research paper

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD 10.1146/annurev.publhealth.25.102802.124348

Annu. Rev. Public Health 2004. 25:119–38 doi: 10.1146/annurev.publhealth.25.102802.124348

Copyright c© 2004 by Annual Reviews. All rights reserved First published online as a Review in Advance on October 15, 2003

EMISSION TRADING AND PUBLIC HEALTH

Alexander E. Farrell1 and Lester B. Lave2 1Energy and Resources Group, University of California, Berkeley, California 94720-3050; email: [email protected] 2Graduate School of Industrial Administration and Department of Engineering and Public Policy, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213-3890; email: [email protected]

Key Words regulation, pollution, hot spots, marketable permits, incentives

■ Abstract Emission trading policies are increasingly important in environmental protection, especially in controlling air pollution in the United States. Their popularity results in part from the limitations and frustrations of centralized command and control regulation. Well-designed emission trading programs can achieve the same or better environmental and health outcomes as command and control regulations but provide regulated industries with greater flexibility. This lowers costs and improves the process of negotiating environmental policy. We present key concepts and important applica- tions of emission trading, focusing on health impacts. These programs generally are well-designed, speeding emission abatement; improving health outcomes; and reduc- ing compliance costs. Flaws have stopped a few poorly designed emission trading programs. A key issue in evaluating emission trading is our inability to know what the environmental outcome would have been under another approach.

Air Pollution and Health

Virtually all the current experience with emission trading is found in air pollution policy in the United States [see http://www.colby.edu/personal/t/thtieten/ for a lengthy bibliography (38)]. Emission trading is a form of government regulation that controls pollution but provides significant flexibility to regulated sources who buy and sell emission allowances or credits to lower the cost of reducing emissions. We begin with a brief review of a few summaries on the links between air pollution and health (8, 47, 48, 56). The pollutant posing the greatest risk to public health in the United States appears to be fine particles (<2.5 µm), although ozone and carbon monoxide (CO) also can be important. Currently, nitrogen oxides (NOx) are controlled to lower ozone formation. In the next several years, both NOx and sulfur dioxide (SO2) are expected to be controlled to lower atmospheric concentrations of fine particles. The mechanisms of action by which air pollutants cause health problems are not fully understood, especially for fine particles. Largely unknown is the importance of particle size, number, surface area, and chemical composition, as well as the effects of copollutants, which may be adsorbed onto the surface of

0163-7525/04/0421-0119$14.00 119

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

120 FARRELL ¥ LAVE

solid particles. It is not even clear if there is a threshold below which exposure has no effect on health.

Epidemiological studies around the world associate a wide range of health ef- fects with air pollution, including increased mortality, respiratory disease, hospital admissions, and emergency room visits, and reduced lung function in children. Studies examine the association of health effects with air pollution across different areas or day-to-day changes in health and pollution levels within a particular pop- ulation. Several studies have followed individuals over many years. The aggregate level of many studies engenders some skepticism that the observed associations are causal. However, because little changes from day to day in a city other than weather and air pollution levels, the association between daily levels of air pollu- tion and mortality, hospital admissions, and other health measures has convinced many public health officials that the association is causal. Because air pollution levels are much lower than during the famous episodes that resulted in numerous deaths and much of our current understanding of air pollution, there is uncertainty as to which pollutant is the culprit, making it difficult to specify cost-effective abatement policy.

In addition to these pollutants, toxicants such as mercury and persistent organic compounds (e.g., polychlorinated naphthalenes) have emerged as important health concerns. These materials can move long distances, often traveling toward the poles by natural distillation processes. Once absorbed by living creatures (metallic mercury must first be converted into methylmercury) these compounds can bio- accumulate, especially in the tissues of predatory fish. People, especially pregnant women, who consume significant amounts of fish from northern waters (including the Great Lakes) may be at risk of exposure to harmful amounts of mercury and persistent organics. Adverse effects of these compounds include neurotoxicity as well as cardiovascular problems. A recent National Academy of Sciences panel reported that “individuals with high methylmercury exposures from frequent fish consumption might have little or no margin of safety” (47, p. 7).

Table 1 presents U.S. Environmental Protection Agency (EPA) estimates of the annual health effects avoided by the 1970 Clean Air Act (CAA), providing an indication of the command and control (CAC) approach up to that point (64). In their retrospective benefit-cost analysis, the EPA found that lead emissions fell 99% and NOX emissions fell by about 20%, with the other pollutant emissions in between. The total benefits of the CAA were $22.2 trillion, with an uncertainty range from $5.6 to 49.4 trillion, and the estimated cost of control was $0.5 trillion, or 1/40th the cost.

These benefits are underestimated because the study did not estimate the health effects of air toxics, numerous health effects that could not be quantified, ecosystem damage, reduced agricultural yield, or damage to building materials (43). However, because more than 95% of the estimated benefits came from controlling gross particulate matter (up to 10µm, often called PM) and lead, costly programs that focused specifically on controlling other pollutants might have had costs greater

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

EMISSION TRADING AND PUBLIC HEALTH 121

TABLE 1 Criteria pollutant heath benefits in the United States, 1990

Annual Affected cases avoided

Endpoint Pollutant(s) population (mean, thousands)

Premature mortality PM Age 30+ 184 Lead All 22

Chronic bronchitis PM All 674

Lost IQ points Lead Children 10,400

IQ less than 70 Lead Children 45

Hypertension Lead Men 20–74 12,600

Coronary heart disease Lead Age 40–74 22

Atherothrombotic brain infarction Lead Age 40–74 4

Initial cerebrovascular event Lead Age 40–74 6

Hospital admissions PM, CO, Ozone All 209

Other respiratory-related PM, Ozone, All 88,200 ailments NO2, SO2

Restricted activity/work-loss days PM, Ozone Age 18–65 147,600 days

Source: Reference 64.

than benefits (43). Future benefits from air pollution controls are expected to be smaller, while future emission control costs may rise, since the cheapest controls were probably made first. [A somewhat controversial counterargument is that strong environmental regulations lead to greater innovations and ultimately to greater economic growth, not added costs (2).]

