environmental science

Page 206 Chapter 11 Command-and-Control Strategies: The Case of Standards A command-and-control (CAC) approach to public policy is one where, in order to bring about behavior thought to be socially desirable, political authorities simply mandate the behavior in law, then use whatever enforcement machinery—courts, police, fines, and so on—that is necessary to get people to obey the law. In the case of environmental policy, the command-and-control approach consists of relying on standards of various types to bring about improvements in environmental quality. In general, a standard is simply a mandated level of performance that is enforced in law. A speed limit is a classic type of standard; it sets maximum rates that drivers may legally travel. An emission standard is a maximum rate of emissions that is legally allowed. The spirit of a standard is, if you want people not to do something, simply pass a law that makes it illegal, then send out the authorities to enforce the law. Figure 11.1 is our familiar graph showing marginal abatement costs and marginal damages related to the rate at which some production residual is emitted into the environment. Suppose that initially the actual level of effluent is at e1, a rate substantially above the efficient rate of e*. To achieve e* the authorities set an emission standard at that level; e* becomes a mandated upper limit for the emissions of this firm. The standard is then enforced by sending out whatever enforcement authorities are necessary to measure and detect any possible violations. If infractions are found, the source is fined or subject to some other penalty. If the firm reduces emissions in accordance with the standard, it will be incurring an amount equivalent to area a per year in total abatement costs. These total abatement costs are the compliance costs of meeting the standard. Standards are popular for a number of reasons. They appear to be simple and direct. They apparently set clearly specified targets. They appeal, therefore, to the sense that everybody has of wanting to come directly to grips with environmental pollution and get it reduced. Standards also appear to be congenial to our ethical sense that pollution is bad and ought to be declared illegal. The legal system is geared to operate by defining and stopping illegal behavior, and the standards approach conforms to this mind-set. Page 207 FIGURE 11.1 Emission Standards We will see, however, that the standards approach is a lot more complex than might first appear. Standards appear to offer a method to take away the freedom of sources to pollute, replacing it with mandated changes in behavior. In fact, a very practical reason for the popularity of standards is that they may permit far more flexibility in enforcement than might be apparent. What appears to be the directness and unambiguousness of standards becomes a lot more problematic when we look below the surface. Types of Standards There are three main types of environmental standards: ambient, emission, and technology. Ambient Standards Ambient environmental quality refers to the qualitative dimensions of the surrounding environment; it could be the ambient quality of the air over a particular city or the ambient quality of the water in a particular river. So an ambient standard is a never-exceed level for some pollutant in the ambient environment. For example, an ambient standard for dissolved oxygen in a particular river may be set at 3 parts per million (ppm), meaning that this is the lowest level of dissolved oxygen that is to be allowed in the river. Ambient standards cannot be enforced directly, of course. What can be enforced are the various emissions that lead to ambient quality levels. To ensure that dissolved oxygen Page 208never falls below 3 ppm in the river, we must know how the emissions of the various sources on the river contribute to changes in this measure, then introduce some means of controlling these sources. Ambient standards are normally expressed in terms of average concentration levels over some period of time. For example, the current national primary ambient air quality standard for sulfur dioxide (SO2) is 80 μg/m3 on the basis of an annual arithmetic mean and 365 μg/m3 on a 24-hour average basis.1 The standard, in other words, has two criteria: a maximum annual average of 80 μg/m3 and a maximum 24-hour average of 365 μg/m3. The reason for taking averages is to recognize that there are seasonal and daily variations in meteorological conditions, as well as in the emissions that produce variations in ambient quality. Averaging means that short-term ambient quality levels may be worse than the standard, so long as this does not persist for too long and it is balanced by periods when the air quality is better than the standard. Emission Standards Emission standards are never-exceed levels applied directly to the quantities of emissions coming from pollution sources. Emission (or effluent) standards are normally expressed in terms of quantity of material per some unit of time—for example, grams per minute or tons per week. Continuous emissions streams may be subject to standards on “instantaneous” rates of flow: for example, upper limits on the quantity of residuals flow per minute or on the average residuals flow over some time period. It is important to keep in mind the distinction between ambient standards and emission standards. Setting emission standards at a certain level does not necessarily entail meeting a set of ambient standards. Between emissions and ambient quality stands nature, in particular the meteorological and hydrological phenomena that link the two. Research to study the linkage between emission levels and ambient quality levels is an important part of environmental science. The environment usually transports the emissions from point of discharge to other locations, often diluting and dispersing them along the way. Chemical processes occur in all environmental media that often change the physical character of the pollutant. In some cases this may render the emitted substance more benign. Organic wastes put in rivers and streams will