Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
2Conceptual Framework for Regulatory Design This chapter starts with definitions of several key terms used in this report, including a discussion of what constitutes âregulation,â both broadly de- fined and in the narrower context of government safety regulation of high- hazard industries. Although the detailed structure of individual regulations can vary widely, this chapter offers a conceptual framework for distinguish- ing among four basic design types. The framework is founded on two key considerations: (a) whether a regulation commands the use of a means or the attainment or avoidance of some ends and (b) whether the command targets individual components of, or pathways to, a larger problem (micro- level) or directs attention to that larger problem itself (macro-level). This chapter closes with a discussion of what are often viewed as the general advantages and disadvantages of regulations falling within each of the four basic design types from the conceptual framework. Claimed ad- vantages and disadvantages are considered again later in the report, after design types have been reviewed in more detail and the circumstances in which the design types may be applied have been considered. DEFINITIONS Regulation can be construed narrowly in the context of the particular administrative processes of one agency or one country. It can also be con- strued more genericallyâas intended in this reportâto avoid associations limited to the specific procedures of individual agencies and jurisdictions. Because the terms âregulation,â âregulatory regime,â âregulator,â and âregulated entityâ are used throughout this report, their intended mean- 22
CONCEPTUAL FRAMEWORK FOR REGULATORY DESIGN 23 ings deserve explanation. A âregulation,â as construed in this report, is a binding command, not a voluntary guideline or statement of an aspira- tion or goal. Although regulations in their broadest sense can include the commands contained in standards developed by nongovernmental bodies, such as professional and industry trade associations that have sanctioning authority, legally binding regulations are those promulgated by governmen- tal bodies (Sinclair 1997). Such regulations include commands found in statutes, rules issued by government agencies that administer statutes, and legal principles arising from court decisions. By defining regulations as legally binding commands and by not fo- cusing on the provenance of those commands, this report distinguishes an individual regulation from a final rule developed through the U.S. fed- eral administrative rulemaking process.1 A final rule refers to a document promulgated by a federal agency and published in the Federal Register. A single published final rule can contain an entire suite of regulations (i.e., binding commands), each having a different design. The entire collection of regulatory commands that govern a particular industry or type of activ- ity, whether in final rules, statutes, or court opinions, forms a âregulatory regime,â which can comprise hundreds of regulations of different design types. The âregulatorâ is the government entity that creates the command, monitors compliance, and dispenses the consequences for noncompliance; the term is used even when more than one such entity regulates a particular firm or industry. In conventional parlance, the term âregulatorâ is usually reserved for the administrative agencies to whom regulatory implementa- tion and enforcement responsibilities are delegated; however, as already suggested, the regulator may also be the legislature or even the judiciary. The regulatorâs aim is to establish and enforce a command to bring about a socially desirable outcome that would not occur otherwise. The âregulated entityâ is the individual or organization to which the command appliesâ that is, the party that has the legal obligation to comply and bears the consequences of noncompliance. This report focuses on the design of regulationsâthat is, of legally binding commands. It does not examine the many factors that a regulator will wish to consider in devising specific strategies to facilitate and enforce compliance with its regulations. However, some examples of strategies are mentioned because a regulatorâs enforcement capabilities can be important in the choice of a regulatory design, as explained in this report. A regulator may use a single strategy or a combination of strategies to facilitate and enforce compliance. For example, it may require preapproval of activities 1 The administrative rulemaking procedure of the U.S. federal government is discussed in more detail in Chapter 4.
