Framing Environmental Policy Instrument Choice

Beginning of article

I. INTRODUCTION

The student of environmental policy is exposed to a myriad of policy instruments the government can employ: emissions taxes, abatement subsidies, marketable allowances, regulation based on performance standards or technology, property rights, deposit-refund schemes, information programs, liability rules, and a number of related policy tools. When Congress crafts environmental legislation or administrative agencies promulgate rules to implement environmental policy goals, they must choose from among these policy tools. In an ideal world, they would employ those instruments that will allow the government to meet its goals at the lowest possible cost subject to external constraints.

One consistent message from the environmental economics literature is that incentive-based instruments are a more cost-effective means to achieve environmental goals than alternative policy instruments such as technology-based standards.(1) In practice, however, this counsel has only rarely been heeded.(2) In the United States, environmental protection schemes have evinced more diversity than uniformity with respect to key characteristics. Perhaps this diversity should not be surprising. Consider the wide variety of activities that policy instruments are designed to promote: nonpoint source water pollution control, protection of wetlands, reduction of sulfur dioxide emissions, clean-up of hazardous waste sites, and protection of endangered species. Given this tremendous variation among environmental goals and the means of achieving those goals, is there any reason to expect that when it comes to implementation we can exercise a one-size-fits-all approach to instrument choice?

In fact, what we observe is a plurality of instruments and combinations thereof that have steadfastly defied economists' and policy analysts' prescriptions.(3) Part of this deviation between normative prescription and actual practice can be attributed to political considerations,(4) and part to legal constraints. However, there may also be normative justifications for choosing instruments outside the increasingly standard incentive-based dyad of taxes and marketable allowances. While those instruments may minimize the direct cost of producing environmental services, such as emissions abatement, they do not necessarily minimize the total social costs. To see this, however, requires a reformulation of the standard evaluation criteria that are applied in most policy instrument studies and a broadening of the range of instruments under consideration.

The purpose of this article is to provide a conceptual framework for understanding the policy instruments that the government uses when it implements environmental policy goals. Unlike other surveys that focus on the contrasts among instruments, the emphasis here is on exploring the continua of instruments that are available for a variety of applications. In the process of developing the framework, the article will advance key ideas related to the choice and structure of evaluation criteria, the design and role of a taxonomy of instruments, and the application of lessons from a variety of distinct literatures.

Many studies of policy instruments have identified evaluation criteria by which the performance of policy tools may be judged.(5) This article expands the insights from those analyses by recognizing that two key criteria can be added to their lists: (1) the differential effects of policy instruments on the system of public finance and (2) legal constraints that may render optimal solutions (from a cost-minimization perspective) infeasible. Reformulating the evaluation criteria as a cost-minimization problem provides additional insight into the nature of the tradeoffs that the policymaker faces in choosing among the various instruments.(6)

This article also demonstrates that the taxonomies of policy instruments can be re-framed to bring out several factors. First, while liability rules, property rights, taxes, marketable allowances, and subsidies are often grouped together as incentive-based approaches,(7) the new framework separates the first two from the latter three to emphasize the fundamentally different role that the government plays.(8) Second, whereas recent studies have often depicted instrument choice as a matter of command-and-control versus incentive-based instruments,(9) the new framework suggests that this approach is a false dichotomy, stemming in part from an overly narrow depiction of the range of instruments.(10) Third, several studies have focused on the quantity-versus-price relation between marketable allowances and taxes,(11) both of which are coercive tools of the government. This new framework demonstrates that the duality relationship of quantity-versus-prices also exists when government uses its enterprise tools of subsidies and contracts.(12) Fourth, whereas broad review studies of instrument choice have largely focused on how policy tools provide incentives for efficient pollution abatement, this framework also addresses the question of who bears the costs of unabated pollution.(13) Finally, the new framework suggests that information-related programs, identified as separate categories of instruments in the policy literature,(14) are in fact addressing different goals, not employing different instruments.(15)

By restructuring the evaluation criteria and taxonomies of environmental policy and economics studies, this new framework facilitates the application of the lessons from several distinct literatures: New Institutional Economics (transaction cost economics),(16) public economics,(17) and administrative and constitutional law.(18) Applying lessons from the first two bodies of work elucidates the continuous, rather than distinct, nature of the relation among the instruments. Applying lessons from administrative and constitutional law illustrates the legal constraints on the instrument choice problem.

In addition to these specific points, this article advances a more general theme. In the recent past academics and policy analysts have emphasized the important efficiency advantages of incentive-based instruments for addressing pollution abatement. In that process the focus of the discussion has narrowed to a subset of environmental policy tools--specifically, marketable allowances and emissions taxes.(19) While it is important to remember the allocative efficiency advantages of incentive-based instruments, there is also a reason that other, seemingly less efficient, instruments exist. Those reasons are not purely political or distributional. In some settings, instruments such as command-and-control regulation and government production may provide greater overall economic efficiency than their incentive-based counterparts. The optimal choice of policy instrument to implement a particular pollution abatement goal depends upon 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. In short, when it comes to environmental policy instrument choice, one size does not fit all.