Although the CAC-oriented CAA has improved air quality in the United States over the past three decades, government standards for air quality have not been achieved (55, 64). The 1970 Act established deadlines for the attainment of these standards nationwide by 1975. For several pollutants, these deadlines were never achieved and were pushed back to 1987 in the 1977 Amendments and then to 2010 in the 1990 Amendments. In addition, CAC regulations contain important provisions and features that are not commonly discussed. First, it is generally possible for firms to receive waivers from some requirements if they can show the need. This problem, along with other imperfections leads to typical “rule effectiveness” values of as low as 80%, which means that a rule written to achieve, for instance, 100 tons per day of emission reductions might be expected to attain only 80. Second, the operating permits associated with CAC regulations allow the holder to continue to emit pollution forever, unless the law is changed. There is no practical difference between this arrangement and a property right to pollute. Thus when considering the health impacts of emission trading programs, it would be incorrect to assume that a perfect alternative exists.

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

122 FARRELL ¥ LAVE

Meeting air quality standards will be an enormous task, requiring further cuts in emissions of some pollutants (e.g., SO2 and NOX) by almost tenfold from current levels in the United States (64). Similar levels of control may be needed globally, especially for bio-accumulative toxics (8, 13, 56). Various international efforts designed to control such pollutants have also been developed in recent decades, although they are still in the early stages of implementation (40). Although international environmental protection actions have not tended to use emission trading or other market-based incentive mechanisms, experience in the United States may prompt imitations, particularly as the costs of emission control rise under more stringent international agreements.

Protecting public health requires further control of air pollution, but achieving the goals will not be easy, especially since the future is likely to hold higher costs and lower benefits. Similarly, mitigating climate change could have very large costs to the point that it could slow economic growth, creating both political opposition and possibly even negatively impacting public health (37). Establishing environmental policy requires political agreements about conflicting goals, which can be eased by reducing the costs, and other negative attributes such as rigidity, intrusiveness, and the prevalence of lawsuits that are associated with regulation. A key question is how to get the controls at the lowest cost and with the least amount of delay.

The Emergence of Emission Trading

Pollution control policies typically consist of overarching goals set by legislators (e.g., protect public health) and detailed rules developed by regulators to achieve those goals. Most air quality regulation uses a centralized CAC approach, such as specifying which emission control technologies different kinds of industries must use. States are then required to craft detailed implementation plans. When an individual polluter fails to comply, often it is the Department of Justice that must bring the matter to a civil or criminal court as part of an enforcement program de- signed to implement these regulations (55, 57). These choices are rightly political acts because they require balancing contradictory social goals (e.g., environmen- tal quality versus cost, simplicity versus social justice) and invoke the power of government. However, the groups making these choices face severe information limitations. For instance, government generally does not have the detailed knowl- edge to deal with the problems faced by individual emissions sources or people living in each neighborhood; government generally knows little about technologi- cal innovation that could lower the cost of pollution control. In addition, monitor- ing is expensive and necessarily incomplete, and enforcing standards is arduous and time consuming, since the courts are central to the enforcement mechanism (61).

The more detailed and specific environmental regulation is, the more these shortcomings matter. As local and state air pollution regulation advanced in the mid-twentieth century and more stringent federal regulations were developed after

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

EMISSION TRADING AND PUBLIC HEALTH 123

1970, such problems became more obvious. One particularly difficult issue was that as originally written, the CAA made no allowances for new industrial and economic activities (i.e., growth) in areas that do not meet air quality standards. As localities realized this implication of the CAA, they demanded change (5, pp. 164–206; 42).

By this time, economists had already worked out many of the key theoretical issues associated with emission trading (6, 16, 20, 68). Empirical research showed that the inefficiencies of CAC regulation were pervasive and large (3, 60). Sev- eral key concepts emerged. Most important, emission trading could provide firms with flexibility by allowing individual sources to vary the level of control emis- sions as long as the net effect was the same (or better) as that achieved by more rigid environmental regulations. This could readily save money or allow deeper emission reductions at the same cost. The simplest case involves uniformly mixed pollutants where the location of emissions did not matter. However, the possibility of spatial effects (e.g., hot spots) was clearly recognized by economists, and they developed potential solutions, such as trading in ambient permits, which accounted for location, instead of emission permits (3, 46). Other approaches have since been developed, such as zones and trading ratios (62).

Emission trading thus began with regulatory actions by the EPA in 1975, when a policy (eventually called the Emission Trading Program, or ETP) was intro- duced to allow for offsets, bubbles, netting, and banking (see Reference 23 for a summary). Trading was introduced to lower compliance costs, provide firms with flexibility, and allow for new business activity (i.e., growth) in places that had poor air quality, without compromising environmental goals. These steps lowered the level of protest by industry, thus allowing the political and legal processes that lead to emission controls to proceed more quickly.

A review of the first 10 years of this program found that it had little, if any, negative effect on environmental quality but lowered compliance costs by $1– 12 billion (33). The program likely reduced emissions below the levels required by the CAC regulations, since many transactions required a 10%–20% emission reduction for each trade. An important aspect of the ETP was that it often brought to light flaws, with the centralized CAC approach, that had been previously hidden by its complexity (22). Important issues included how to treat plant shutdowns and how to determine the baseline from which emission reductions would be calculated. Nonetheless, the ETP has been criticized as overly restrictive and thus failing to realize further gains (34, 45).

TYPES OF EMISSION TRADING

Since emission trading first emerged in the mid-1970s in the United States, two distinct types of regulatory systems called emission trading have been developed, along with a number of variations and combinations. Each requires resolution of a large number of issues associated with economic efficiency, political viability, and

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

124 FARRELL ¥ LAVE

practicality. We ignore these for the most part, focusing on health-related issues. For more detail see References 23, 61, and 62.

Emission Reduction Credits

Emission reduction credit (ERC) programs assume the existence of some prior baseline, often the previous CAC regulation, and operate by allowing firms to trade credits and thus operate above or below the baseline. ERCs can be used to avoid expensive emission reductions, or to allow for new or expanded operations. ERCs are usually denominated in terms of mass emissions per time period and are considered permanent, just as the implied property right of the underlying CAC is considered permanent. Generally, there is a tax of 10%–20% on each trade to ensure that the program creates an environmental improvement. Some key issues for ERC programs include the definition of baselines, the treatment of actions that would take place without the ERC program, administrative procedures for certifying credits, and verification of performance.