normally be subject to natural degradation processes, which will break them down into constituent elements. Thus, the ambient quality of the water at various points downstream depends on the quantity of emissions as well as the hydrology of the river: its rate of flow, temperature, natural re-aeration conditions, and so on. The link between emissions and ambient quality also can be vitally affected by human decisions. A classic case is automobiles. As part of the mobile-source air-pollution program, emission standards have been set for new cars in terms of emissions per mile of operation. But because there is no effective way of controlling either the number of cars on the roads or the total number of miles each Page 209is driven, the aggregate quantity of pollutants in the air, and thus, ambient air quality, is not directly controlled. Emission standards can be set on a wide variety of different bases. For example: Emission rate (e.g., pounds per hour). Emission concentration (e.g., parts per million of biochemical oxygen demand, or BOD, in wastewater). Total quantity of residuals (rate of discharge times concentration times duration). Residuals produced per unit of output (e.g., SO2 emissions per kilowatt-hour of electricity produced). Residuals content per unit of input (e.g., SO2 emissions per ton of coal burned in power generation). Percentage removal of pollutant (e.g., 60 percent removal of waste material before discharge). In the language of regulation, emission standards are a type of performance standard because they refer to end results that are meant to be achieved by the polluters that are regulated. There are many other types of performance standards: for example, workplace standards set in terms of maximum numbers of accidents or levels of risk to which workers are exposed. A requirement that farmers reduce their use of a particular pesticide below some level is also a performance standard, as is a highway speed limit. Technology Standards There are numerous standards that don't actually specify some end result, but rather the technologies, techniques, or practices that potential polluters must adopt. We lump these together under the heading of technology standards. The requirement that cars be equipped with catalytic converters or seat belts is a technology standard. If all electric utilities were required to install stackgas scrubbers to reduce SO2 emissions,2 these would be, in effect, technology standards because a particular type of technology is being specified by central authorities. This type of standard also includes what are often called design standards or engineering standards. There are also a variety of product standards specifying characteristics that goods must have and input standards that require potential polluters to use inputs meeting specific conditions. At the edges the difference between a performance standard and a technology standard may become blurred. The basic point of differentiation is that a performance standard, such as an emission standard, sets a constraint on some performance criterion and then allows people to choose the best means of achieving it. A technology standard actually dictates certain decisions and techniques to be used, such as particular equipment or operating practices to be used by polluters. For illustrative purposes, Exhibit 11.1 shows some typical standards, applicable in this case to snowmobiles. The carbon monoxide, hydrocarbon, and noise limits are emission standards; the limit on snowmobiles entering Yellowstone National Park can be thought of as a technology standard, as it restricts the use of certain machines in this setting. Page 210 Standards Applicable to Snowmobiles EXHIBIT 11.1 Snowmobiling has become a major wintertime activity in the United States. Historically, snowmobiles were built with two-stroke engines, the same kind of engine that has been used to power lawn mowers and outboard motors. In a two-stroke engine, fuel enters the combustion chamber at the same time that exhaust gases are expelled from it. As a result, as much as one-third of the fuel passes through the engine without being combusted. This causes poor fuel economy and high levels of emissions, particularly hydrocarbons and carbon monoxide. In one hour, a typical snowmobile emits as much hydrocarbon as a 2001 model automobile emits in 24,300 miles of driving. Snowmobiles also emit as much carbon monoxide in an hour as a 2001 model auto does in 1,520 miles of driving. They are also very noisy. In the last decade there has been a political struggle over emission standards applicable to snowmobiles. Another aspect of the fight has been over efforts to control the entrance of snowmobiles into national parks, especially Yellowstone National Park. The tabulation shows standards recently proposed by the EPA. As of now they are tied up in court battles, as the snowmobile industry regards them as too restrictive, and environmental groups regard them as not restrictive enough. Sources: James E. McCarthy, Snowmobiles: Environmental Standards and Access to National Parks, Congressional Research Service, December 3, 2007; U.S. National Park Service, The Future of Winter Use in Yellowstone National Park, October 11, 2011. Page 211 Standards Used in Combination In most actual pollution-control programs, different types of standards are used in combination. National air-pollution-control policy contains all three, as we shall see in Chapter 15. In the Total Maximum Daily Load program for water-pollution control, authorities establish ambient standards for water quality, emission standards to reduce incoming pollution loads, and technology standards in the form of best management practices. We will encounter this program in Chapter 14. The Economics of Standards It would seem to be a simple and straightforward thing to achieve better environmental quality by applying standards of various types. Standards appear to give regulators a degree of positive control to get pollution reduced, but standards turn out to be more complicated than they first appear. The discussion in the rest of this chapter will focus on the efficiency and cost-effectiveness