Abatement-Cost Heterogeneity and Anticipated Savings from Market-Based Environmental Policies

Over the past decade, policy makers in many parts of the world have given increasing attention to market-based instruments for environmental protection, including various types of tradeable permit and charge systems (Stavins 1999). Whereas market-based environmental policy instruments were controversial just ten years ago, they have now evolved in political circles to the point of becoming a conventional wisdom, at least in the United States (Keohane, Revesz, and Stavins 1998). This change may please many economists concerned with environmental policy, but it also highlights the importance of identifying the appropriate policy instrument for each environmental problem that is faced in its particular socio-economic context. In some cases, market-based instruments may be highly desirable, but in other cases their advantages may be relatively small. Our fundamental purpose is to provide some relatively simple rules-of-thumb for policy analysts and policy makers engaged in the early stages of exploring alternative policy instruments, with the hope that such rules-of-thumb can help identify policy instruments that merit more detailed investigation.

A variety of criteria can and have been brought to bear upon the choice of policy instruments to achieve environmental goals. As the stringency of environmental targets has increased, cost-effectiveness has become a more important criterion for instrument choice.For a list of additional candidate criteria for policy instrument choice, see Bohm and Russell (1985). A key factor affecting relative aggregate abatement costs under alternative policy instruments is the heterogeneity of pollution control costs across sources. There are many reasons why the costs of complying with environmental regulations tend to be heterogeneous, including differences in plant location, size, age, and production technology. Location, for example, can affect costs due to differences in the quality and price of inputs (e.g., proximity to clean inputs), physical characteristics (e.g., urban or rural), and political jurisdiction (e.g., pre-existing regulations). While it is widely recognized that abatement-cost heterogeneity is a fundamental determinant of the potential cost-savings associated with market-based policy instruments, there is surprisingly little analysis of the relationship between the nature and magnitude of such heterogeneity and potential cost savings associated with these innovative policy instruments.

Although there is an absence of such studies, there is a relevant albeit small theoretical literature on the relationship between potential gains from trade and the underlying heterogeneity of consumer preferences and production technology. Most prominent in this literature are studies by: Weitzman (1977) and Suen (1990) on the effects of diversity in consumer preferences on the relative efficiency of the price system; Krueger and Sonnenschein (1967) on the relationship between price divergence across countries and the gains from international trade; and Mendlesohn (1986) on regulating pollution in the presence of heterogeneous benefits and costs. None provide a simple framework for directly estimating the cost-savings associated with using market-based policy instruments.

There is also a substantial empirical literature that explores the costs of using alternative environmental policy instruments for particular environmental problems.See, for example: Atkinson 1983; Atkinson and Lewis 1974; Atkinson and Tietenberg 1982, 1991; Carlson, Burtraw, Cropper, and Palmer 1997; Coggins and Swinton 1996; Gollop and Roberts 1983, 1985; Hahn and Noll 1982; Kolstad 1986; Krupnick 1986; Maloney and Yandle 1984; McConnell and Schwartz 1992; O''Neil, David, Moore, and Joeres 1983; O''Ryan 1996; Perl and Dunbar 1982; and Seskin, Anderson, and Reid 1983. These studies are concerned with specific environmental pollutants in specific contexts, ranging from dissolved oxygen in the Fox River of Wisconsin to particulates in Santiago, Chile. Each study uses different types of data at different levels of aggregation, each makes different simplifying assumptions, and each employs different methods of analysis. Although this differentiation in data and methods may be appropriate for the situations being studied, little intuition thereby emerges regarding the general relationship between the nature and magnitude of cost heterogeneity and the potential cost savings associated with cost-effective, market-based instruments.

The data requirements and analytical methods of many of the approaches utilized to compare the expected costs of alternative policy instruments render them inappropriate for the early stages of policy development. Instead, relatively simple rules-of-thumb that can be employed with minimal amounts of data may be preferable for conducting initial screenings of environmental problems, so that analysts and decision makers can focus their attention on cases where potential cost savings are greatest. Such screening is important because market-based instruments are by no means a panacea; in some cases, they hold tremendous promise of providing environmental protection cost effectively, but they are not well suited in other cases for a variety of reasons,See, for example, Hahn 1984, Misolek and Elder 1989, Malueg 1990, and Stavins 1995. and there are significant political barriers to their adoption (Keohane, Revesz, and Stavins 1998). Therefore, it is important for policy analysts and policy makers to have some idea in advance of the potential cost savings that may be associated with using a market-based instrument for a particular environmental problem.

Our approach is to develop three highly stylized models of alternative means of achieving an aggregate environmental target: two command-and-control policy instruments and one market-based instrument. A fundamental feature of the models is that a subset of their parameters represent the nature and degree of abatement cost heterogeneity. We solve each model for the aggregate abatement cost it would imply, and then by comparing results, we derive expressions for the absolute and percentage cost-savings attributed to adopting a market-based instrument - our "rules-of-thumb." Subsequently, we demonstrate the potential use of these rules-of-thumb with a specific application using readily available information.

Any empirical cost-savings potentials that are identified with the approach we offer might be thought of as lower-bound estimates, because we abstract from several other dimensions along which the costs of market-based instruments are anticipated to be less than or equal to those of command-and-control approaches. First, firms can in theory reduce emissions through three types of activities: product output reduction; input substitution; and end-of-pipe abatement. We abstract from the first type of activity by treating output as exogenous, which seems reasonable, since pollution-control costs are typically a very small fraction of total production costs (Jaffe, Peterson, Portney, and Stavins 1995).There are, of course, exceptions to this, one prominent example being carbon dioxide emissions associated with electricity production. Further, we do not differentiate between input substitution and end-of-pipe treatments, considering both as emission abatement. Hence, our stylized command-and-control policy instruments are (uniform) performance standards, not technology mandates. This is important to recognize, since even if firms were perfectly homogeneous, a true technology mandate would not be cost-effective, since it provides no latitude for firms to substitute "cleaner" inputs.

Second, our focus is exclusively on static cost-effectiveness, but it is well known that the incentive-structure of market-based instruments can lead to "dynamic cost-effectiveness," that is, decreased abatement costs over time as a result of technological innovation and diffusion (Milliman and Prince 1989; Jaffe and Stavins 1995; Jung, Krutilla, and Boyd 1996; Newell, Jaffe, and Stavins 1999). Third and finally, our analysis is partial-equilibrium in nature. In a general-equilibrium context, particular types of market-based instruments, such as those that combine revenue generation with cuts in pre-existing distortionary taxes, can enjoy additional cost advantages over commensurate command-and-control regulations (Goulder, Parry, Williams, and Burtraw 1999).

In Section 2, we develop our stylized models of three basic types of policy instruments, carry out some comparative statics analysis, and identify key general findings. In Section 3, we apply the approach to the policy problem of reducing nitrogen oxide emissions in the northeastern United States, and in Section 4, we conclude.