A major problem with ERCs, such as the ETP system described above, is that they generally require that credits be certified on a case-by-case basis. Regulated sources typically are required to submit detailed compliance plans explaining how they will meet these requirements. Regulators, usually at the state level, approve these plans and review and approve trades, and inspectors periodically verify com- pliance. These requirements were developed as part of the political process and were designed to ensure that the ETP resulted in real emission reductions and allowed EPA regulators to retain enforcement powers. The rule governing the ETP published in theFederal Registerran to 47 pages of fine print (51FR43814, September 1986). Another important limitation is that firms often have little or no interest in creating credits—such projects are unusual, attract unwanted attention, distract management from core issues for the firm, and must compete for attention and capital with other projects that are often more profitable. This has led some environmental organizations to worry that if too many regulated firms optimisti- cally plan to buy allowances rather than reduce emissions, shortages will occur (e.g., 31). In addition to causing some level of harm, they are also worried that if shortages were widespread enough, regulators would find it legally or politically impossible to hold accountable firms that went to the market in a good faith effort only to find it lacking.

Cap-and-Trade

Cap-and-trade (C/T) systems have been used increasingly over ERC programs. As the name implies, C/T creates a permanent limit on total emissions but allows emissions to be transferred among sources. Environmental advocates feel that setting a cap on emissions is a key virtue. The first steps in developing a C/T system are for the government to define the regulated sources and the total amount of pollution that they can emit, the cap, for each time period. Typically, the cap is set in mass units (e.g., tons), is lower than historical emissions, and declines over time.

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

EMISSION TRADING AND PUBLIC HEALTH 125

The government creates allowances equal to the size of the cap and then distributes them to the regulated sources. Allowances can be auctioned by the government to help provide price signals and stimulate emission control investment, although they are typically given for free to firms based on their historical emissions (35). C/T systems do not rely on any preexisting regulatory system (although they can be added to preexisting systems) and typically have fewer regulatory barriers (e.g., certification of credits) than do ERC programs. Another interesting feature of C/T systems is that regulators need much less information about the regulated industry in order to obtain the desired environmental and health outcomes and can focus on setting the correct cap.

The government requires regulated facilities to surrender emission allowances that equal their emissions on a regular basis (sometimes called “true up”). It will also set standards for emission monitoring and establish rules, for how allowances may be used, and enforcement measures. Because the allocation to each firm is smaller than its historical emissions, regulated firms have four basic options: (a) control emissions to exactly match their allocation; (b) undercontrol and buy allowances to cover their emissions; (c) overcontrol and then sell their excess; or (d) overcontrol and bank allowances for use in future years (when even fewer allowances will be allocated). The reason companies might buy (or sell) allowances is that facilities will have different emission control costs, or they might change operations (e.g., install newer, cleaner production equipment) so that they needed more (or fewer) allowances. Those companies with higher costs will be able to save money by undercontrolling and buying allowances from those plants with lower control costs, which make money by overcontrolling and selling.

One particularly important regulatory decision is if and how allowances can be saved (or banked) from one period to another. The economic rationale behind banking is simple, banking allows for further flexibility. The environmental effect of banking on emissions, relative to a counter-factual CAC regulation, is shown in Figure 1. Aggregate emissions move from the baseline level to a new regulated level as shown. In CAC, individual facilities all act alike, they all reduce emissions at the

Figure 1 The effect of banking.

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

126 FARRELL ¥ LAVE

same time, and thus the total effect is a step function. In a C/T system, some facilities will reduce emissions earlier and some later, so there is a smoother transition. Here, we assume that in both cases the same new regulated level is achieved and that banking induces a linear reduction. Banking provisions typically require that the number of allowances withdrawn from the bank is equal to (or less than) the number deposited. In terms of the figure, this means the size of trianglesA andB are equal or that triangleA is larger than triangleB. Thus, banking results in earlier reductions (A) but delays in reaching the new regulated level (B). Proposals have been made to allow for borrowing from future periods, but these have been rejected.

The net effect of banking emission allowance on health will vary from case to case depending on the dose-response (DR) curve. For a linear curve (or for health impacts that respond to cumulative emissions) there would be no difference for ei- ther trajectory. If the DR curve were convex (more than linear), the linear reduction trajectory due to banking would create a net benefit because the early reductions from the higher baseline have a greater effect than the slower achievement of the later reductions. If the DR curve was concave, a net disbenefit would be created for the opposite reason. If the use of emission trading with banking allowed for a lower final regulated level, a health benefit would likely occur in all cases. Gener- ally, advocates of emission trading refer only to the early emission reductions (A), whereas critics only refer to the delay in reaching the regulated level (B).

Other Approaches

In the United States, a number of other regulatory averaging programs have been implemented. These are similar to ERC programs discussed above in that they involve variations from regulatory limits that are quantified and traded. Often averaging programs are more automatic than ERC programs and do not require case-by-case certification (23).

The term discrete emission reduction (DER) is often used when sources outside of an emission trading system create credits or allowances to be sold to sources that are, in fact, regulated by an emission trading program. [Other terms for similar ideas include “opt-in,” “voluntary compliance,” “open market trading” or OMT, and in the case of greenhouse gases (GHGs) “verified emission reductions” or VERs.] Importantly, DERs are voluntary in that the nonregulated source creating the credit willingly becomes part of a regulatory program. Often, DERs are based on past reductions. These aspects of DERs make them controversial; their retrospective nature makes problems of quantifying emission baselines even more difficult, and critics charge they are also improper because they reward behavior that would have happened anyway, even without the program.

Considerable effort was made during the 1990s to develop OMT programs. The EPA went so far as to publish a proposed rule and several states initiated such programs (63). Advocates of this approach often look to create DERs in the mobile source sector (by buying and scrapping old vehicles or paying for upgrades to cleaner vehicles) to sell to stationary sources, as well as nonregulated stationary sources. These advocates emphasize the benefits of innovation by entrepreneurial

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

EMISSION TRADING AND PUBLIC HEALTH 127

firms and wish to encourage their entry into the market. However, environmental groups (including those that normally advocate for emission trading) and the EPA’s Inspector General found major flaws in the proposed EPA rule, subsequent EPA Guidance, and two state OMT programs (26, 31). The main problems are the lack of adequate baselines or credit creation protocols, the lack of verifiability of the data (i.e., a virtual honor system), the use of past emission reductions to offset future requirements, fraudulent behavior, uncertainty, liability rules, and a lack of state oversight. These problems were so severe that New Jersey closed its OMT program in late 2002, and it appears that no more will be introduced.