of standards, as well as the problem of enforcement. Setting the Level of the Standard Perhaps the first perplexing problem is where to set the standard. We saw in the case of the decentralized approaches to pollution control—liability laws and property rights regimes—that there was, at least, the theoretical possibility that the interactions of people involved would lead to efficient outcomes. But with standards we obviously can't presume this; standards are established through some sort of authoritative political/administrative process that may be affected by all kinds of considerations. The most fundamental question is whether, in setting standards, authorities should take into account only damages or both damages and abatement costs. Look again at Figure 11.1, particularly at the marginal damage function. One approach in standard setting has been to try to set ambient or emission standards by reference only to the damage function. Thus, one looks at the damage function to find significant points that might suggest themselves. A principle used in some environmental laws has been to set the standard at a “zero-risk” level: that is, at the level that would protect everyone, no matter how sensitive, from damage. This would imply setting emission standards at the threshold level, labeled et in Figure 11.1. This concept is fine as long as there is a threshold. Recent work by toxicologists and other scientists, however, seems to indicate that there may be no threshold for many environmental pollutants, that in fact marginal damage functions are positive right from the origin. In fact, if we followed a zero-risk approach, we would have to set all standards at zero. This may be appropriate for some substances, certain highly Page 212toxic chemicals, for example, but it would be essentially impossible to achieve for all pollutants. The standard might instead be set at a level that accepts a “reasonably small” amount of damages, for example, e0, the point where the marginal damage function begins to increase very rapidly. Here again, however, we would be setting the standard without regard to abatement costs. A different logic might suggest that in setting the standard, damages ought to be balanced with abatement costs. This would put us squarely within the logic used in discussing the notion of economic efficiency and, in this way, lead us to set the standard at e*, the efficient emission level. Exhibit 11.2 discusses some of these issues in the context of the recent controversy about how the Environmental Protection Agency (EPA) sets ambient air quality standards. Note that there is, in effect, a certain amount of “balancing” going on when standards are set on the basis of an average over some time period. In this case short-run periods, when ambient quality is relatively low, are considered acceptable as long as they do not last “too” long. A judgment is being made, in effect, that it is not necessary to install enough abatement technology to hold ambient quality within the standard under all conceivable natural conditions. In other words, an implicit trade-off is being made between the damages that will result from the temporary deterioration of ambient quality below the standard and the high costs that would be necessary to keep ambient quality within the standard under all conditions. Uniformity of Standards A very practical problem in standard setting is whether it should be applied uniformly to all situations or varied according to circumstances. This can be illustrated by using the problem of the spatial uniformity of standards. The ambient air quality standards in the United States, for example, are essentially national. The problem with this is that regions may differ greatly in terms of the factors affecting damage and abatement cost relationships, so that one set of standards, uniformly applied across these local variations, may have serious efficiency implications. Consider Figure 11.2. It shows two marginal damage functions, one of which (labeled MDu) is assumed to characterize an urban area, whereas the other (labeled MDr) applies to a rural area. MDu lies above MDr because there are many more people living in the urban area, so the same quantity of emissions will affect the health of more people there than in the rural region. Assume that marginal abatement costs (labeled MAC) are the same in the two regions. Because the marginal damages are much higher in the urban than in the rural area, the efficient ambient level of benzene is much lower in the former than in the latter region; the efficient level is er in the rural region and eu in the urban area. Thus, a single, uniform standard cannot be efficient simultaneously in the two regions. If it is set at eu, it will be overly stringent for the rural area, and if it is set at er, it will not be tight enough for the urban region. The only way to avoid this would be to set different standards in the two areas. Of course, this confronts us with one of the great policy trade-offs: the more a policy is tailored so that it applies to different and heterogeneous situations, the more efficient it will be in terms of its impacts, but also the more costly it will be in terms of getting the information needed to set the diverse standards and enforcing them once they have been established. Page 213 The Search for an “Intelligible Principle” Setting Air Quality Standards Under the Clean Air Act EXHIBIT 11.2 Under the Clean Air Act the U.S. Congress sets emission standards for cars. But to set ambient air pollution standards for the criteria pollutants, it gives the job to the EPA; in effect, it delegates to that agency the legal power to set and enforce these standards. So in the EPA the battles take place over where the standard should be set, with people on one side saying it shouldn't be too strict and people on the other saying it should be stricter. There appears to be no major agreed-upon procedure for setting the standard. In 1999 the U.S. Court of Appeals for the District of Columbia made a novel decision: It found, at the legal request of the American Trucking Association, that Congress had engaged in an