This discussion focuses on the normative issues related to environmental policy design, and addresses only tangentially the important questions of political economy. The discussion is based on the assumption that Congress or an administrative agency has adopted a new pollution abatement goal and is searching for the policy instrument(s) that will most efficiently implement the program.(20) The treatment is broad brush in nature, attempting to synthesize the vast amount of analysis available in the environmental policy literature.

Most instrument choice studies first describe the many policy tools available and then provide evaluation criteria for comparing the instrument. This analysis reverses that order on the premise that the taxonomy should be structured to facilitate the evaluation.

II. EVALUATION CRITERIA

Articles that review policy instruments and the instrument choice problem often provide criteria to assess the relative merits of the various policy instruments.(21) This section provides a brief overview of the criteria from several studies, suggests logical relations among the criteria, and identifies useful additions to the list. This leads to a reformulation of the criteria as a constrained cost-minimization problem.

A. Standard Treatment

Table A1 provides a sampling of the evaluation criteria described in a number of the studies. What is perhaps most interesting about these lists is that they are generally unstructured. Referring to the criteria listed in his own work, Majone commented:

   The first observation to be made about the criteria is their heterogeneity
   of values; the second is their lack of specificity. The heterogeneity is
   exemplified by the difficulty of reconciling political feasibility with the
   conditions for economic efficiency. The lack of specificity is particularly
   obvious in the case of the polluter-pays principle, which, as usually
   interpreted, rules out only the subsidies alternative; and even this
   limitation is weakened by a number of exceptions.(22)

Although the studies use different terms, they are actually surprisingly consistent on a conceptual level. All studies listed in Table A1 refer to some form of efficiency or cost-effectiveness, while three refer to some form of dynamic incentives to innovate and develop new abatement technologies. All of the studies also include criteria related to the administrative burden, information requirements, or monitoring costs associated with particular instruments. The studies all include criteria related to political feasibility, distribution and equity effects, or public acceptance. Four studies incorporate effectiveness or efficacy as a criterion. Interestingly, only two of the six studies reviewed in Table A1 explicitly list flexibility, or adaptability in the face of change,(23) as an important criteria for evaluating policy tools. This later point is particularly important, because the ability to adapt to change can be an important factor that favors non-incentive-based instruments under certain circumstances.

B. Expansion and Reformulation of the Evaluation Criteria

Two important criteria are conspicuously absent from the lists in Table A1. First, there is no mention in those lists of the costs that are imposed on society by the distortionary effects that taxes have on production and consumption decisions. Generally, when the government intervenes in the market, for example, by taxing production or consumption, it distorts prices, the signals by which the market allocates resources. This can lead to significant social welfare losses known as "excess burden."(24) If the instruments have a differential impact on the system of public finance, then it is important to include that effect in the evaluation criteria.

Second, while all of the studies in Table A1 include criteria for some type of political constraint, none include legal constraints. It is important to recognize that for some applications, instruments that may be theoretically most cost-effective are not available due to legal constraints.(25)

If the objective of the instrument choice exercise is to minimize the social cost of achieving a given environmental goal (as it is assumed to be in the normative context of this framework), then the evaluation criteria can be structured as a constrained optimization problem. Under the cost-minimization formulation, we can identify three types of economic costs: production costs (PC), implementation costs (IC), and public finance impacts (TX). The stylized policy problem then becomes:

Minimize (PC + IC + TX)
Subject to the constraints:
Pollution abatement requirement
Legal constraints
Political constraints

Production costs (PC) refer to the actual capital, training, operation, maintenance and management costs of producing emissions abatement or other environmental services. They are generally included in evaluation criteria as static efficiency,(26) dynamic incentives,(27) or cost-effectiveness.(28) These are the costs that are most commonly addressed in the environmental economics literature.

Public finance impacts (TX) have not generally been mentioned in discussions of environmental policy instrument choice. In this cost-minimization problem, TX refers to the costs imposed on the system of public finance related to instruments' revenue-raising requirements, including excess burden and administration costs.

