risk regulation standard is composed of the sum of different kinds of costs, including “production” costs (such as capital, training, and operation and maintenance costs), “implementation costs” (such as measurement and enforcement costs), and “public finance” impacts. Although incentive-based regulation can have lower “production” costs than traditional regulation, higher “implementation” costs can offset any cost advantage that incentive-based regulation might have when only “production” costs are considered.
The process of examining the degree to which emissions trading and marketable permits are capable of achieving risk reduction goals more cost effectively than command-and-control instruments has begun. Further experience with the use of such incentive-based instruments may in the future reveal that they are more cost-effective in most cases, or that they are more cost-effective in particular contexts. A pragmatic approach to risk regulation would encourage the proponents of incentive-based instruments to continue trying to convince the relevant “critical communities” that such is the case. To date, the available evidence does not seem to warrant a wholesale shift to these kinds of instruments of the nature that would result from adoption of a presumption in favor of incentive-based techniques. The evidence is weaker still for instruments such as pollution taxes, which have not been applied to any significant degree to federal risk reduction programs.
A related form of regulatory adjustment mechanism—the waiver, exception, or exemption—subjects a regulated entity to more lenient treatment or exempts it completely from regulatory obligations instead of extending the time for compliance. This “option can ... be viewed as a recognition that formal rules are unlikely to capture the infinite varieties of empirical reality and that increased flexibility in the rulemaking process is necessary.” The issuance of a waiver or an exception is typically based on equitable grounds or on the grounds that insistence on compliance with a generic rule by the particular entity involved would be inconsistent with underlying regulatory objectives. Had the agency adopting a rule been aware of the unique circumstances that make its application to a particular entity unfair or counterproductive, it presumably would (or could) have carved out an exception in the rule itself. Having failed to do so, the issuance of a waiver or an exception simply represents an alternative procedural mechanism for accomplishing the same result—essentially, promulgation of a rule applicable to a category of one entity. Several different kinds of individualized adjustment mechanisms have been made available to entities subject to risk regulation statutes. First, “hardship exceptions” represent adjustments based on the adverse economic impact of regulation on an individual firm or on the absence of available technology to comply with regulatory controls. According to Alfred Aman: “Though Congress may have decided that industries should internalize certain environmental costs ... Congress’s broad legislative objectives do not automatically outweigh the continued survival of regulated firms. The regulatory cures for environmental pollution . . . should not necessarily cripple the industries to which they apply. Stability and preservation of economic order go hand in hand with environmental or economic reforms. The regulatory preference for individual firm survival does not necessarily mean that shutdowns must always be avoided, but such extreme consequences should be the result of a considered process, not the unintended or unconscious fallout of an overbroad statute or rule”.
Many of the federal risk regulation statutes provide explicitly for hardship exceptions. In other cases, the courts have interpreted general statutory standards in a way that minimized economic disruption which would result from strict application of the standard. State agencies operating under authority delegated to them by federal statutes have created analogous individual adjustment mechanisms.
Adjustments also may be based on a demonstration that a particular regulated activity will not create the kinds of risks the regulatory scheme was designed to control. The Clean Water Act, for example, allows EPA to modify
the technology-based effluent limitations applicable to point sources discharging certain kinds of pollutants if an individual point source can show that less onerous limitations will neither result in the imposition of additional effluent restrictions on other sources nor interfere with attainment or maintenance of fishable/swimmable water quality. The Occupational Safety and Health Act allows the Occupational Safety and Health Administration (OSHA) to issue variances from occupational safety or health standards to an employer able to demonstrate that the alternative practices it proposes to use will provide places of employment as safe and healthful as those that would have resulted through compliance with the standard. Under the Clean Air Act, EPA may waive the requirement that states containing areas not in attainment with the national ambient air quality standard for carbon monoxide require the sale of gasoline with a higher than normal oxygen content. To qualify for such a waiver, the state must show that mobile sources of carbon monoxide do not contribute significantly to excessive concentrations of that pollutant. And the Resource Conservation and Recovery Act establishes a petition process pursuant to which generators may procure delisting of hazardous wastes at particular facilities.
Third, individual adjustments may be available on equitable grounds. A regulated entity may qualify for one of these “fairness exceptions” if it can show that the costs it would incur if it complied with the regulations would be disproportionately higher than those incurred by similarly situated firms, that the regulation inadvertently punishes the entity for good-faith activities, or that the costs imposed on the firm are not justified by the resulting social benefits. “In other words, though the costs involved do not jeopardize the economic viability of the firm, and therefore do not justify an economic-hardship exception, the rule as applied to the petitioner is nevertheless unreasonable.”