unconstitutional delegation of power to the EPA to set the standards for ozone and particulate matter. According to the U.S. Constitution, Congress is the only body that may legislate new federal statutes. Many years ago the Supreme Court decided that this power may be delegated to a regulating agency provided Congress provides that agency with an “intelligible principle” for making the decision. The trucking association and its allies asserted that the Clean Air Act (CCA) contains no such intelligible principle, hence the EPA's standard-setting was effectively illegal. Of course, the CAA does contain language containing the criteria that EPA is supposed to follow in setting the standards: It is supposed to set standards “requisite to protect the public health” with “an adequate margin of safety.” There is no mention of cost here (i.e., it doesn't say something like “tighten the standards until the added costs exceed the added benefits”), and in fact the Appeals Court concurred that cost factors could not legally be considered by the EPA. So what “intelligible principle” is the EPA supposed to follow? Numerous people have suggested alternative approaches: Significant tightening: tighten the standard until there is no significant improvement that could be gained by further improvement. Knee-of-the-curve: tighten the standard until there is a significant drop-off in added benefits from further tightening. De minimis rule: tighten the standard until further tightening would produce benefits that are too small to worry about. But do these qualify as “intelligible principles”? It would seem that whatever rule the EPA should choose, it's always going to be faced with the question of whether it would be worth it to tighten the standards a little more; in other words, will the benefits of tightening the standard exceed the cost? Despite the fact that the Clean Air Act does not contain explicit language of this kind, the U.S. Supreme Court reversed the decision of the Appeals Court, saying in effect that the EPA had, in fact, developed over the years procedures for intelligent interpretation and implementation of the health-related criterion contained in the CAA. In effect it has taken the general criteria as expressed in the law and evolved procedures for setting standards in a reasonably intelligent way. Of course, that does not necessarily satisfy the combatants in the policy process, who still fight over where the standards should be set. Page 214 FIGURE 11.2 Regional Variation in Efficiency Levels The curves in Figure 11.2 could be used to represent other heterogeneous situations as well as differences in geographical regions. For example, MDu might represent marginal damages in a particular region under some meteorological conditions, or in one season of the year, whereas MDr could represent the marginal damage function for the same area but under different meteorological conditions or at a different time of year. Now a single standard, enforced throughout the year, cannot be efficient at all points in time; if it is efficient at one time, it will not be at the other. Standards and the Equimarginal Principle Having discussed the issue of setting the standard at the efficient level of emissions, we needs to remember that the efficient level itself is defined by the minimum marginal abatement cost function. This means that where there are multiple emissions sources producing the same effluent,3 the equimarginal principle must hold. The principle states that in order to get the greatest reductions in total emissions for a given total abatement cost, the different sources of emissions must be controlled in such a way that they have the same marginal abatement costs. This means that different sources of a pollutant would normally be controlled to different degrees, depending on the shape of the marginal abatement cost curve at each source. A major problem with standards is that there is almost always an overwhelming tendency for authorities to apply the same standards to all sources. It makes their regulatory lives much simpler, and it gives the impression of being fair to everyone because all are apparently being treated alike. But identical standards will be cost-effective only in the unlikely event that all polluters have the same marginal abatement costs. Page 215 FIGURE 11.3 Marginal Abatement Costs for Two Sources Consider Figure 11.3, showing the marginal abatement cost relationships for two different sources, each emitting the same waste material. Note that the marginal abatement cost functions differ; for Firm A they increase much less rapidly as emissions are reduced than they do for Firm B. Why the difference? They may be producing different outputs with different technologies. One firm might be older than the other, and older technology may be less flexible, making it more costly to reduce emissions than at the plant with the newer equipment. One plant may be designed to use a different type of raw material input than the other. This, in fact, mirrors the situation in the real world. Normally one can expect considerable heterogeneity in abatement costs among groups of firms even though they are emitting the same type of residual. The essential numbers are summarized in Table 11.1. Uncontrolled emissions are 20 for each firm, and control costs, marginal and total, are zero. Page 216 TABLE 11.1 Illustrative Values for Polluters Shown in Figure 11.3 Under an equiproportionate rule,4 each firm would cut back by 50 percent, to 10 tons each. Marginal control costs are then unequal, $16.50 for A and $204.90 for B. Total pollution control costs are $75.90 for A and $684.40 for B, for a total of $760.30. With equimarginal cutbacks, A goes to 5 tons and B to 15 tons. Pollution control costs total $272.30 for this situation. Compliance costs for a cutback that satisfies the equimarginal criterion are only about 36 percent of compliance costs in an equiproportionate reduction. To summarize: Standards are usually designed to be applied uniformly across emission sources. This practice is almost inherent in the basic philosophy of the standards approach, and to many people this strikes