CRITICISMS

The main criticism of emission trading programs are that they will not effectively protect the environment and health. We show in the following five cases this criticsm is not valid except in one important example. Four other criticisms of emission trading have been raised.

First, some scholars argue that emission trading is immoral because it transforms a “wrong” into a “right,” calling allowances “medieval indulgences” or “licenses to pollute” (21, 32). These same critics, however, fail to note that the operating permits issued by environmental regulators in CAC systems have the same properties, save that they are permanently assigned to a single firm. Enabling plants to trade emissions rights lowers costs and enables regulators to tighten standards (53). Further, these critics do not acknowledge that CAC regulations allow for waivers.

Second, some critics are concerned that the greater flexibility afforded firms in emission trading systems would lead to increased environmental health risk among minority and low-income communities (14, 21, 58). This concern for en- vironmental justice arose originally with respect to the proximity of landfills and polluting industries near such communities, although there is considerable contro- versy about whether the effect exists even in this context (7, 19, 52). We review the evidence supporting this argument, where it exists, for the cases below.

The third concern is about procedural justice—neighboring communities may desire to have a say in, or at least be informed about, pollution control and emis- sion trading decisions made by nearby facilities. However, under the present CAC system, neighboring communities have little influence on the emissions of exist- ing facilities because these are generally based on control technology regulations made at the national or state level. To some degree, of course, environmental or- ganizations may act in the place of community members and thus seek to monitor these decisions. Allowing direct community oversight of emission trading deci- sions might guard against adverse health outcomes, but it would probably create barriers to emission trading and thus possibly lead to a loss of its benefits, including more healthful legislative and regulatory decisions.

Fourth, there are considerable concerns about methods of allocating emission allowances to regulated firms in C/T systems. Critics complain about both giving

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

128 FARRELL ¥ LAVE

them away for free and about auctioning or otherwise charging for them (32). Regu- lated firms, of course, argue strenuously that they deserve to be granted allowances for free, but by doing so governments create windfalls. However, allocation de- cisions are powerful tools that politicians have used to gain agreement on the introduction of emission trading programs (39).

CASES

Phasing Lead Out of Gasoline

Tetraethyl lead, a known neurotoxin, boosts gasoline octane cheaply and was used in most gasoline sold in the United States by 1970. It was first regulated in the mid- 1970s to improve the performance of the catalytic converters then being introduced to control automobile emissions. Increasing evidence of the toxic effects of this compound led to the EPA imposing much more stringent limits and eventually to ban it (49). In phasing out lead from gasoline, the EPA recognized that the cost of producing high-octane gasoline without lead varied dramatically across refineries. To reduce business protests and give flexibility to refineries, the EPA first allowed a company to average across refineries and expanded this program in 1982 to allow trading across firms. These rules allowed a refiner to sell gasoline with higher than prescribed lead content if it purchased an equivalent number of rights from other refineries that had reduced their own lead content more than the required amount. When further reductions were planned in 1986, concerns about an inability of the refinery (especially smaller firms) to install the correct technology quickly led the EPA to introduce a banking provision for 1985–1987 as well. Eventually, the EPA banned the sale of leaded gasoline.

This program was a success in many ways. It lowered industry protests, al- lowed lead to be phased out quickly, and is estimated to have reduced the cost of the program by 20% (33). The flexibility provided by the program avoided pos- sible shortages in gasoline and provided incentives for more efficient technology adoption decisions (41). The temporal pattern of emission reductions matched Figure 1: Early reductions were in 1985, but the new regulated level was not achieved until 1987. Because lead neurotoxicity (at relevant levels) is a cumu- lative phenomenon, lead banking most likely had no health effect. Because this regulation affected emissions from cars nationwide, not specific refineries or fac- tories, there seems to be no environmental justice issue.

Eliminating Ozone-Depleting Substances

To comply with the Montreal Protocol, the EPA phased out ozone-depleting sub- stances with a C/T system. This system allowed for trading among different chem- icals on the basis of their ozone-depleting potential and for international trading (61). This program succeeded in eliminating some chemicals and drastically re- ducing the use of others, helping to stop and now reverse stratospheric ozone

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

EMISSION TRADING AND PUBLIC HEALTH 129

depletion. The details of the trading system are not available to the public, so it is hard to tell what the cost savings have been, but one estimate suggests about 10% of all permits have been traded. An interesting feature of this regulation is that the initial allocation of allowances created windfalls for a small number of large producers, which led Congress to impose a tax as well. As a global phenomenon, albeit one that affects countries close to the poles (especially the South Pole) most, there seems to be no environmental justice issue.

Controlling Tailpipe Emissions

Although little known, vehicle emissions in the United States include an important emission trading component, the Mobile Source Averaging, Banking, and Trading (ABT) program (23). As for stationary sources, the 1970 CAA embodied an in- flexible approach to controlling mobile source emissions: Emission standards were set for each new car. Automakers complained bitterly about the inflexibility of the rule and pleaded for an averaging provision that would allow the high emissions of one model to be offset by the lower emissions of another. ABT provisions were first allowed for heavy-duty trucks but have since been extended to many other types of vehicles, including automobiles. Provisions for ABT have recently been introduced to reduce sulfur in gasoline and diesel fuel. Experience with the existing ABT programs indicates that averaging and banking are much more heavily used than trading (66).

The ABT programs demonstrate how the flexibility provided by well-designed emission trading programs lowers costs and permits greater health improvements. (This example is from 23.) For small marine engines, for example, the EPA set the average emission standard, in part, on the basis of marginal costs assuming emis- sion trading. If the ABT provisions had not been included (the counter-factual), the average emission standard would likely have been less stringent to accommo- date the higher costs of compliance for some manufacturers and engine families, or different standards would have been established for each engine family, which would be more costly to implement and enforce. Marine engines are distributed nationwide, so there is little likelihood that the small resulting variations around a single standard would adversely affect human health, and again there seems to be no environmental justice issue.