Perhaps the best known example of a fairness exception is the fundamentally different factor (FDF) variance from the Clean Water Act’s technology-based effluent limitations. Although the statute did not originally authorize EPA to issue such FDF variances, the Supreme Court conditioned EPA’s authority to regulate point sources through category wide regulations rather than on a point source-by-point source basis on the availability of individual variances from regulatory standards. EPA fleshed out the conditions under which these variances would be available, and the Supreme Court interpreted the statute to authorize their issuance even for toxic pollutants, which are subject to the act’s most stringent controls. In 1987, Congress endorsed these developments by codifying the FDF variance mechanism for a point source able to convince EPA that its activities are fundamentally different with respect to a factor such as non–water quality environmental impact than the sources the agency considered when issuing the regulations otherwise applicable to the applicant for a variance.
Fourth, individualized adjustments have been authorized when environmental protection objectives clash with other important social values. Provisions that authorize EPA to exempt activities whose pursuit is important to the national security from compliance with regulatory obligations are commonplace. Similarly, the Clean Air Act has authorized the suspension of controls for fuel burning stationary sources in the event of a national or regional energy emergency. The Clean Air Act also delegates to EPA the power to authorize the production of ozone-depleting substances that are otherwise banned to ensure aviation safety and sanitary food supplies and for use in medical devices, critical agricultural uses, and fire and explosion prevention.
The risk regulation statutes have also acknowledged and accommodated important social values by affording preferential treatment to small business entities. Some of these accommodations take the form of regulatory exemptions or deadline extensions, while others are reflected in direct and indirect subsidies. Agencies such as EPA have supplemented these accommodations by choosing not to rigorously enforce some obligations against small entities. Such accommodations may be based on the premise, which has been vigorously contested, that the adverse environmental impacts generated by small businesses are sufficiently small that allowing them to avoid regulations applicable to larger concerns will not cause unacceptable setbacks to environmental protection goals. Alternatively, they may be based on a desire to protect small businesses because of the vital role they have traditionally played in providing for a diverse economy.
The foregoing survey indicates that back-end adjustments can prevent command and control regulations from producing needlessly inefficient and otherwise perverse results. These types of back-end adjustments also reduce the necessity for a comprehensive assessment of potential regulatory impacts before a regulation is promulgated. Indeed, after-the-fact adjustments of the sort involved in issuance of variances and exceptions will often provide a better forum for policy-makers to fine-tune regulation, because experience with operation of the regulations may reduce uncertainties and resulting bounded rationality. Likewise, the smaller scope of the exercise may make it easier for a decision-maker to understand the implications for relevant values of a decision to grant or deny relief in a particular case than it would be to predict such consequences on a larger scale.
Agencies such as EPA have engaged in unilateral as well as negotiated adjustments to traditional regulatory programs to increase their efficiency or enhance their performance. Of the pollution control programs administered by EPA, none has been the subject of more sustained and virulent criticism than the hazardous substance cleanup program authorized by the Comprehensive Environmental Response, Compensation, and Liability Act. In response to those criticisms, and perhaps as a means of staving off legislative reforms to which the agency would have objected, EPA adopted a series of administrative reforms during the Clinton administration that were designed to increase enforcement fairness, cleanup effectiveness, and consistency, as well as state and public participation, while reducing transaction costs. The agency’s Brownfields Economic Redevelopment Initiative, for example, was meant to empower states, local governments, communities, and other stakeholders interested in economic redevelopment to work together to prevent, assess, safely clean up, and sustainably reuse brownfields. ... One of the major enforcement activities under this initiative is to remove liability barriers to the cleanup and redevelopment of these contaminated properties, thereby giving prospective purchasers, lenders, and property owners more assurances of a safe investment. EPA also took steps to mitigate the difficulties experienced by owners of property containing contaminated aquifers in selling their properties or obtaining financing for development because of the risk of incurring cleanup liability. If statutory delegations are sufficiently flexible to allow for the exercise of agency discretion and initiative, similar efforts to adjust ongoing regulatory programs could yield the same kinds of beneficial consequences as the issuance of deadline extensions and exceptions.
A form of back-end adjustment mechanism that has received considerable attention in the field of public natural resource management is adaptive management. Originally conceived of by researchers at the Institute of Animal Resource Ecology at the University of British Columbia, and later popularized by political scientist Kai Lee, adaptive management has been described by its proponents as “learning by doing.” It involves a willingness to act in the absence of complete information, coupled with a commitment to evaluate systematically the results of that action as it occurs. New actions are then formulated in light of the information derived from previous experience. This approach is designed to reflect a bias toward action rather than study without “losing sight of the perils of scientific uncertainty.” It is based on the recognition that the complexity and unpredictability of natural systems necessarily renders management efforts experimental. Thus, adaptive management presumes ongoing institutional transformation; entities should develop their philosophies and strategies in an evolutionary way through continuous adaptation and assessment. These changes should be driven by a constant flow of information gathered from purposeful experiments. Adaptive management may be thought of as a research strategy designed to generate feedback.