them as an equitable way to proceed. But if marginal abatement costs in the real world vary across sources, as they usually do, the equal-standards approach will produce less reduction in total emissions for the total compliance costs of the program than would be achieved with an approach that satisfied the equimarginal principle. The greater the differences in marginal abatement costs among sources, the worse will be the performance of the equal-standards approach. We will see in the chapters ahead that this difference can be very large indeed. Could command-and-control-type standards be set in accordance with the equimarginal principle? Unless the applicable law required some sort of equiproportional cutback, there may be nothing to stop the authorities from setting different standards for the individual sources. To get an overall reduction to 20 tons/month in the previous example, they could require Source A to reduce to 5 tons/month and Source B to cut back to 15 tons/month. The difficult part of this, however, is that to accomplish this, the authorities must know what the marginal abatement costs are for the different sources. This point needs to Page 217be stressed. For almost any real-world pollution problem, there will normally be multiple sources. For a public agency to set individual standards in accordance with the equimarginal principle, it would have to know the marginal abatement cost relationship for each of these sources. We talked in Chapter 9 about the problem of asymmetric information.5 Polluters normally have a substantial amount of private information about pollution-control costs. These costs will usually vary among sources, so for the regulators to establish cost-effective pollution-control regulations, they will have to find some way to obtain this information. The primary source of data would have to be the polluters themselves, and there is no reason to believe they would willingly share this information. In fact, if they realize, as they certainly would, that the information would be used to establish individual source standards, they would have every incentive to provide the administering agency with data showing that their marginal abatement costs rise very steeply with emission reductions. Thus, there are real problems with authorities attempting to establish source-specific emission standards. Nevertheless, a considerable amount of this is done informally, through the interactions of local pollution-control authorities, charged with enforcing common standards, and local sources, each of which is in somewhat different circumstances. We will come back to this later when we discuss issues of enforcement. Standards and Incentives An important issue for any policy is whether it creates incentives for sources to reduce emissions to efficient levels and in cost-effective ways. The command-and-control approach based on standards is seriously deficient in this regard. A basic problem is that standards are all or nothing; either they are being met or they are not. If they are being met, there is no incentive to do better than the standard, even though the costs of further emission reductions may be quite modest. By the same token, the incentives are to meet the standards, even though the last few units of emission reduction may be much more costly than the damages reduced. It is easy to deal with the case of technology standards. Here the incentives to find cheaper ways (considering all costs) of reducing emissions are effectively zero. If control authorities dictate in detail the specific technology and practices that polluters may legally use to reduce emissions, there are no rewards to finding better approaches. Now consider emission standards. Figure 11.4 shows marginal abatement costs of a firm in two situations: MAC1 refers to such costs before a given technological improvement; MAC2 is the marginal abatement cost curve the firm could expect to have after investing some large amount of resources in an research-and-development (R&D) effort to develop better treatment or recycling technology. Without any pollution-control program at all, there is absolutely no incentive to spend the money on the R&D. But suppose the firm is now faced with having to meet emission standards of e2 tons/year. With the original marginal abatement costs the total annual cost of compliance for this firm is (a + b) per year. If the R&D program is successful, compliance costs would be only b/year. The difference, a/year, is the amount by which compliance costs would be reduced and represents, in fact, the incentive for engaging in the R&D effort. We will see in the next chapter that this is a weaker effect than is provided by economic-incentive types of programs. Nevertheless, it is an incentive, which is more than we could say for technology standards. Page 218 FIGURE 11.4 Cost Savings from Technological Change: The Case of Standards To understand fully the incentive effects of standards, one has to look closely at the details. Figure 11.4 depicts a standard applied to total emissions. Historically, most standards have been applied to emissions per unit of input or output of industrial firms. For electric utilities, an emission standard per unit of fuel burned is a standard per unit of input. There are important incentive implications of setting standards this way. Consider the following expression, showing how total emissions from an industrial operation are related to underlying performance factors: Suppose authorities apply an input standard to, for example, coal-burning power plants. The standard could be expressed in terms of maximum amounts of SO2 emissions allowed per ton of coal burned. This is a standard applied to the last term of the equation, and so the power plant will presumably reduce its emissions per unit of input to the level of the standard. But there are two other ways of reducing total emissions, as depicted in the first two terms to the right of the equals sign. Page 219One is to reduce total output through, for example, encouraging consumers to conserve electricity. The other is to reduce the amount of coal needed per unit of electricity generated, in