24 DESIGNING SAFETY REGULATIONS FOR HIGH-HAZARD INDUSTRIES before they are allowed to commence (e.g., permits and licenses); actively inspect or audit regulated entities to motivate desired levels of compliance; or take a more reactive approach by responding to complaints or incident reports, in effect sanctioning violators after the fact in an effort to deter noncompliance in the future. Compliance may also be facilitated through features in the regulations themselves. Among such features are require- ments for a regulated firm to provide information useful to the regulatorâs enforcement actions and disclosure of which may motivate the firm to pur- sue higher levels of compliance for fear of attracting unwanted public and marketplace attention to violations (Bennear and Olmstead 2008; Hindin and Silberman 2016). Regulators often carry out other activities that may be considered in- strumental to facilitating regulatory compliance and furthering the goals of the regime, such as educational and awareness campaigns, joint research with industry, data collection and analysis, and participation in third-party standard- and guideline-setting activities. Regulators sometimes advise on the use of such external standards or guidelines. How to design such vol- untary guidance to encourage its use is of interest to regulators and the subject of research (Coglianese and Nash 2006). However, in accordance with the study charge, the subject of this report is the design of legally bind- ing commandsâgovernment regulationsârather than voluntary programs, guidelines, and other nonbinding standards. CONCEPTUAL FRAMEWORK As discussed in Chapter 1, the existence of a variety of often ambiguous and sometimes misleading classifications of regulations has contributed to confusion about the choices available to regulators. The labels that others have used to describe regulation have focused on characteristics such as whether a regulation creates incentives for technological development and application, mandates data reporting, requires risk analysis, and specifies design parameters. Richards (2000) summarizes dozens of classification schemes in the literature, many of which contain different labels used to de- scribe regulations that have fundamentally similar regulatory designs. The varied labels and taxonomies have often clouded, rather than clarified, the key dimensions of regulatory design and the choices confronting regulators when they select a design type. Accordingly, this study required a common terminology and framework to make conceptual distinctions among regulations. The conceptual frame- work that is developed next in this chapter was derived from the scholarly literature on regulation and provides the basis for the nomenclature and organizing principles of the remainder of this report. The framework dis- tinguishes between regulating specific means in a pathway of industrial
CONCEPTUAL FRAMEWORK FOR REGULATORY DESIGN 25 activities and regulating the ends, or outcomes, of those activities. It further distinguishes between regulating components of a system (micro-level) and regulating the system as a whole (macro-level). This conceptual framework is applied throughout this report to describe and compare safety regulations used in high-hazard industries. Because regulations often carry vernacular labels, some of the more common labels are mapped to this reportâs typol- ogy of the four basic design types. Means Versus Ends A regulation can command that the regulated entity take or avoid an ac- tion, with the intention of furthering a regulatory goal or achieving an outcome related to that regulatory goal. âMeans-basedâ regulation focuses on actions, such as the use of a technology or practice. For example, safety regulators may require firms to install a particular type of valve, retain cer- tain documents, conduct certain observations or measurements, or inspect the condition of equipment at specified intervals. A required âmeansâ in occupational safety regulation might be the installation of a hazard warning sign or the establishment of a worker hazard awareness training program. Alternatively, a regulation can mandate the achievement or avoidance of certain ends. âEnds-basedâ regulation may require that a code-compliant building be capable of evacuating all occupants in a designated time, that a factory keep its emission of air pollutants below certain levels, or that an employer keep the workplace free of all identifiable hazards. The commands in some regulations contain both means- and ends- based elements. For example, a regulation may require the use of protective equipment (means) that has passed approved testing standards for fire and impact resistance (ends). Micro Versus Macro A regulationâwhether means-based or ends-basedâcan be distinguished along a second dimension relating to the regulationâs focus or target. The regulation can be described as âmicro-levelâ when it is targeted to a specific contributor or causal pathway to the ultimate problem that motivates regu- lation, or it can be described as âmacro-levelâ when its focus is widened to the ultimate problem itself. Whether the regulation directs the attention of the regulated entity to the ultimate problem or to a causal pathway leading to that problem constitutes a crucial distinction in the types of obligations that the regulation imposes, as discussed in more detail in this report. Micro-level regulations are more common because they are often bun- dled to address a problem. For example, the ultimate problem addressed by a traffic safety regulation is to reduce the harm caused by motor vehicle
