Reforming Hazardous Waste Policy

PART I

Public policies for hazardous waste are an important part of environmental regulation in the United States. Complex regulations address hazardous wastes under the federal Resource Conservation and Recovery Act (RCRA). Firms face high costs to comply with these rules. The Environmental Protection Agency (EPA) estimates that complying with hazardous waste regulations will cost $32 billion in 2000, about 20 percent of the cost for all U.S. pollution control laws.1

Hazardous waste regulation, however, has not received as much critical attention as other major environmental policies. This lack of attention does not indicate success. Indeed, there are serious problems with the current policy. The regulations do not prioritize risks and thus fail to direct resources toward the worst environmental contamination. The most important example of this problem is a rule that severely restricts land disposal of all wastes, regardless of the costs and benefits. Such lack of prioritization is unfortunately common in U.S. environmental policy. The problem is particularly severe for hazardous waste regulation because of the diversity of the substances and waste management methods regulated.

In addition to their failure to prioritize risks, the current hazardous waste regulations also create strong incentives for illegally disposing of hazardous waste. Illegal disposal can damage the environment much more severely than legal waste management, so policies may even be counterproductive. Few U.S. environmental policies are easy to enforce. Hazardous waste regulations, however, may suffer especially from enforcement problems because of the harms of illegal dumping.

Although some small changes would improve the policy, fundamental reforms are necessary to address these problems fully. Such fundamental reforms include a switch to economic incentive policies. In particular, taxes on environmental releases for waste management facilities provide the most robust way to make the hazardous waste policies responsive to the differential risks from different types of wastes and methods of waste management. Better still, a modified deposit/refund policy could implement the same desirable incentives, without the enforcement problems that plague taxes as well as current regulations.

This essay begins with an overview of current hazardous waste regulations and then documents their problems. After this discussion, I propose policy reforms, beginning with a simple modification of current policies and moving on to more radical changes. The essay focuses on policies that address the current management of hazardous wastes by industrial and commercial generators. Other public policies address contamination from waste management activities after it occurs; the Comprehensive Environmental Response, Liability, and Compensation Act (CERCLA), better known as Superfund, concerns abandoned contaminated sites, and RCRA's Corrective Action program concerns contamination at active waste management sites. The problems with programs for cleanup of contaminated sites are well known; I give center stage to the less familiar hazardous waste regulations.

CURRENT POLICIES FOR HAZARDOUS WASTE

The Resource Conservation and Recovery Act regulates currently generation and management of hazardous wastes. When Congress initially passed RCRA in 1976, it gave the Environmental Protection Agency great flexibility in designing a program for hazardous waste regulation. The EPA struggled with the task. It did not issue the first set of regulations until 1980 and had not issued many permits to facilities when Congress significantly amended RCRA in 1984.2

Both the EPA's speed and approach dissatisfied Congress, which decided to take matters into its own hands. In the 1984 amendments, Congress specified extraordinarily detailed regulatory requirements and fixed timetables for the EPA to issue regulations. The changes greatly increased the stringency of the regulations but maintained much of the earlier regulatory framework.

Under RCRA, the EPA dubs a broad group of wastes hazardous and thus subject to regulation. Wastes are considered hazardous if they test positive for certain characteristics, such as being toxic or creating a risk of fire, or if the EPA explicitly lists them as hazardous waste. Facilities produced about 214 million tons of regulated wastes in 1995.

Standards for Treatment and Disposal Facilities

Wastes classified as hazardous are subject to two important types of regulations. First, the regulations set standards for facilities that manage hazardous wastes. Standards for these treatment, storage, and disposal facilities began with the interim regulations issued in 1980 and were replaced by final regulations in 1982. Although the requirements have changed over time, they continue to have the same basic structure.

For most treatment facilities, the regulations set standards for removing contaminants from the waste and for emissions. For example, incinerators--facilities that burn waste--must destroy 99.99 percent of the principal organic constituents (and a larger percentage of dioxin-bearing wastes). They must also achieve certain air emissions standards, which the 1990 Clean Air Act Amendments have strengthened. Finally, incinerators must manage their ash as a hazardous waste.

For land disposal facilities, RCRA attempts to lower the risk that wastes migrate into groundwater. Following detailed specifications by Congress in the 1984 amendments, land disposal facilities are required to have thick liners (barriers to keep contaminated liquids, called leachate, from leaving the landfill), leachate collection systems, and groundwater monitoring to make sure that contaminants have not migrated.