other words, for the plant(s) to increase fuel efficiency. But the plants will have no incentive to reduce emissions in these last two ways because the standard has been written in terms only of the last factor in the expression. In recent years, regulators have been moving more toward output-based standards, that is, standards expressed in terms of allowable emissions per unit of output.6 If you multiply together the last two terms of the expression, you get emissions per unit of output. If you now place a standard on this factor, note that polluters can reduce it in two ways: by reducing inputs per unit of output and by reducing emissions per unit of input. The incentives of the polluters have been broadened. So in the case of a power plant, an output-based standard would involve both incentives: to reduce emissions per unit of coal burned (perhaps by switching to low-sulfur coal) and to become more fuel efficient (perhaps by upgrading the boilers of the plant). Political-Economic Aspects of Standards The theory of standards is that they are established by regulatory authorities, then responded to by polluters. In fact, this process can lead to patterns of political give-and-take between the parties that will substantially affect the outcome. Suppose that the authorities are making every effort to set the standard at something approaching the efficient level of emissions. In Figure 11.4, e2 is their view of the efficient level before the technical change. But the new technology lowers the marginal abatement cost curve, and we know from Chapter 5 that this will reduce the efficient level of emissions. Suppose the authorities estimate that, given their view of marginal damages, the new technology shifts the efficient emission level to e3 in Figure 11.4, and that they now change the standard to reflect this. Now the firm's compliance cost will be (b + c) per year. The difference is now (a − c). So the firm's cost savings will be substantially less than when the standard was unchanged; in fact, compliance costs may actually be higher than before the R&D program. In other words, the firm could suppose that because of the way regulators may tighten the standards, they would be worse off with the new technology than with the old methods. The standard-setting procedure in this case has completely undermined the incentive to produce new pollution-control technology. This is a case of what might be called perverse incentives. A perverse incentive is one that actually works against the objectives of the regulation. In this case, standard setting can work against long-run improvements in pollution-control technology.7 Page 220If emission standards create incentives for technological change, is it not desirable to establish very stringent standards so as to increase that incentive? This is another place where political considerations come into play. If, in Figure 11.4, the standard is set at e3 right at the beginning, this would mean cost savings of (a + d + e) with the new technology rather than just a as it would be with the standard set at e2. This type of approach goes under the heading of technology forcing. The principle of technology forcing is to set standards that are unrealistic with today's technology in the hope that it will motivate the pollution-control industry to invent ways of meeting the standard at reasonable cost. But stricter standards also create another incentive: the incentive for polluters to seek relief from public authorities by delaying the date when they become applicable. In an open political system, firms may take some of the resources that might have gone for pollution-control R&D and devote them instead to influencing political authorities to delay the onset of strict standards. The stricter and more near-term the standards, the more of this activity there is likely to be. Thus, technology forcing is another one of those strategies where the effectiveness of moderate amounts does not imply that more will be even more effective. It needs to be remembered also that to a significant extent new R&D for pollution control is carried out by a pollution-control industry rather than the polluting industries themselves. Thus, to draw conclusions about the incentives of pollution-control policy for technological change means to predict how these policies will contribute to the growth and productivity of the pollution-control industry. Technology standards are stultifying on these grounds because they substantially drain off the incentives for entrepreneurs in the pollution-control industry to develop new ideas. Emission standards are better in this respect, as we have seen. The evidence for this is the fact that representatives of the pollution-control industry usually take the side politically of stricter environmental standards; in fact, they see the fortunes of their industry tied almost directly to the degree of stringency in the emissions standards set by public authorities. The Economics of Enforcement All pollution-control programs (perhaps with the exception of voluntary programs, mentioned in the last chapter) require enforcement. Much of the ongoing political conflict over environmental regulations involves questions of enforcement, with one side often saying that it is too harsh, the other side maintaining that it isn't harsh enough. In this section we deal briefly with some economic issues related to enforcement, discussing these particularly in their relation to enforcing standards. In later chapters we will discuss enforcement issues related to other types of policy instruments. Page 221 FIGURE 11.5 The Economics of Enforcement Enforcing Emission Standards There are two primary dimensions of enforcement, monitoring and sanctioning. Consider Figure 11.5. This shows a marginal abatement cost function (MAC) representing as usual the marginal costs to the firm of reducing emissions. But on the other side, instead of a marginal damage function as in the standard model, there is a marginal penalty function. The line marked MPC represents the expected penalties that firms can be expected to face for violating an emission standard. Penalties arise