26 DESIGNING SAFETY REGULATIONS FOR HIGH-HAZARD INDUSTRIES crashes. To achieve that purpose, the regulators will typically disaggregate the traffic safety problem into its component parts and issue regulations tar- geted to each part. The hundreds of Federal Motor Vehicle Safety Standards issued by the National Highway Traffic Safety Administration (NHTSA) require that automakers build cars with features providing various capabili- ties, such as exterior lighting, stability, braking, and occupant protection, that bear on a vehicleâs crash potential and severity. These micro-level regu- lations are accompanied by many other government regulations, such as state requirements controlling driver qualifications and the wearing of seat belts, that affect the incidence and severity of motor vehicle crashes. In this regard, each individual traffic safety regulation is designed to target, almost in isolation, one of the many factors in the causal pathways or network leading to the ultimate problem of traffic fatalities and injuries. In contrast to targeting intervention at a micro-level, a regulation can be designed to draw the regulated entityâs attention to the ultimate problem that motivates government intervention. For example, NHTSA also has responsibility for establishing regulations to increase motor vehicle fuel economy. To achieve this purpose, NHTSAâs regulations do not target the many individual attributes of a vehicle that can affect fuel economy, such as its weight, engine displacement, or aerodynamics.2 Instead, the regulations require automakers to achieve a specified average fleetwide fuel economy level. In this way, NHTSAâs fuel economy regulations impose obligations that directly target the ultimate goal while allowing automakers to adjust fleet mix and vehicle attributes as they see fit. Macro-level regulations like these are not directed at means or ends related to individual causal factors leading to an ultimate problem; they impose means- or ends-based obliga- tions that focus directly on that ultimate problem itself. Distinguishing Regulatory Designs Whether the target of a regulation is focused at a point distant from (micro- level) or closer to (macro-level) the ultimate regulatory problem can have implications for the number of regulations that are needed, the knowledge a regulator must possess about the causes of the problem, and the ability of the regulator to monitor and verify compliance. Whether the regulatory command is means- or ends-based can also have important implications for the ability of the regulated entity to innovate and find lower-cost ways to comply as well as for the enforcement burden of the regulator. Because such implications need to be considered by regulators when they decide on 2 To ensure that its regulations can be reasonably met by automobile manufacturers and suppliers and to further industry compliance, NHTSA does support research and technology activities that concern many of these specific vehicle attributes.
CONCEPTUAL FRAMEWORK FOR REGULATORY DESIGN 27 a regulatory design, a conceptual framework that differentiates regulations according to these basic design features can be a valuable decision-making tool. Table 2-1 illustrates the relevance of these distinctions by mapping some example regulatory commands to the four regulatory design types that follow from the conceptual framework of this report. The terms used to describe the four design typesâmicro-means, micro-ends, macro-means, and macro-endsâare not currently part of the common parlance of safety or regulatory professionals. They are used in this report because they cap- ture more precisely and accurately the underlying differences that most professionals have in mind in discussing regulatory designs. The terms used in common parlance are often imprecise, inconsistent, or overlapping. Some are normatively loaded as well. What this report calls âmicro-meansâ regulation, for example, is often referred to as âprescrip- tiveâ regulation, which often carries a negative connotation. What this report calls âmicro-endsâ regulation is often referred to as âperformance- basedâ regulation, which, as noted in Chapter 1, can carry a more positive connotation. Furthermore, some regulatory professionals use âperformance-based regulationâ to refer to what this report calls âmacro-meansâ regulationsâ that is, to requirements that firms adopt certain management programs and TABLE 2-1 Four Basic Types of Regulations with Examples Means Ends Micro Micro-Means â¢ Install a hazard warning sign having a certain color scheme â¢ Install a particular type of valve â¢ Inspect the condition of equipment at a defined time interval â¢ Construct a pipeline by using a specified grade of steel Micro-Ends â¢ Ensure that an electrical component of a product passes a test for shock resistance â¢ Limit sulfur dioxide emissions to certain levels â¢ Demonstrate the capability to evacuate all occupants from a building in a designated time Macro Macro-Means â¢ Engage in threat and risk analysis â¢ Establish and execute a safety management program â¢ Reevaluate and revise safety management plan at regular intervals Macro-Ends â¢ Keep the workplace free from recognized hazards â¢ Design and maintain a facility to prevent releases of hazardous substances â¢ Avoid a transportation accident SOURCE: Adapted from Coglianese 2010.