Controlling Acid Rain

The best-known C/T system is the EPA’s Acid Rain Program for SO2 emissions from coal-fired power plants (9, 24, 65). Key features include the strict emission monitoring provisions; the national scope of the program; the relatively deep cuts in emissions (50%); completely unrestricted trading and banking; and a small auction program in the early years of the program. The program has succeeded in a major reduction in SO2 emissions after more than a decade of political de- bate because the cost of the program was believed to be too high. The introduc- tion of emission trading as a policy tool is widely understood to have broken

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

130 FARRELL ¥ LAVE

the political logjam and allowed deep cuts in emissions to be enacted into law (5, 17).

The Acid Rain Program achieved substantial emission reductions, and because banking was permitted, the temporal pattern of emissions shown in Figure 1 again emerged. Over the period 1990–2002, SO2 emissions dropped by about one third while electricity generation had increased nationally by one third. Compared to a CAC system, C/T reduced compliance costs by about one third to half over its first five years; savings estimates are $350–$1400 million (12). Trading is unregulated (except in an accounting sense, which the EPA takes care of), monitoring require- ments are stringent and very straightforward, and enforcement mechanisms are automatic and tough. The result has been virtually perfect compliance, compared to rule effectiveness for CAC regulations that can be as low as 80%. This essen- tially eliminates the discretionary role of bureaucrats and the need for adjudication in the courts, and it also led to very small operational costs compared to CAC programs.

Unrestricted trading led to fears that trading would create hot spots detrimen- tal to public health, as noted above. However, the preexisting, health-based EPA standards for ambient SO2 concentrations remain in place and can be used to stop emission trades that would violate the standards. This demonstrates that emission trading and CAC regulations can be combined, although it is difficult to imple- ment successfully. Studies looking for evidence of environmental justice problems have found none (19, 59). In the presence of the health-based standard, it is not clear what positive findings would imply, perhaps, that although the community’s health was being adequately protected, the community (or the shareholders?) was not contributing sufficiently to the relief of acid rain. Studies that have gone be- yond merely looking at emissions to estimating health and other endpoints find that the health benefits of the Acid Rain Program are substantial and largely due to reductions in fine particles (10, 11, 15). The effects of banking are not completely clear, but because acidification is sensitive to cumulative emissions, there may be none.

Managing Photochemical Smog

Two C/T programs focus on reducing emissions of NOX in order to help manage regional photochemical smog: California’s RECLAIM system and the NOX Budget program implemented by several northeastern states (30, 44). (RECLAIM also regulates SO2 emissions.) Smog (usually measured as ozone concentration) is a challenging pollutant to regulate with an emission trading approach because it has strong spatial and temporal properties and is a key public health issue (13, 48, 50, 51). In addition, ozone concentrations tend to rise to unhealthful levels only on hot, sunny days, and pollutants that form ozone have atmospheric half-lives of only a few days. Thus, very short-term changes in emissions could have important health outcomes. Even small differences may be important because ozone concentrations are highly nonlinear functions of local NOX concentrations.

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

EMISSION TRADING AND PUBLIC HEALTH 131

RECLAIM regulates emissions from industrial sources in southern California. To control for the spatial features of the ozone problem there, the region was divided up into three zones, which were allowed only limited trading with each other. To control for the temporal features, RECLAIM does not permit banking; allowances must be used in the year they are issued or lost. However, it is not clear that prohibitions against banking adequately address this problem because of the mismatch between the periodicity of the environmental problem (several days) and the control period (one year).

When RECLAIM was first implemented in 1994, the cap was generous, allow- ing for an increase in emissions over historical levels for many sources; however, the caps declined steadily each year, aiming at an overall reduction of about 75% by 2003. Emissions in 1993–1998 did not decline nearly as fast as the cap owing to a failure of many (but not all) participants to invest in emission control equip- ment (36). Although the state regulatory agency amply warned participants of a looming problem, many firms were unwilling to take appropriate actions because of management’s failure to consider future emission allowance markets and its belief that the government would bail them out in case of serious problems.

By early 2000, prices for NOX allowances skyrocketed to over $40,000/ton, and several companies went seriously out of compliance (a first for C/T systems) and paid record fines (54). Facing significant political pressure, the state regulatory agency decided to return to a CAC approach for electric power plants.

The partial breakdown of the RECLAIM program resulted in emissions greater than were planned for, and most likely caused a health disbenefit, although we could find no estimate of the size. However, given the political situation in southern California during the 1990s (including a recession and electricity crisis), it is not clear what a CAC alternative would have been like.

Several papers harshly critical of RECLAIM appeared even before its break- down in 2000 (14, 21). These papers repeat some of the generic environmen- tal justice arguments above and examine in detail one relatively small part of RECLAIM that was problematic: mobile source emission credits. Under these pro- visions (Section 1610), DERs were generated either by purchasing and scrapping older (and presumably much dirtier than average) automobiles or by retrofitting them with emission controls. Stationary sources, in this case refineries, could buy these DERs and avoid emission controls. This feature of RECLAIM was chal- lenged in court on the grounds that it violated the Civil Rights Act of 1954 by having a disparate impact on different groups. The plaintiffs argued that a dis- parate impact occurred when region-wide emission reductions (from the mobile sources) substituted for potential localized emission reductions (at the refineries). The communities that surrounded these sites were minority and low-income, and the pollutants contained toxics as well as smog precursors. The courts overturned this provision, but not until after it had been in place for a few years. During this time, some health disbenefit undoubtedly occurred, but it is not clear how much.

The Ozone Transport Commission (OTC) NOX Budget controls emissions from industrial facilities in the states on the eastern seaboard from the District

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

132 FARRELL ¥ LAVE

of Columbia to Pennsylvania and Maine. It uses a C/T system that lowered emis- sions by 55%–65% in 1999–2002 and 65%–75% from 2003 on. To control for the spatial features of the ozone problem there were several trading zones proposed, but none were ever adopted. To control for the temporal features, the OTC NOX Budget operates only in the summertime (May through September) and features a restriction on banking called progressive flow control.