Transaction or implementation costs (IC) have been analyzed in the transaction costs/contracts literature,(29) but this work has not been extensively applied to environmental policy.(30) The evaluation criteria in Table A1 include terms for information requirements,(31) ease of monitoring and enforcement,(32) demand on government resources,(33) and adaptability and flexibility in the face of changes in tastes, technology, resource use, and the economy.(34) Lessons from the environmental economics and New Institutional Economics literatures both provide insight into how policy instrument selection affects implementation costs.(35)

The first constraint in the cost-minimization formulation is that the instrument employed to implement environmental policy must accomplish the desired policy goal or target level of abatement. That is, the instrument must provide efficacy or assurance that the goals will be met.(36) In addition, legal and political constraints limit the range of instruments that can be used under particular circumstances. The cost-minimizing instrument, in other words, may not be in the feasible set. Table A1 includes references to political constraints,(37) understandability to the public,(38) distributional issues,(39) fairness,(40) and environmental equity and justice.(41) These are all incorporated in the cost-minimization formulation under the rubric of political constraints. Interestingly, none of the works cited in Table A1 recognized legal constraints that may limit the range of instruments available for a particular application.(42)

III. TAXONOMY OF ENVIRONMENTAL POLICY INSTRUMENTS

A. Standard Treatment

The environmental policy literature has produced many taxonomies of policy instruments over the past 30 years. Table A2 provides a summary of the taxonomies contained in a sample of works that were chosen to span much of the history of the environmental policy literature and to reflect both recognized academic authorities, government studies, and international agency works.(43) The degree to which the studies cited in Table A2 explicitly outline or simply imply the taxonomies varies substantially across the studies.(44)

The taxonomies may be more remarkable for their similarities than their differences. Still, there are several important patterns that emerge. First, the recognition of direct government expenditure or government investment virtually disappeared during the 1980s.(45) Second voluntary agreements are emerging as a separate category of instruments. (46) Third, several of the taxonomies adopt categories primarily based on the comparison of command-and-control or regulatory instruments versus incentive-based or market-based instruments.(47) Fourth, it is common to include information, education, and research-oriented instruments with pollution regulation and incentive-based instruments.(48) Interestingly, information, education, and research are included more commonly in government studies than in the academic works.(49)

Finally, and perhaps most importantly, most of the studies treat the instruments in an ad hoc manner; they are essentially descriptive lists or menus, lacking any structure that explains the systemic differences and relationships among the instruments. The next step in developing a new framework for instrument choice is to create a taxonomy that provides the needed structure.

B. Reformulation of the Taxonomy

The discussion in the previous section illustrates several different methods for categorizing policy instruments. This section proposes a new taxonomy that will facilitate analysis of the constrained optimization formulation of the instrument choice problem.

1. Principles of a Taxonomy

A useful taxonomy reflects general principles of scientific classification, informing the user about the important similarities and differences among the various items in the classification.(50) Taxonomy is a "scheme that partitions a body of knowledge and defines the relationships among the pieces. It is used for classifying and understanding the body of knowledge."(51)

Taxonomy generally employs an organizing principle to differentiate among the elements of the classification.(52) In the taxonomy described below, the instruments are organized according to the role that the government plays in each. This approach recognizes that, across instruments, there is significant variation with respect to the type and degree of government involvement in determining the marginal cost of abatement, setting and enforcing goals, controlling how, where, and when pollution abatement will occur, and bearing the costs of pollution abatement and unabated pollution. Not only does this approach elucidate the relation among the instruments in terms of the government function, but it also provides a map that assists in the cost-minimization problem.

2. Dimension 1: Fundamental Role of Government--Entitlement Assigner vs. Regulator

Individuals or groups often make decisions to engage in activities that affect others who are not parties to those decisions. Negative environmental externalities occur when the spillover effects of the chosen actions degrade the quality of the environment, possibly leading to allocative inefficiency. The Coase Theorem states that if all parties have full information about their costs and benefits, if there are zero transaction costs, and if property rights are fully assigned and understood, then parties will bargain to allocatively efficient outcomes.(53) However, if property rights are incomplete or there are high transaction costs, Coasean bargaining may not occur.(54)

One efficiency-enhancing role for the government in environmental management, then, is to clarify the assignment of initial entitlements. Several of the lists in Table A2 include instruments that address the assignment or redefinition of property rights (55) or liability. so The government uses these instruments to address the problem of incomplete rights and responsibilities, mentioned above in the context of Coasean bargaining.

Calabresi and Melamed have provided a framework for understanding the relation between the property rights and liability approaches to controlling externalities.(57) Like Coase, Calabresi and Melamed recognize that to protect entitlements, the state must decide whom to entitle.(58) They go on to describe a second order decision that the government must make: whether to protect entitlements with property rules or liability rules.(59) Under a property rule, "someone who wishes to remove [an] entitlement from its holder must buy it from him in a voluntary transaction."(60) The current property holder has the right to refuse a sale.(61) The government role, then, is only to assign the property right and enforce it.(62) In contrast, a liability rule allows someone to "destroy the initial entitlement if he is willing to pay an objectively determined value for it."(63) Under a liability rule, the government not only assigns the initial entitlement, but determines the value that must be paid in the case of involuntary transfer.