In addition, all treatment, storage, and disposal facilities face financial responsibility requirements. They must either carry insurance against sudden and accidental pollution incidents or provide assurances that they have adequate assets to cover such incidents. Disposal facilities must also have liability insurance for gradual releases, as well as sudden and accidental releases, and guarantee that they can finance care of the facility after it closes (such care may include monitoring the groundwater for contamination and maintaining structures that contain the waste).

How stringent are these requirements? Changes in the costs of commercial hazardous waste management give some indication of the effects of the regulations. Table 1 reports the average of low-end and high-end prices quoted by major waste management firms in 1981, 1987, and the early 1990s for various management methods. As table 1 shows, most types of waste management became much more expensive during the 1980s.

Prices for incineration, a widely used treatment method, skyrocketed during the 1980s. In 1981, incinerators sometimes paid waste generators for clean wastes with high energy content. By 1987, incinerators charged at least $400 per metric ton for these wastes. Land disposal also increased in cost, although less dramatically than incineration.

The RCRA regulations phased in during the 1980s are not solely responsible for these price changes. During the same time, communities became more aggressive in fighting local waste management facilities, making it more costly to site new facilities and expand existing ones. Waste management facilities (particularly land disposal facilities) also began to face much higher threats of litigation. In addition, the early land disposal restrictions created an increase in demand for incineration; expanded capacity in the late 1980s caused prices to settle back somewhat by the 1990s, as seen in table 1. Nonetheless, the RCRA regulations probably caused much of the cost increase.

Land Disposal Restrictions

The EPA had only begun to implement the first standards for treatment, storage, and disposal facilities when Congress decided to restrict waste management more severely. In the late 1970s, widespread concern developed about the health effects of abandoned hazardous waste sites. This concern, which also prompted the Superfund program, led Congress to tighten RCRA's regulation of land disposal. Trying to prevent a repeat of the past, it ordered the EPA to eliminate nearly all land disposal of untreated wastes.

The land disposal restrictions prohibit any land disposal--including landfilling, land treatment (spreading waste on land), surface impoundment (containing liquid waste in lagoons), and underground injection--unless facilities first treat the wastes. If possible, treatment must destroy hazardous constituents, for example, by incinerating them. When destruction of the hazardous constituents is not possible, the treatment must reduce the mobility of hazardous substances.

Frustrated with the slow initial progress on RCRA regulations, Congress gave the EPA strict deadlines to establish treatment standards for different groups of wastes. If the EPA did not keep up with Congress's schedule, "hammer" provisions would ban all land disposal even of treated wastes. Congress specified that the first regulations should address solvents and dioxin-bearing wastes by November 1986 and a group of wastes already banned from land disposal in California (known as the "California list" wastes) by June 1987. The remaining regulated wastes were divided into three groups and restricted sequentially in 1988 through 1990. Since then, the EPA has continued to issue regulations pertaining to wastes newly classified as hazardous.

Again, the commercial waste management costs in table 1 indicate the stringency of these regulations. Even with the RCRA standards for landfills in place by 1987, land disposal appears much less expensive than many alternative management methods. By precluding this option, the land disposal restrictions dramatically increased waste management costs for many facilities.

PROBLEMS WITH HAZARDOUS WASTE POLICIES

The problems with these hazardous waste regulations arise from several sources, including the political forces behind their creation, their reliance on traditional regulatory approaches, and the technical complexity of managing hazardous wastes. In this section, I discuss three major problems with the current approach: excessive emphasis on reducing land disposal, failure to prioritize risks, and incentives for illegal disposal.

Excessive Emphasis on Reducing Land Disposal

Since the 1984 amendments to RCRA, reducing land disposal of hazardous wastes has been the top priority of hazardous waste policy. The EPA has made a vast rule-making effort to specify the technology standards for the land disposal restrictions. The standards involve not only extremely technical rules but also a lengthy petition process for firms to request exemptions based on the specific attributes of their waste.

The rules successfully deflected a large amount of waste from land disposal. Data (from the reports that waste generators must submit every two years) show that disposal in landfills declined dramatically after the RCRA land disposal restrictions began in late 1986. Between 1987 and 1993, landfill disposal of the affected wastes declined by as much as 75 percent.(3)

Despite the apparent success of the program in reducing land disposal, the public health benefits may have been slim. Estimates of the public health effects of this program, derived from the analyses conducted for the EPA before the regulations went into effect, are seen in table 2. The table needs some explanation. As mentioned above, the federal land disposal restrictions took place in several stages for different types of waste. Table 2 shows the cost and health benefits for each group of wastes that became subject to the regulations at the same time. The first wastes addressed were solvents and dioxins and a group of wastes known as the "California list" wastes because California had already restricted them. The rules then divided the remaining wastes into three groups, called "first third," "second third," and "third third" wastes. In addition, regulations for surface disposal (which includes landfills and surface impoundments) are separated from later regulations for underground injection.