28 DESIGNING SAFETY REGULATIONS FOR HIGH-HAZARD INDUSTRIES risk analysis activities. Yet such macro-means regulations are âprescriptiveâ in that they require firms to take certain actions (i.e., in ordinary language, they prescribe actions). Those actions are intended to cause a firmâs manag- ers to identify strategies for achieving the end state reflected in the regula- torâs ultimate outcome of concern (Bardach and Kagan 1982; Coglianese and Lazer 2003; Huising and Silbey 2011). Macro-means regulations do not require firms to adopt specific risk reduction technologies or practices or even to achieve specific limits on risk levels or other measures of safety performance. Thus, this type of regula- tion gives firms flexibility in choosing their micro-level behavioral routines and technologies. However, macro-means regulations do not allow the substitution of some other type of action; they literally tell firms how to manage themselves. The regulations mandate that firmsâ managers study their operations comprehensively and develop strategies suited to mitigat- ing the risks they identify (Coglianese 2010). Often, this type of regulation imposes on firms the obligation to âplan-do-act-checkâ with respect to ad- dressing a problem (Coglianese 2010). The regulations usually define the basic elements of a compliant management system. For example, firms may be expected to conduct an internal risk analysis; identify and evaluate risk control options; implement preferred controls; prepare a written plan for communicating safety-related work rules and ensuring they are understood and observed; and establish procedures supporting manager and worker training, documentation, and compliance monitoring (Silbey and Agrawal 2011). In addition, the regulations may require periodic program audits and feedback mechanisms to support efforts to improve the firmâs manage- ment (Chinander et al. 1998; Coglianese and Lazer 2003; Kunreuther et al. 2002). The legal commands in the cell in Table 2-1 labeled âmacro-endsâ deserve mention because they do not bring to mind regulations in the classic sense. That is because they have not been operationalized into any specific proactive and narrowly defined obligations designed to prevent the occurrence of the ultimate problem. However, the imposition of liability or penalties if such a problem does occur brings about a type of regulatory obligation. A good example of a macro-ends command is a general duty provision in a statute or regulation, such as the Occupational Safety and Health Actâs requirement that employers ensure that their workplaces âare free from recognized hazards that are causing or likely to cause death or serious physical harm to . . . employees.â3 Other examples are the liability provisions contained in the Oil Pollution Act and the Clean Water Act, 3 Occupational Safety and Health Act Section 5(a)(1).