In the context of pollution control, property and liability rules, coupled with the choice of initial entitlement, create four possible arrangements (Figure 1).(64) The first, and perhaps best recognized, is a property right that entitles a pollutee to enjoin a pollution nuisance. To avoid an injunction, a polluter must negotiate a price directly with the pollutee. The alternative arrangement, a liability rule, wherein the pollutee is the party endowed with the entitlement, requires only that the polluter compensate the pollutee for damages in an amount determined by the state. Agreement between the polluter and pollutee is not a prerequisite as it is in the case of the property rule. Alternatively, of course, the entitlement to pollute may be vested in the polluter. Here we are most accustomed to seeing a property rule, requiring the pollutee to pay a price acceptable to the polluter to effect the cessation of the offending emission.

Figure 1: Calabresi and Melamed Framework on
Property Rights and Liability Rules

                                 ENTITLEMENT

                POLLUTEE                     POLLUTER

PROPERTY        Pollutee may enjoin nui-     Polluter may pollute;
RULE            sance; polluter must pay     pollutee must pay pol-
                pollutee's price for right   luter's price to stop
                to pollute                   pollution

LIABILITY       Polluter may pollute, but    Pollutee may stop pol-
RULE            must compensate pol-         lution, but must com-
                lutee for damages            pensate polluter for loss

The three approaches above are moderately familiar to the student of law and frequently observed in practice. Calabresi and Melamed extend the range of options to include a now operatively evident, if previously overlooked, instrument based on an entitlement to pollute, protected by a liability rule.(65) They describe this design as "a kind of partial eminent domain coupled with a benefits tax."(66)

The property rights approach is most applicable when the conditions for Coasean bargaining are approximately met, i.e., when there are low transaction costs between emitters and receivers of pollution and where there is sufficient information on the marginal costs of control and environmental damage.(67) Where negotiations are not possible, either because there is a large number of parties or because the environmental injury is accidental in nature, then liability rules may be appropriate if the government is able to establish a proxy valuation of the environmental damage caused by one party against another.(68)

In many important situations, the bargaining process does entail significant transaction costs, such as coordinating parties, managing free riders, and gathering information. Additionally, the government often is not in the position to establish values to be assessed for liability by one party against another, nor to identify all parties that have been injured. So, in the real world, both approaches are often insufficient. This raises another important role for the government that of regulator.

The first step in the instrument choice process is for the government to decide whether to adopt the "Calabresian" role of assigning property rights and liability rules, or a more involved role as an initiator of environmental protection. Figure 2 depicts the first dimension along which policy instruments are distinguished: the role of government in determining the degree of environmental protection. In one role, depicted on the left-hand-side of Figure 2, the government defines property rights and liability rules and enforces their integrity when invited to do so. Here the government plays a relatively passive role. In the case of property rights, the government relies upon Coasean bargaining among individuals to determine the efficient level of environmental protection, intervening only to protect individuals' assigned rights. For liability rules, the government not only assigns entitlements but assesses values for various environmental injuries between parties, hearing cases as they are brought for adjudication.

[Figure 2 ILLUSTRATION OMITTED]

In its role as regulator, depicted on the right-hand side of Figure 2, the government intervenes when neither property rights nor liability rules are practical because of transaction costs, coordination costs, free-riding, difficulty in setting damages values, or difficulty in identifying injured parties. In this regulatory capacity, the government sets environmental quality targets, chooses instruments to accomplish those goals, monitors compliance, and initiates actions to enforce the rules.(69) This role of government as regulator of environmental quality is the primary focus of this study.

3. Instruments for Implementing Environmental Goals: The Regulator Role

The remainder of this article focuses entirely on the right-hand-side of Figure 2, the instruments that the government uses when it directly intervenes in environmental protection decisions.

a. Dimension 2: Locus of Discretion

Abating pollution requires a decision-maker to choose which abatement technologies or practices to use. In some cases the choice is obvious; in others there are many options. The second dimension for distinguishing among policy instruments, then, is the extent to which the government controls the selection of abatement practices. If the government retains control for itself, as in the case for command-and-control regulations, it employs a hierarchical approach to decision-making. In contrast, the policy tools commonly referred to as incentive-based instruments allow the polluter or private parties themselves to identify the best ways to meet pollution abatement requirements.

b. Dimension 3: Distribution of Abatement and Environmental Damage Costs

Choosing a level of environmental protection commonly involves balancing two costs--the cost of pollution abatement versus the cost of environmental damage. Unless the government chooses an extreme option, such as no abatement or complete abatement, both of these costs will be part of any final solution. The third dimension of the taxonomy addresses which, if any, of these costs will be borne by government and society as a whole and which will be borne by the polluter.

c. A Preliminary Taxonomy

Combining the second and third dimensions, it is possible to sketch a rudimentary …