Both the cost and the benefit estimates in table 2 have substantial flaws. The estimates may exaggerate the costs of the policies for several reasons. They assume the preregulation quantity and mix of wastes and thus do not anticipate flexibility on the part of waste generators, who experience higher waste management costs from the regulations. To avoid these costs, firms may modify their production processes to reduce the quantity of waste they generate or make it more easily treatable. Such adjustments would make compliance with the regulations less costly than reported. In addition, the strict regulations could prompt research and development of alternative treatment technologies, potentially resulting in new, lower-cost treatments.

The land disposal restrictions also give rise to other benefits, such as avoiding ecological damage, beyond those in the table. When land disposal of hazardous waste contaminates groundwater, it may damage ecosystems if the groundwater eventually flows into surface water. The damages in the table are exclusively human health consequences from contact with contaminated groundwater.

With these caveats in mind, we still might ask whether the health risk reductions in table 2 merit the estimated costs. This assessment requires comparing health risk reductions with dollar costs, which we can do by looking at the decisions that people make about risk in their own lives. Although few people make life or death decisions based on costs, they often trade off money against small amounts of risk, for example, in choosing whether to accept a more dangerous job for a higher wage. Scaling these private trade-offs up to a 100 percent risk provides the value of a "statistical" life. Although neither the public policies nor the private trade-offs actually create a certainty of death for anyone, scaling to 100 percent probability is convenient for comparing the trade-offs across different risk levels. Typical estimates for this value range from between $3 million and $8 million per statistical life (in 1996 dollars).(4)

Comparing these values to those in table 2 suggests that some of the benefits of the surface disposal restrictions might merit their costs. If we assume that the health consequences of land disposal occur continuously over time, the surface disposal restrictions on California list waste cost about $3.6 million per adverse outcome avoided. Some of these adverse health effects are not fatal and thus probably worth less than reducing the risk of death. Nonetheless, the benefits of this regulation may exceed its costs. It is not surprising that the regulations on this group of wastes look the most favorable. Wastes fall into this group if California chose to restrict their land disposal--because they are hazardous and easy to treat--prior to the federal restrictions.

By contrast, the surface disposal regulations on the remaining wastes have high costs compared with their benefits. For example, the $1.2 billion spent on "first third" surface disposal restrictions amounts to about $240 million per cancer case avoided. In addition, the analysis mentions 455 reduced exposures annually to toxic chemicals above the "reference dose"--a maximum daily exposure to a substance "likely to be without an appreciable risk of deleterious effects during a lifetime," even for sensitive populations. Given the conservatism in defining reference doses, few of those exposed to the chemical above the reference dose will experience harmful consequences. Thus, the benefits of the "first third" regulations are unlikely to justify their costs. Similarly excessive costs per cancer case characterize the remaining "thirds."

In addition to restricting landfilling, the land disposal restrictions also sharply curtailed underground injection, in which wastewaters are injected into deep wells and salt caverns. As the table reports, no health benefits were estimated for several of these underground injection regulations. Ecological consequences of this activity are also slight because these wastes are unlikely to come into contact with groundwater or surface water. Thus, there is no gain from the millions of dollars spent on diverting wastewaters from underground injection to treatment.

Unnecessary restriction of land disposal may have harmful environmental effects. For many organic wastes, the EPA has chosen to require incineration prior to land disposal. Incineration, however, can give rise to harmful air emissions. For instance, the EPA estimates that air emissions of dioxins from hazardous waste facilities may cause 49 to 140 cases of cancer annually.(5) A few of the analyses in table 2 do consider the possibility of harms from the treatments to which waste are diverted. In all cases in which they are mentioned, the analyses conclude that these harms are small. Nonetheless, trade-offs in health harms mean that it is a mistake to treat land disposal as always the least desirable waste management method.

The overemphasis on land disposal is not the EPA's fault. Congress insisted that the EPA ban land disposal. The threat that Congress held over the agency to assure that it met the deadlines is telling: if the EPA did not issue regulations specifying suitable treatment before land disposal, then land disposal would be precluded altogether. Thus, the EPA had to act quickly to avoid even more draconian regulations.