when firms are detected to be exceeding their emission standard and when fines or other penalties are levied as a result. Suppose a standard is set at e*. Perhaps this was established by comparing abatement costs with damages, or perhaps on some other criterion. What is relevant here is the way firms will actually be motivated to reduce their emissions.8 MPC is zero below e*; the firm is only penalized for emissions in excess of the standard, and the shape of the MPC curve shows how penalties would increase as the size of the violation increases. If current emissions are at e0, the firm will clearly reduce them because marginal abatement costs at this point are well below the marginal penalty costs currently in effect. But it will stop reducing emissions at e1, because to go lower than this would require higher abatement costs than it would save in terms of reduced penalties. Unless something changes, therefore, the firm's emissions will end up at e1, and the amount of noncompliance will be e* − e1. The only way to reduce noncompliance is to raise the penalty function. Basically, there are two ways of doing this: raise monitoring activities so as to be Page 222better able to detect noncompliance, or raise fines for those who have been detected as noncompliant. Of course, authorities could do both and, if they did so enough, the penalty function could be raised to something like the dotted line labeled MPC’, which would then ensure a noncompliance rate of zero. This analysis shows several things. It shows the basic result that there is a trade-off in enforcement; to get higher levels of compliance authorities will normally have to devote more resources to enforcement. There may be some trade-off also between monitoring and fines. The MPC curve of Figure 11.5 can be raised or lowered by changes in either one of these enforcement elements. We should also note that monitoring in this case requires measuring, or estimating, the quantities of emissions, because the MPC function is essentially expressed in term deviations of actual emissions from the emission levels set in the standard. This deviation could apply to hourly, daily, or annual quantities. Over the years there have been major advances in monitoring technology. In the early days of rigorous pollution control, much of the enforcement effort depended on self-monitoring, where sources themselves kept the books on emissions flows over time. This permitted the agencies to visit firms periodically to audit the records at each source. Agencies could also make random checks to measure emissions. The rate of auditing and random visits could be varied according to agency budgets. More recently, technologies have been developed for undertaking continuous measurement and electronic reporting (via the Internet, for example) of emissions in some cases. The future will undoubtedly also see new developments in remote monitoring technology. The other main factor behind the MPC function of Figure 11.5 is the size of the fines or other sanctions (e.g., jail terms). Most pollution-control statutes contain provisions on the size of the fine (or jail term) that may be levied against violators if and when they are caught and found guilty. In many cases, fines have been set too low, lower than the abatement costs required to meet the standards.9 In these situations firms can actually save money by dragging their feet on compliance. With low sanctions like this, enforcement is therefore likely to be much more difficult and costly than if sanctions are higher. Sources faced with the possibility of having to pay substantially higher fines would presumably have a stronger incentive to come into compliance. It needs to be kept in mind also that sanctioning ordinarily involves using the court system to pursue legal action. The functioning of courts may put some limits on what enforcing agencies may do. For example, if monetary fines or other penalties expressed in the law are very high, courts may be reluctant to hold sources in rigorous noncompliance because of the economic dislocation this may produce. Page 223 Enforcing Technology Standards Technology standards require that sources adopt and operate approved technical means of pollution control. In this case an important distinction is between initial compliance and continued compliance. Initial compliance is where a polluter charged with meeting a particular technology standard installs the appropriate equipment. To monitor initial compliance it is necessary to have inspectors visit the site, check to see that the equipment is installed, and make sure it will operate in accordance with the conditions of the standard. Having ascertained this, the administering agency can then give the firm the necessary operating permit, but this does not ensure that the equipment will continue to be operated in the future in accordance with the terms of the permit. It may deteriorate through normal use, it may not be maintained properly, future operating personnel may not be properly trained, and so on. Without some amount of monitoring, therefore, there is no assurance that the source will continue to be in compliance. But here, again, the administering agency has great flexibility in setting up a monitoring program. It can vary from very infrequent visits to randomly selected sites all the way up to permanent observers stationed at each source. General Issues When enforcement costs are included in the analysis, the question arises of whether standards should be set, at least in part, with enforcement costs in mind. Stricter standards may involve larger enforcement costs because they require larger operating changes on the part of sources. Less strict standards may be achievable with fewer enforcement resources for the opposite reason. Public environmental agencies are usually faced with budget stringencies. In some cases, greater overall reductions in emissions may be obtained by using less strict standards that can be easily enforced than by stricter standards involving higher enforcement costs. One very common feature of environmental standards is that they are usually set and enforced by different