CONCEPTUAL FRAMEWORK FOR REGULATORY DESIGN 29 which impose financial consequences on offshore oil and gas operators for oil spills.4 By imposing liability or penalties when problems arise, macro-ends commands can create incentives that operate behaviorally in a manner similar to traditional, ex ante regulation (Kolstad et al. 1990). For example, a commonly observed safety practice is the placement by building mainte- nance personnel of small warning signs notifying passersby of wet floors, a practice that is intended to prevent someone from slipping. When macro- ends commands result in such behaviors, they can, in practice, yield results similar to what other types of means-based regulation might produce. NOMENCLATURE MAPPED TO COMMON REGULATORY LABELS As noted earlier, Richardsâ (2000) review of the scholarly literature identi- fied a wide range of labels used to describe different types of regulations. The ambiguity and diversity of existing labels led the committee to refrain from trying to identify and map existing labels definitively to the four cells in Table 2-1. However, for readers with an interest in knowing where some common labels used to describe regulatory designs might fit into the orga- nizing scheme of this report, a loosely matched list is provided in Table 2-2. In the sections of this report that follow, these more common labels are sometimes provided in parentheses next to the four regulatory design terms simply to remind the reader of the types of regulations being discussed. How some frequently mentioned types of regulation fit into the four cells of Tables 2-1 and 2-2 may not be immediately obvious. For example, âmarket-basedâ regulationsâsuch as permit trading regimes, cap-and-trade systems, and emissions taxes (Schmalensee and Stavins 2017; Tietenberg 2006)âare actually just a type of micro-ends regulation. Micro-ends regu- lations often require each regulated entity to achieve the same outputs, but with market-based regulation, outputs can vary across regulated entities. A market-based emissions tax imposes an obligation to pay a tax on outputs measured on a per unit (marginal) basis. This is a micro-ends regulation un- der which the âpenaltiesâ for producing emissions are meted out marginally and called a tax instead of a penalty. Under a cap-and-trade or emissions trading scheme, just as with an emissions tax, the performance of regu- lated facilities can vary. With emissions trading, the particular micro-ends obligation that each firm must meet will simply depend on the quantity of tradable permits it holds. So-called âinformation disclosure regulationâ is another example of a type of regulation that might at first appear to be difficult to place in one of the four cells in Tables 2-1 and 2-2. The reason lies in the different purposes 4 Clean Air Act Section 112(r)(1).
30 DESIGNING SAFETY REGULATIONS FOR HIGH-HAZARD INDUSTRIES behind the information generation and reporting that are required (Fung et al. 2007; Hamilton 2005; Kleindorfer and Orts 1998; Sunstein 1999). The cell in Table 2-1 into which information disclosure requirements fit will depend on their purpose. When information disclosure is intended as a means to a desired end state, such as an informed consumer, it can be characterized as micro-means. If the purpose of information disclosure is to increase the regulated entityâs awareness of its contribution to the ultimate problemâand thus to prompt it to address that problem5âthe requirement can be characterized as macro-means. ADVANTAGES AND DISADVANTAGES COMMONLY ASSOCIATED WITH REGULATION DESIGN TYPES These examples of the different terms used to describe regulatory designs indicate how much variation can exist across individual regulations even though they can be grouped generally into four main design types. Regula- tions can vary still further within each design type, not just across the four main types. In other words, not all micro-means regulations are the same; not all macro-means regulations are the same; and so forth. Differences in how regulations falling within the same category in Table 2-1 are struc- tured and in the exact legal duties they impose can affect what a regulation 5 Thaler and Sunstein (2009, 191) have considered certain environmental information disclosure requirements to constitute what they call a âsocial nudge,â because disclosure of pollution can serve as a means of informing community members, who in turn put pressure on firmsâ managers to improve performance. TABLE 2-2 Map of Common Regulation Descriptors to Conceptual Framework Means Ends Micro Micro-Means â¢ Prescriptive regulation â¢ Design standards â¢ Technology-based regulation â¢ Specification standards Micro-Ends â¢ Performance-based regulation â¢ Output-based regulation â¢ Market-based regulation Macro Macro-Means â¢ Management-based regulation â¢ Performance-based regulation â¢ System regulation â¢ Goal-based regulation â¢ Safety case regulation â¢ Enforced self-regulation Macro-Ends â¢ Tort and ex post liability â¢ General duty provisions â¢ Outcome-based regulation SOURCE: Adapted from Coglianese 2010.