Public concerns about contaminated sites motivated Congress's insistence on this issue. These concerns began in the 1970s with the discovery of several areas of contamination. In perhaps the most dramatic incident, residents of Love Canal (a suburb of Niagara Falls, New York) noticed discoloration of their lawns and basement walls. It turned out that the Hooker Chemical Company had used the area to dispose of chemical wastes from 1942 to 1952. The company then sold the property for a token sum to the community, which built a subdivision and school there, despite being warned of the contamination. By 1978, the plight of Love Canal residents had captured national attention, prompted in part by early reports of health harms to the community. Although later scientific research called these findings into question, the incident sparked a public outcry that led Congress to pass Superfund in 1980 and the 1984 amendments to RCRA that included the land disposal restrictions.

Allowing the EPA more discretion could have resulted in better regulations. With its greater scientific expertise, the EPA might have treated some of the early research more skeptically than the public and ultimately Congress did. Of course, given the strength of the political pressures to address contaminated sites, the EPA might also have made the same choices as Congress.

Failure to Prioritize Risks

The overemphasis on avoiding land disposal, although perhaps the single greatest mistake in the program, is symptomatic of a more general failure to prioritize risks. Another example of this failure is the lack of differentiation among wastes in terms of the hazards they pose. The program has only two categories of hazardous waste: ordinary hazardous wastes and a very small quantity of acutely hazardous waste. Any waste categorized as hazardous faces the same regulatory requirements, although wastes may vary greatly in the nature and extent of the dangers they pose.

As a consequence, the regulations do not reserve the most costly treatment for the most risky waste. Forcing a single requirement on all wastes could result in undercontrol of the most dangerous wastes and overcontrol of the least dangerous wastes. Overcontrol appears most common. Trying to maintain safety for even the most dangerous wastes has resulted in unnecessarily aggressive standards for relatively benign wastes.

The costs and benefits of the land disposal restrictions in table 2 illustrate this problem. Some of the land disposal restrictions had much more favorable cost-to-health benefit ratios than other restrictions. For the restrictions on the California list waste, the health risk reductions may merit the costs of the program. Other restrictions spend large amounts of money for negligible health improvements. Even this presentation disguises the true variation in the cost-benefit ratios for the land disposal restrictions across wastes. For the "thirds," the estimated health benefits often come from a handful of the hundreds of wastes regulated in each of these rule makings. If the program had permitted land disposal of some less hazardous waste, it might have accomplished nearly the same health benefits with much lower costs.

In addition to driving up costs, failure to prioritize wastes sacrifices the environment. The current program provides limited incentives for waste generators to shift away from especially hazardous wastes since all hazardous wastes face the same regulatory standards. If more hazardous substances had stronger controls, these differences would have encouraged waste generators to substitute away from these substances when possible in their production processes.

Incentives for Illegal Disposal

The effects of regulations depend on the enforcement threats that lie behind them. Although few environmental regulations are easy to monitor and enforce, hazardous waste regulations pose a particular problem. Waste generators can transport waste off-site for illicit disposal, making it much more difficult to detect evasion than with conventional pollutants that cannot be so easily distanced from their source.

The RCRA regulations raise the costs of legal hazardous waste management and thus provide incentives for waste generators to substitute deliberate illegal disposal. The incentives created are large, as suggested by the commercial waste disposal prices in table 1. For more direct evidence of the cost savings from illegal disposal, consider the experience of a New York police sting:

When undercover detectives, posing as illegal dumpers, went into the business of disposing toxic waste from small businesses for $40 a barrel, no questions asked, they found competition so fierce that they had to lower their price . . . One business asked for a payment plan; another insisted it was accustomed to receiving a forged disposal manifest from other illegal dumpers. (Barbanel 1992, B5)(6)

By comparison with the $40 price for illegal disposal, the cost of legal disposal averaged $568 per barrel for the wastes the detectives collected. The detectives' experience also indicates that illegal disposal services are readily available (at least in New York) and convenient.

Some of my recent research looks for evidence that these incentives increase the amount of illegal disposal.(7) I studied illegal dumping of used oil, a waste that small businesses such as auto repair shops frequently generate. Federal regulations only classify used oil as hazardous under special conditions, but many states have stepped in with regulation. Thirty-four states ban disposal of used oil, requiring that it be reused as fuel or recycled, thus raising legal waste management costs for some generators. I estimate these rules increase illegal dumping of oil by 28 percent. Because illegal dumping damages the environment much more severely than legal waste management, the regulations may actually provoke greater environmental damage than they prevent.