Although the OTC NOX Budget suffered some serious start-up problems, it did not collapse like part of RECLAIM, and overall emissions have been significantly reduced at what appears to be reasonable costs so far (18, 28). However, it is not clear that progressive flow control adequately addresses this problem of a mismatch between the time period of the environmental problem (2–5 day episodes) and the control period (5 months). No evidence of either spatial or temporal hot spots has been found, and no concerns about environmental justice have been raised (29). A sizeable bank was built up by the start of 2002, but it was too early to tell what health effects this will have. One possibility is that much of this bank will be lost owing to changing rules, creating a health benefit. Importantly, the OTC NOX Budget program is the first example of a multilateral emission trading program that does not rely on the authority of a central government, providing insight into how international emission trading may come about (28a).

LOOKING AHEAD

Several challenging new applications for emission trading are on the horizon, including the control of multiple pollutants and GHGs.

Multiple Pollutants

In the United States many multipollutant emission control bills have been proposed over the last few years, generally with a C/T approach (25). The reason for including multiple pollutants in one law is that this would greatly reduce the regulatory uncertainty for companies (especially electric power producers) that have been, over the last 30 years, subject to a long sequence of ever tighter CAC regulations on different pollutants. Such an approach is very expensive, especially for energy- related infrastructure that is designed (and financed) to last several decades. Capital investment in new facilities and emission control retrofits for existing facilities will be much less risky and probably less expensive for firms if they can count on predictable, even if increasingly stringent, environmental regulations. Thus, industry seems to be willing to strike a political deal in which uncoordinated and rigid CAC regulation is replaced by stricter, but more flexible, multipollutant C/T regulation that will remain fixed for an extended period (a concept called the safe harbor). As always, it is hard to evaluate these proposals because it is impossible to tell what might be feasible with a CAC counter-factual.

Major issues in considering multipollutant emission are whether to include toxic pollutants, such as mercury (Hg), and GHGs, such as CO2, as well as what level of

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

EMISSION TRADING AND PUBLIC HEALTH 133

control and predictability to include. An important consideration is that different environmental problems will almost always have different levels of consensus, both scientific and political, at any given time, so the length of time a safe harbor might last will likely be different for various pollutants. Another is that no consensus exists about whether it is appropriate to allow emission trading to apply to toxics, since the issue of hot spots may be very significant.

Greenhouse Gases

Avoiding dangerous changes to the climate system will require reducing GHG emission by approximately an order of magnitude within the next 50 years. Initial steps toward this goal are being taken by most industrialized countries (but not the United States) under the United Nations Framework Convention on Climate Change and the subsequent Kyoto Protocol. Because CO2 is the principle GHG, and because most anthropogenic CO2 comes from fossil fuel combustion, avoiding dangerous changes to the climate system will require radical changes in the world’s energy systems, a potentially costly and disruptive transformation. Thus, it is not surprising that emission trading (with its flexibility and lower costs) occupies a central place in current and expected climate policy.

As a long-lived, globally mixed pollutant, CO2 would appear to be an excellent candidate for the application of emission trading, and indeed many proposals and existing regulations take this approach. Interestingly, the European Union, which was strongly opposed to emission trading in the 1990s, changed tracks dramatically after 2000, as further analysis demonstrated just how costly implementing the Ky- oto Protocol might turn out to be (27). The current European Union plan for climate policy relies centrally on emission trading, and several such systems are already in place in the United Kingdom and Denmark. These proposals are sophisticated, and the first year of the United Kingdom’s program has proved very successful: More emission reductions than were expected were made cost-effectively.

However, there are likely to be limitations as well, particularly at the inter- national level, because international law is not very effective at promoting and protecting property rights (67). A particularly difficult example of the allocation problem exists in regard to climate change and how countries with differing lev- els of population and of industrialization can fairly be allocated CO2 emission allowances (4).

Nonetheless, so widespread is the belief that national or international law will one day result in a CO2 emission trading system that numerous firms have already started trading (1). These trades are all being made in advance of the regulatory pro- grams, so it is not clear what value they have, and as noted above, DERs and OMT programs have a very poor track record. Uncertainties include future baselines, monitoring requirements, timeframes, and other factors. Possible benefits include establishing arguments for baselines, becoming familiar with CO2 markets, and learning about the firm’s CO2 emissions and regulatory risk.

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

134 FARRELL ¥ LAVE

CONCLUSIONS

Emission trading emerged in the United States in the mid-1970s, as centralized command and control regulation of air pollution proved to have severe limitations, and it now plays a major role in improving air quality and thus human health. Emission trading concepts have been applied in other media and for other reasons, for instance to control water pollution effluent, but these cases are relatively rare.

Well-designed emission trading programs have provided flexibility to regulated firms while protecting the environment and important social values, such as fair- ness. Examples include the netting, bubbling, and banking of criteria pollutants, the phaseout of lead from gasoline, the elimination of some ozone-depleting sub- stances, the control of acid rain, and the management of regional photochemical smog. So far, and with a few minor exceptions, harmful local effects (e.g., emission hot spots) have not been found, no evidence for significant health impacts due to the banking of allowances has been found, and no environmental justice problems have been found. The exceptions are parts of the RECLAIM program in California (specifically, mobile source emission reduction credits) and Open Market Trading, such as New Jersey’s failed program.

This experience shows that emission trading programs have lowered costs, often by 20%–50%, and have allowed for political negotiations about environmental pol- icy to proceed more quickly and often to more stringent (and thus more healthful) outcomes. However, emission trading programs are not really about free markets; they typically require unequivocal, detailed rules that have to be defined carefully to ensure both that the resulting market is viable and that the desired environmen- tal outcomes are obtained, which includes ensuring that no local population faces a significant decrease (or inappropriately small improvement) in air quality. An important part of such rules is that overall pollution reduction is required, although individual firms retain flexibility in deciding how to participate in the reduction.

Environmental organizations have been important in the creation and review of emission trading programs, promoting them and stopping the poorly designed programs. However, some organizations and individuals see emission trading as unethical, despite the pragmatic advantages and despite the fact that command and control regulation has many of the same problematic features. These concerns appear to have had little effect on the use of emission trading so far.

Based on the cases reviewed here, which include almost all of the major emission trading programs in existance today, emission trading has been a major benefit to public health. As society demands still greater emission reductions in the future, achieving the goals at the lowest cost and with a minimal administrative burden becomes still more important. Recognition of the long-range transport of pollutants and global warming problem has focused a spotlight on developing nations, with demands that they reduce pollution and greenhouse gas emissions. A key question is how to get the controls at the lowest cost and with the least amount of delay. Emission trading has much to offer, as long as it is done carefully.