groups of people. Standards often are set by national authorities; enforcement is usually done by local authorities. For example, the air quality standards established under the Clean Air Act are set at the federal level, but enforcement is mostly carried out by state-level agencies. This has a number of important implications. One is that standards often are set without much thought to costs of enforcement; it is more or less assumed that local authorities will find the necessary enforcement resources. Of course, this is not the case in practice. With limited enforcement budgets, local authorities may react to new programs by reducing resources devoted to other programs. Another implication is that, in practice, environmental policies incorporating standards end up having a lot more flexibility than might at first appear. Laws written at national levels are specific and apparently applicable everywhere. But at the local level, “where the rubber meets the road,” as they say, it's a matter of local pollution-control authorities applying the law to local sources, and in this process there can be a great deal of informal give-and-take between the authorities and local plant managers, with participation by local environmental Page 224groups as well. A cynic, or a political realist, might conclude that standards approaches are favored because of the very fact that in the real world of tight public agency budgets, they permit partial or incomplete compliance. One of the advantages (some might say disadvantages) of policies using standards is that they permit flexibility in enforcement. Summary The most popular approach to environmental pollution control historically has been the setting of standards. This has been called the command-and-control approach because it consists of public authorities announcing certain limits on polluters, then enforcing these limits with appropriate enforcement institutions. We specified three primary types of standards: ambient, emission, and technology. Initial discussion centered on the level at which standards should be set and the regional uniformity of standards. A leading problem with standard setting is the question of cost-effectiveness and the equimarginal principle. In most standards programs the administrative bias is to apply the same standards to all sources of a particular pollutant. But pollution control can be cost-effective only when marginal abatement costs are equalized across sources. When marginal abatement costs differ among sources, as they almost always do, uniform standards cannot be cost-effective. In practice, differences among sources in their marginal abatement costs often are recognized informally by local administrators in applying a uniform national standard. We dealt at length also with the question of the long-run impact of standards through their effects on the incentives to look for better ways of reducing emissions. Technology standards completely undermine these incentives. Emission standards do create positive incentives for R&D in pollution control, although we will see that these are weaker than those of economic-incentive types of pollution-control policies, the subject of the next two chapters. Finally, we discussed the all-important question of enforcement. Questions for Further Discussion Environmental protection programs are frequently designed to require all polluters to cut back emissions by a certain percentage. What are the perverse incentives built into this type of program? If emission standards are ruled out because of, for example, the impossibility of measuring emissions (as in nonpoint-source emissions), what alternative types of standards might be used instead? In Figure 11.2, show the social cost of setting a uniform national standard, applicable to both rural and urban areas (to do this, you can assume that the national standard is set at either eu or er). Page 225Consider the example of Figure 11.3. Suppose we define as fair a cutback in which the two sources have the same total costs. Would an equiproportionate reduction be fair in this sense? A reduction meeting the equimarginal principle? Is this a reasonable definition of fair? It is sometimes suggested that the most equitable way to resolve the trade and environment problem would be for all countries to adopt the same emission standards. What are the pros and cons of this from an economic standpoint? For additional readings and Web sites pertaining to the material in this chapter, see www.mhhe.com/field6e. 1 μg/m3 stands for micrograms per cubic meter. 2 A scrubber is a device that treats the exhaust gas stream so as to remove a substantial proportion of the target substance from that stream. The recovered material then must be disposed of elsewhere. 3 That is, in cases of “uniformly mixed” emissions. 4 An equiproportionate cutback is one that reduces each source by the same percentage of its original emissions. In the example in the text, the 10-ton cutback for each source was equal in absolute terms and also equiproportionate, as each source was assumed to be initially at an emission level of 20 tons per month. 5 See earlier discussion, p. 000. 6 Automobile emission standards have always been in terms of output, for example, grams of pollutant per mile driven. We will talk about this in Chapter 15. 7 There is another perverse incentive lurking in equiproportionate reductions. If polluters realize that they will be subject to an equiproportionate cutback in the future, it is better for them to increase their base now by increasing their emissions. When the cutback is imposed, they will be able to emit higher amounts than they would have had they not inflated their base. 8 In the last chapter we discussed liability rules as a way of controlling emissions. In penalty function terms, liability rules turn the marginal damage function into an MPC curve, because they would make firms responsible for paying for the damages their emissions produce. 9 U.S. Environmental Protection Agency, Consolidated Report on the National Pollution Discharge Elimination System Permit Enforcement Program (EPA/IG E1H28-01-0200-0100154), Washington, D.C., 1990.