It is possible to avoid these problems by stepping up enforcement against violators. Increased enforcement could take two forms: high penalties for violations or intensive monitoring to be sure that violations are caught. With high enough penalties, even the slight probability of being caught in violation scares polluters into compliance and intensive monitoring is unnecessary. With a few exceptions, however, penalties are low in practice. The General Accounting Office found that penalties sometimes do not even recapture the gains from noncompliance.(8) Low penalties may be appropriate if there is a great chance of mistakenly penalizing law-abiding firms. A more likely cause for low penalties, however, is the lack of political will to inflict serious damage on firms that may be major local employers. Thus, it seems unlikely that large penalties will ever solve the enforcement problem in practice.

Given the infeasibility of high penalties, the other way to offset RCRA's incentives for illegal dumping is to increase the chance that violators are caught and face penalties. But detecting violations is difficult and costly. The government must devote resources to monitoring firms, through visiting plants and inspecting records. It also must engage in costly legal processes to get penalties assessed on violators.

Congress made this enforcement problem more difficult in 1984, when it greatly expanded the number of regulated waste generators. In its initial regulations, the EPA opted not to include wastes generated by facilities that generate under 1,000 kilograms of regulated wastes per month. In 1984, however, Congress extended most hazardous waste regulations to facilities that generate between 100 and 1,000 kilograms of hazardous waste per month, known as small-quantity generators (SQGs). In 1984, there were about 113,000 SQGs, mostly automobile repair shops.(9) (For comparison, there are fewer than 20,000 large-quantity generators). Despite the large number of SQGs, they generate a tiny percentage of total hazardous waste.

Expanding the number of generators strained the limited enforcement resources available. For example, in fiscal year 1995, state and federal authorities conducted 20,764 RCRA facility inspections.(10) These inspections might effectively monitor the 1,718 permitted treatment, storage, and disposal facilities and possibly the 19,908 large-quantity generators in that year. The more than 100,000 SQGs, however, probably face little risk of detection for RCRA violations. Evidence suggests that they take advantage of the lax enforcement; a survey of small-quantity generators in the San Francisco Bay Area found that 57 percent relied on some form of illegal waste management.(11)

In addition to direct enforcement, RCRA has another mechanism to reduce illegal disposal, its "cradle to grave" waste-tracking system. Under this system, a "waste manifest" must accompany all hazardous wastes sent off-site for treatment, recycling, storage, or disposal. When the waste reaches its final destination, this facility must send a copy of the manifest back to the waste generator and to state authorities.

There is little evidence that waste manifests have helped enforce the hazardous waste laws. Most states do not have the resources to deal with the blizzard of paper. Only a few states computerize the data and thus have any chance of catching inconsistencies. A study of state enforcement programs in the early 1980s found no instances in which manifests tipped off authorities to violations, although use of the manifests may have improved subsequently.(12) The requirements are also easy to avoid, as confirmed by the forgery that the New York police sting encountered.

Thus, stepping up existing measures to combat dumping are likely to be costly and ineffectual. A new approach to controlling hazardous wastes is necessary to avoid enforcement problems. Current regulations will continue to encourage illegal disposal and thus risk counterproductive effects on the environment.

PART II

POLICY REFORMS

Policymakers have shown little interest in modifying the RCRA program for hazardous waste management. Although Congress has repeatedly attempted major reforms of Superfund, RCRA's companion program, RCRA itself has no drawn no attention. As the discussion above suggests, however, RCRA is broken in significant ways. This section suggests ways to fix it, from simple changes to fundamental restructuring.

Relax Land Disposal Restrictions

As discussed above, RCRA devotes excessive expenditures to diverting waste from land disposal, including landfilling and underground injection. A solution to this problem does not require a radical policy redesign. Congress could eliminate these excess costs by relaxing the land disposal restrictions that it imposed in 1984.

Although it would require congressional action, relaxing the land disposal restrictions would not be complicated. The analyses behind table 2 point out the wastes for which eliminating land disposal resulted in significant health benefits. Restrictions might remain on this much narrower group of wastes, mostly included in the early land disposal bans. However, the restrictions should be lifted for remaining chemicals. The goal of eliminating land disposal may sound attractive, but it wastes resources on changes that have little clear environmental benefit.