CONCEPTUAL FRAMEWORK FOR REGULATORY DESIGN 31 achieves and what costs or other adverse effects it produces. In Chapter 4, a more detailed discussion of the various ways that regulations of the four basic design types can be structured is provided. Implications follow from distinguishing the four main types of regula- tory designs. Each of the four types is associated with what are generally considered to be different advantages and disadvantages, which can be useful as a starting point in deciding which regulatory design would best be used in a particular case. A clear understanding of the differences between the four main design types is essential in selecting and justifying the design type most suitable for solving a given problem. Micro-means regulations (which, as noted, are often called âprescrip- tiveâ) provide clear instructions about actions that must be taken by the regulated party, which can be easily monitored to verify compliance. How- ever, these regulations are further removed from the ultimate health, safety, and environmental concerns that motivate government intervention. They offer little flexibility (short of waiver or revocation) for the regulated in- dustry in responding to the regulation, even if better means are or may become available. Micro-ends regulations (which are often called âperformance-basedâ) have the advantage of being closer to the ultimate health, safety, or environ- mental concern of the regulator. For example, setting standards with regard to oil concentration in emissions from offshore oil platforms is closer to the end (prevent damage to the offshore environment) than is requiring a specific waterâoil separation technology. Micro-ends regulations also allow more flexibility by the regulated industry in meeting the regulation. How- ever, such regulations can carry disadvantages; in some contexts a firmâs attainment of the required ends can be difficult to monitor. For example, an ends-based mandate to reduce emissions can require investments in monitoring, testing, and modeling technologies that are costly or that may be insufficiently reliable for verifying compliance. Macro-means regulations (which are often called âmanagement-basedâ) have the advantage of providing flexibility for the regulated entity with regard to operational actions to undertake, technologies to use, or plans to be formulated by management. They can be easier for the regulator to develop than a collection of highly targeted, micro-means regulations. They can be used when outcomes are difficult to measure directly. Such regula- tions may also infuse a greater sense of responsibility and accountability (i.e., safety culture) into the regulated firms. Among their disadvantages may be the limitations of regulatory agencies in monitoring and enforcing them if the agencies lack the expertise to review firm-specific management plans and the execution of those plans. Comparable limitations may exist among firms, especially smaller ones, which may need to hire personnel with new skills (or train existing personnel) to conduct risk analyses, design
32 DESIGNING SAFETY REGULATIONS FOR HIGH-HAZARD INDUSTRIES and implement complex management systems, and monitor management plans and their execution. Macro-ends regulations (such as tort liability and âgeneral dutyâ pro- visions) have the advantage of focusing directly on the ultimate ends. However, such ex post liability and penalties may not always be viewed as adequate for spurring efforts to achieve those ends. For example, in the safety context, firms may underestimate their liability because they believe that incidents will not occur, because bankruptcy and insurance protections limit their liability as a practical matter, or because the benefits of misconduct may exceed the ultimate penalties (Bennear 2012; Manski 2004; Manski and Molinari 2010). Furthermore, in cases where the harm created by noncompliance can be catastrophic, ex post liability and penalty determinations can be complicated and unacceptable as the exclusive means of regulatory control. These claimed general advantages and disadvantages are discussed in more detail in Chapter 4, where they are also considered in the context of structural differences in regulations as well as differences in the nature of regulatory problems, variation in regulatorsâ capabilities, and differences in firmsâ characteristics. The asserted generic advantages and disadvantages were raised here to show that much can be at stake when a regulator, con- fronting a specific safety challenge in a high-hazard industry, must choose between different regulatory designs. Chapter 3 considers case studies showing how, in multiple countries, the four main regulation design types are used in different ways, and in combination with one another, to regulate high-hazard industries. REFERENCES Bardach, E., and R. A. Kagan. 1982. Going by the Book: The Problem of Regulatory Unrea- sonableness. Temple University Press, Philadelphia, Pa. Bennear, L. S. 2012. Beyond Belts and Suspenders: Promoting Private Risk Management in Offshore Drilling. In Regulatory Breakdown? The Crisis of Confidence in U.S. Regula- tion (C. Coglianese, ed.), University of Pennsylvania Press, Philadelphia, Chapter 3. Bennear, L. S., and S. M. Olmstead. 2008. The Impacts of the âRight to Knowâ: Information Disclosure and the Violation of Drinking Water Standards. Journal of Environmental Economics and Management, Vol. 56, No. 2, pp. 117â130. Chinander, K. R., P. R. Kleindorfer, and H. C. Kunreuther. 1998. Compliance Strategies and Regulatory Effectiveness of Performance-Based Regulation of Chemical Accident Risks. Risk Analysis, Vol. 18, No. 2, pp. 135â143. Coglianese, C. 2010. Management-Based Regulation: Implications for Public Policy. In Risk and Regulatory Policy: Improving the Governance of Risk (G. M. Bounds and N. Malyshev, eds.), Organisation for Economic Co-operation and Development, pp. 159â183. Coglianese, C., and D. Lazer. 2003. Management-Based Regulation: Prescribing Private Man- agement to Achieve Public Goals. Law and Society Review, Vol. 37, No. 4, Dec., pp. 691â730.