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

EMISSION TRADING AND PUBLIC HEALTH 135

The Annual Review of Public Healthis online at http://publhealth.annualreviews.org

LITERATURE CITED

1. Akhurst M, Morgheim J, Lewis R. 2003. Greenhouse gas emissions trading in BP. Energy Pol.31:657–63

2. Allenby B, Day R, Farrow S, Florida R, Morgenstern RD, et al. 1999. The Porter hypothesis: Can business benefit econom- ically from regulation?Environ. Forum 16:44–51

3. Atkinson SE, Lewis DH. 1974. A cost- effectiveness analysis of alternative air quality control strategies.J. Environ. Econ. Manag.1:237–50

4. Baer P, Harte J, Haya B, Herzog AV, Hol- dren J, et al. 2000. Equity and greenhouse gas responsibility.Science289:2287

5. Bailey CJ. 1998.Congress and Air Pollu- tion. New York: Manchester Univ. Press. 292 pp.

6. Baumol WJ, Oates WE. 1971. The use of standards and prices for protection of the environment.Swed. J. Econ.73:42– 54

7. Bowen W. 2002. An analytical review of environmental justice research: What do we really know?Environ. Manag.29:3– 15

8. Brunekreef B, Holgate ST. 2002. Air pollution and health.Lancet 360:1233– 42

9. Burtraw D. 1998.Cost Savings, Market Performance, and Economic Benefits of the U.S. Acid Rain Program. Washington, DC: Resour. For the Futur. (RFF)

10. Burtraw D, Krupnick A, Mansur E, Austin D, Farrell D. 1998. Costs and benefits of reducing air pollutants related to acid rain. Contemp. Econ. Pol.16:379–400

11. Burtraw D, Mansur E. 1999. Environmen- tal effects of SO2 trading and banking.En- viron. Sci. Technol.33:3489–94

12. Carlson C, Burtraw D, Cropper M, Palmer KL. 2000. Sulfur dioxide control by elec-

tric utilities: What are the gains from trade?J. Polit. Econ.108:1292–326

13. Chameides WL, Kasibhalta PS, Yienger J, Levy H. 1994. Growth of continental- scale meso-agro-plexes, regional ozone pollution, and world food production.Sci- ence264:74–77

14. Chinn LN. 1999. Can the market be fair and efficient? An environmental justice critique of emissions trading.Ecol. Law Q. 26:80–125

15. Civerolo KL, Brankov E, Rao ST, Zurbenko IG. 2001. Assessing the impact of the acid deposition control program. Atmos. Environ.35:4135–48

16. Coase RH. 1960. The problem of social cost.J. Law Econ.III:1–44

17. Cohen RE. 1995.Washington at Work: Back Rooms and Clean Air. Needham Heights, MA: Allyn & Bacon. 239 pp.

18. Commission OT. 2003. NOx Budget Program: 1999–2002 Progress Report. Ozone Transport Comm., Washington, DC

19. Corburn J. 2001. Emissions trading and environmental justice: distributive fair- ness and the USA’s acid rain programme. Environ. Conserv.28:323–32

20. Dales JH. 1968. Land, water, and owner- ship.Canad. J. Econ.1:791–804

21. Drury RT. 1999. Pollution trading and en- vironmental injustice: Los Angeles’ failed experiment in air quality policy.Duke Environ. Law Pol. Forum9:231

22. Dudek DJ, Palmisano J. 1988. Emission trading: Why is this thoroughbred hob- bled?Columbia J. Environ. Law13:217– 56

23. Ellerman AD, Joskow PL, Harrison D. 2003. Emissions Trading in the U.S.: Experience, Lessons, and Considerations for Greenhouse Gases. Pew Center on

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

136 FARRELL ¥ LAVE

Global Climate Change, Arlington, VA. http://www.pewclimate.org

24. Ellerman AD, Joskow PL, Schmalensee R, Montero J-P, Bailey E. 2000.Markets for Clean Air: The U.S. Acid Rain Pro- gram. Cambridge, UK: Cambridge Univ. Press. 362 pp.

25. Energy Inf. Adm. 2001. Analysis of Strategies for Reducing Multiple Emis- sions from Power Plants: Sulfur Dioxide, Nitrogen Oxides, Carbon Dioxide, Mer- cury, and a Renewable Portfolio Standard. Rep. SR/OIAF/2001–03, U.S. Dep. En- ergy, Washington, DC

26. EPA—Off. Insp. Gen. 2002. Open Mar- ket Trading Program for Air Emis- sions Needs Strengthening. Rep. 2002-P- 00019, USEPA, Washington, DC

27. Eur. Comm. 2001. Proposal for a Frame- work Directive for Greenhouse Gas Emissions Trading Within the Euro- pean Community—COM 2001581. Eur. Union, Brussels

28. Farrell A. 2000. The NOx budget: a look at the first year.Electr. J. March:83– 92

28a. Farrell AE. 2001. Multi-lateral emission trading: lessons from inter-state NOX con- trol in the United States.Energy Policy 29(13):1061–72

29. Farrell AE. 2003. Performance of the OTC NOx budget, 1999–2001. Carnegie Mel- lon Electr. Cent. Work. Pap. No. 03–12, ed. J Apt, p. 35. Pittsburgh

30. Farrell AE, Carter R, Raufer R. 1999. The NOx budget: costs, emissions, and imple- mentation issues.Resour. Energy Econ. 21:103–24

31. Goffman J, Dudek D. 1995.Comments on the Proposed Open Market Trading Rule. Washingtion, DC: Environ. Def. Fund

32. Goodin RE. 1994. Selling environmental indulgences.Kyklos47:573–96

33. Hahn RW, Hester GL. 1989. Marketable permits: lessons for theory and practice. Ecol. Law Q.16:36,181–91

34. Hahn RW, Hester GL. 1989. Where did

all the markets go? An analysis of EPA’s emissions trading program.Yale J. Regul. 6:109–53

35. Hausker K. 1992. The politics and eco- nomics of auction design in the market for sulfur-dioxide pollution.J. Pol. Anal. Manag.11:553–72