Adaptive management has been used for years to assist in ecosystem management, such as the preservation of fish and wildlife resources in the Pacific Northwest and of endangered species more generally, but it has not been widely used in other areas of environmental policy. Recently, however, it has been described as the basis for regulatory reform efforts at EPA, such as the Project XL program, and it has filtered into a variety of other government environmental policy proposals and programs.
CONCLUSION
Pragmatism recognizes that institutional arrangements are context-specific. Instead of assessing the value of risk regulation mechanisms against an ideal standard, a pragmatic approach demands evidence that proposed reforms to such arrangements will yield superior results in practice. We endorse efforts by policy-makers to take steps to reduce the need for the kind of information that bounded rationality is likely to make difficult to obtain. Because of the uncertainty that tends to limit our ability to understand (and certainly to quantify) regulatory costs and benefits, any attempt to fashion a general regulatory standard that universally avoids irrational results is doomed to failure. Regulated entities often have access to more information about the cost of regulation than the regulatory agency does. To the extent that regulators must rely on cost estimates supplied by the regulated community in developing regulatory standards, they are to a certain extent at the mercy of that community. Regulated entities have a built-in incentive to behave opportunistically by overestimating the cost of compliance. Affording regulated entities flexibility in the choice of the means of compliance with regulatory standards turns a potential negative factor at the standard development stage into a positive one at the stage of regulatory implementation. It takes advantage of the superior knowledge of regulated entities by affording them incentives to select the most cost-effective means of compliance possible.
Another component of pragmatism is its commitment to incremental decision-making as a means of addressing bounded rationality problems. In the context of risk regulation, such a commitment entails testing regulatory approaches on an ongoing basis to determine which ones fail to provide useful solutions, so that they can be adjusted or replaced. We therefore endorse commitment by regulatory agencies to incremental change in the implementation of risk regulation through resort to a series of mechanisms by which they may adjust the application or enforcement of general risk regulation standards in particular cases. Pragmatic regulatory implementation of the sort we describe in this chapter allows agencies to take advantage of knowledge of the impacts of regulation – beneficial and adverse – yielded through real world regulatory applications. The current risk regulation statutes provide many opportunities for this kind of “back-end” flexibility, and agencies have invented others even where they are not explicitly authorized. To the extent that flexibility of this sort is not currently available, we favor steps to make it so.
LITERATURE
1 Shapiro, S. Risk Regulation at Risk : Restoring a Pragmatic Approach/S. Shapiro and R. Glicksman. – Stanford: Stanford University Press. – 1995. – 271 p.
2 Keohane, N. The Choice of Regulatory Instruments in Environmental Policy / Nathaniel O. Keohane, Richard Revesz, Robert Stavins // Harvard Environmental Law Review. – 1998. – Volume 22, number 2. – Pp. 313-367.
3 Sunstein, C. Is the Clean Air Act Unconstitutional? / Cass R. Sunstein // AEI-Brookings Joint Center Working Paper. – 1999 – No. 99-07. – Pp. 1 – 104.
4 McGarity, T. Workers At Risk: The Failed Promise Of The Occupational Safety And Health Administration/Thomas O. McGarity, Sidney A. Shapiro – PRAEGER FREDERICK A. – 1993. – 376 p.
5 Crandall, R. An Agenda for Federal Regulatory Reform / Robert W. Crandall [and other] // American Enterprise Institute for Public Policy Research and The Brookings Institution. – Washington, D.C., 1997. – 27 p.
6 Breyer, S. Breaking the Vicious Circle: Toward Effective Risk Regulation / Stephen Breyer. – Harvard University Press, 1999. – 144 p.
7 Hopkins, T. Regulatory Costs in Profile / Thomas Hopkins // CENTER FOR THE STUDY OF AMERICAN BUSINESS. – Policy Study Number 132, August 1996 – 1–26 p.
8 Williamson, O. The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting / Oliver E. Williamson. – Free Press, 1985. – 450 p.
9 Kelman, S. Cost-Benefit Analysis: An Ethical Critique / Steven Kelman // from AEI Journal on Government and Society Regulation – January-February, 1981. – Pp. 33–40.
10 Marin, A. The reward for risk in the labor market: evidence from the United Kingdom and a reconciliation with other studies / Alan Marin and George Psacharopoulos // Journal of Political Economy. – 90 (4). – Pp. 827–853.