CONCEPTUAL FRAMEWORK FOR REGULATORY DESIGN 33 Coglianese, C., and J. Nash. 2006. Management-Based Strategies: An Emerging Approach to Environmental Protection. In Leveraging the Private Sector: Management-Based Strate- gies for Improving Environmental Performance (C. Coglianese and J. Nash, eds.), Re- sources for the Future, Washington, D.C., pp. 3â30. Fung, A., M. Graham, and D. Weil. 2007. Full Disclosure: The Perils and Promise of Trans- parency. Cambridge University Press. Hamilton, J. T. 2005. Regulation Through Revelation: The Origin, Politics, and Impacts of the Toxics Release Inventory Program. Cambridge University Press. Hindin, D., and J. Silberman. 2016. Designing More Effective Rules and Permits. George Washington Journal of Energy and Environmental Law, Vol. 7, No. 2, pp. 103â123. Huising, R., and S. S. Silbey. 2011. Governing the Gap: Forging Safe Science Through Rela- tional Regulation. Regulation and Governance, Vol. 5, March, pp. 14â42. Kleindorfer, P. R., and E. W. Orts. 1998. Informational Regulation of Environmental Risks. Risk Analysis, Vol. 18, No. 2, pp. 155â170. Kolstad, C. D., T. S. Ulen, and G. V. Johnson. 1990. Ex Post Liability for Harm vs. Ex Ante Safety Regulation: Substitutes or Complements? American Economic Review, Vol. 80, No. 4, Sept., pp. 888â901. Kunreuther, H. C., P. J. McNulty, and Y. Kang. 2002. Third-Party Inspection as an Alterna- tive to Command and Control Regulation. Risk Analysis, Vol. 22, No. 2, pp. 309â318. Manski, C. F. 2004. Measuring Expectations. Econometrica, Vol. 72, No. 5, pp. 1329â1376. Manski, C. F., and F. Molinari. 2010. Rounding Probabilistic Expectations in Surveys. Journal of Business and Economic Statistics, Vol. 28, No. 2, pp. 219â231. Richards, K. R. 2000. Framing Environmental Policy Instrument Choice. Duke Environmental Law and Policy Forum, Vol. 10, No. 2, pp. 221â286. Schmalensee, R., and R. Stavins. 2017. Lessons Learned from Three Decades of Experience with Cap and Trade. Review of Environmental Economics and Policy, Vol. 11, No. 1, pp. 59â79. Silbey, S. S., and T. Agrawal. 2011. The Illusion of Accountability: Information Management and Organizational Culture. Droit et SociÃ©tÃ©, No. 77, pp. 69â86. Sinclair, D. 1997. Self-Regulation Versus Command and Control? Beyond False Dichotomies. Law and Policy, Vol. 19, No. 4, pp. 529â559. Sunstein, C. R. 1999. Informational Regulation and Informational Standing: Akins and Be- yond. University of Pennsylvania Law Review, Vol. 147, pp. 613â675. Thaler, R. H., and C. R. Sunstein. 2009. Nudge: Improving Decisions About Health, Wealth, and Happiness. Penguin. Tietenberg, T. H. 2006. Emissions Trading: Principles and Practice, 2nd ed. Resources for the Future, Washington, D.C.