36. Israels K, Grow R, Campbell A, Lid- dell A, Schwarz A, Pekelney D. 2002. An Evaluation of the South Coast Air Quality Management District’s Re- gional Clean Air Incentives Market— Lessons in Environmental Markets and In- novation. U.S. Environ. Prot. Agency Reg. 9, San Francisco http://www.epa.gov/ Region9/air/reclaim/index.html

37. Jacoby H, Eckaus R, Ellerman A, Prinn R, Reiner D, Yang Z. 1997. CO2 emissions limits: economic adjustments and the dis- tribution of burdens.Energy J. 18:31– 58

38. Jarvie M, Solomon BD. 1998. Point- nonpoint effluent trading in watersheds: a review and critique.Environ. Impact Assess. Rev.18:135–57

39. Joskow P, Schmalensee R. 1998. The po- litical economy of market-based environ- mental policy: the US acid rain program. J. Law Econ.41:37–83

40. Keating TJ, West JJ, Farrell AE. 2004. Prospects for international management of inter-continental air pollution transport. In The Intercontinental Transport of Air Pollutants, ed. A Stohl. Berlin: Springer- Verlag. In press

41. Kerr S, Newell R. 2001. Policy-induced technology adoption: evidence from the U.S. lead phasedown. Washington, DC: Resour. For the Futur. (RFF)

42. Kneese AV, Schultze C. 1975.Pollution, Prices, and Public Policy. Washington, DC: Brookings Inst. 125 pp.

43. Krupnick A, Morgenstern R. 2002. The future of benefit-cost analyses of the Clean Air Act. Annu. Rev. Public Health23:427– 48

44. Lents JM, Leyden P. 1996. RECLAIM: Los Angeles’ new market-based smog

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

EMISSION TRADING AND PUBLIC HEALTH 137

cleanup program.J. Air Waste Manag. As- soc.46:195–206

45. Liroff RA. 1986. Reforming Air Pollu- tion Regulation: The Toil and Trouble of EPA’s Bubble. Washington, DC: Conserv. Found. 186 pp.

46. Montgomery WD. 1972. Markets in li- censes and efficient pollution control pro- grams.J. Econ. Theory5:395–418

47. Natl. Res. Counc. 2000.Toxicological Ef- fects of Methylmercury. Washington: Natl. Acad. Press. 304 pp.

48. Natl. Res. Counc. Comm. Tropospheric Ozone Form. Meas. 1991.Rethinking the Ozone Problem in Urban and Regional Air Pollution. Washington, DC: Natl. Acad. Press. 500 pp.

49. Nichols A. 1997. Lead in gasoline. InEco- nomic Analyses at EPA: Assessing Regu- latory Impact, ed. RD Morgenstern, pp. 49–89. Washington, DC: Resour. For the Futur. (RFF)

50. Nobel CE, McDonald-Buller EC, Kimura Y, Allen DT. 2001. Accounting for spa- tial variation of ozone productivity in NOx emission trading.Environ. Sci. Technol. 35:4397–407

51. Nobel CE, McDonald-Buller EC, Kimura Y, Lumbley KE, Allen DT. 2002. In- fluence of population density and tem- poral variations in emissions on the air quality benefits of NOx emission trading. Environ. Sci. Technol.36:3465– 73

52. Perlin SA, Setzer RW, Creason J, Sex- ton K. 1995. Distribution of industrial air emissions by income and race in the United States—an approach using the toxic release inventory.Environ. Sci. Tech- nol. 29:69–80

53. Pezzey J. 1992. The symmetry between controlling pollution by price, and con- trolling it by quantity. Can. J. Econ.25: 983–91

54. Polakovic G. 2000. DWP to pay pollu- tion penalty of $14 million; smog: the util- ity also agrees to slash emissions. Dirtier plants had been pressed into service dur-

ing heat wave. InLos Angeles Times, pp. B1. Los Angeles

55. Portney PR, Stavins RN, eds. 2000.Pub- lic Policies for Environmental Protection. Washington, DC: Resour. For the Futur. (RFF) 294 pp.

56. Rabl A, Spadaro JV. 2000. Public health impact of air pollution and implications for the energy system.Annu. Rev. Energy Environ.25:601–27

57. Raufer R. 1996. Market-based pollution control regulation: implementing eco- nomic theory in the real world.Environ. Pol. Law26:177–84

58. Solomon B, Lee R. 2000. Emissions trad- ing systems and environmental justice: Why are some population segments more exposed to pollutants than others?Envi- ronment42:32–45

59. Swift B. 2001. How environmental laws work: an analysis of the utility sector’s response to regulation of nitrogen ox- ides and sulfur dioxide under the Clean Air Act. Tulane Environ. Law J.14:309– 425

60. Teller A. 1967. Air pollution abate- ment: economic rationality and reality. Deadalus96:1082–98

61. Tietenberg T. 2001. Introduction to vol- ume 1. InEmissions Trading Programs, ed. T Tietenberg, pp. xi–xxviii. Burling- ton, VT: Ashgate

62. Tietenberg T. 2001. Introduction to vol- ume 2. InEmissions Trading Programs, ed. T Tietenberg, pp. xv–xxxiv. Burling- ton, VT: Ashgate

63. US Environ. Prot. Agency. 1995. Pro- posed Rule: Open Market Trading Rule for Ozone Smog Precursors. 60Fed. Reg- ist. 39668–94. Aug. 3

64. US Environ. Prot. Agency. 1997.The Ben- efits and Costs of the Clean Air Act, 1970 to 1990.Washington, DC: USEPA

65. US Environ. Prot. Agency. 1999. Acid Rain Program 1998 Compliance Report. Rep. EPA-430-R-99–010, Acid Rain Div., Washington, DC

66. US Environ. Prot. Agency. 2001. The

13 Feb 2004 14:32 AR AR209-PU25-06.tex AR209-PU25-06.sgm LaTeX2e(2002/01/18)P1: IBD

138 FARRELL ¥ LAVE

United States Experience with Eco- nomic Incentives for Protecting the En- vironment. Off. Pol., Econ., Innov., Washington, DC

67. Victor DG. 2000.The Collapse of the

Kyoto Protocol and the Struggle to Slow Global Warming. Princeton, NJ: Prince- ton Univ. Press

68. Weitzman ML. 1974. Prices vs. quantities. Rev. Econ. Stat.41:477–92