Replace Treatment Standards with Taxes

More fundamental changes would further improve hazardous waste policies. In particular, RCRA relies too heavily on command and control to achieve its objectives. Replacing these requirements with incentives for polluters could substantially reduce the costs of the program and even improve its environmental results.

A variety of incentive-based approaches to environmental policy are available, including so-called green taxes on pollution emissions. An alternative to taxes is a marketable emission permit program, which the United States uses more extensively than emissions taxes. Under a marketable permit system, facilities may buy and sell "rights to pollute." Sometimes facilities create salable rights by reducing their emission below an allocated pollution level. Alternatively, the government may auction rights to polluters.

Implementing incentive-based policies for hazardous waste requires some thought. Hazardous wastes differ in a number of important ways from the air and water pollutants that have been the focus of most economic instruments for environmental protection. The differences will create some difficulties that the policy design should address.

First, waste generators do not release hazardous waste directly to the environment, unlike traditional pollutants. People and ecosystems feel their effects only after facilities manage wastes, whether by disposal or treatment. As a result, taxes or marketable permits should not be applied directly to waste generation, although this approach might resemble other incentive-based policies. Incentives to reduce waste generation would not address environmental releases of wastes at the management stage. Thus, current standards for treatment, storage, and disposal facilities might still need to remain in effect.

Instead, the ideal policy would create incentives for waste management facilities to reduce the contaminants they release to the environment. For example, it would elevate the costs of air emissions from incinerators and groundwater releases from landfills. The government might create these incentives either through taxes or through marketable permits; however, I argue that taxes are better and thus focus on taxes. Taxes on environmental releases could entirely replace the existing hazardous waste policy, including standards for treatment, storage, and disposal of hazardous wastes. Being taxed on their pollution would motivate management facilities to clean up. In addition, such taxes would increase the costs of waste management and thus make the firms that create waste cut back.

The disadvantage of taxes on environmental releases is that they are more difficult to enforce than both taxes on hazardous waste generation and the current regulations. To collect these taxes, the government will need to monitor air emissions and discharges into groundwater and surface water. Monitoring air and surface water releases could require costly equipment but is probably feasible for most pollutants. Monitoring releases to groundwater, however, is trickier. Although it is possible to sample groundwater and detect contamination, isolating the facility that is the source of release and quantifying the amount released (as would be necessary for a tax) often stretches current scientific understanding. As a result, enforcing these taxes is likely to be imperfect, weakening their influence. Even with spotty enforcement, however, taxes would send a clearer signal to polluters than current regulations.

A second consideration in designing incentive policies is the legal liability that facilities already face. Those who suffer from environmental damage may sue and thus give waste management facilities incentives to consider their environmental impacts. In addition, the government can sue under a program called RCRA Corrective Action Program to force facilities to pay for cleanup of environmental contamination. The EPA estimates that Corrective Action will cost $18.7 billion dollars (in discounted 1992 dollars) over the life of the program; these costs could thus be a significant factor in facilities' decisions.(13) If the expected costs from liability are at least as high the environmental costs of the release, then there is no need for a tax on top of the liability. For some releases such as air emissions, however, the injured parties are often too diffuse a group to organize a lawsuit and the release does not lead to contamination requiring cleanup. The expected costs of liability may! then fall below the environmental damage from the release. For these releases, taxes are necessary, but they should be lower than the environmental costs to reflect any costs already imposed by liability.

A third consideration in designing hazardous waste policy is the diversity of substances involved. Hazardous waste policies must control thousands of different chemicals, some generated by only a few plants. Policies should dissuade the environmental release of very hazardous substances more strongly than the release of relatively benign substances. This problem could be addressed by a system of differentiated taxes based on the hazard posed by each chemical and the environmental medium into which the chemical was released.

Setting the right taxes would be complicated because of the number of substances and the lack of toxicological information on many of them. Congress should not take on this task but rather delegate it to the EPA, which has greater access to scientific expertise and more experience with incorporating science into its decision making.

In fact, the EPA has already done much of the work needed to set the relative tax rates. It recently developed an index of the relative harm of the chemicals that firms are required to report to the Toxic Release Inventory. Although not intended for this purpose, the index could be used to set relative tax rates. Even if the judgments about relative risks were rough, any taxes that vary with toxicity would be preferable to current policy, which completely fails to differentiate degrees of hazard.

Although differentiated taxes seem the natural solution to the variety of substances, a permit market might also be feasible. Traditional permit systems have a separate market for each pollutant: for example, the air pollution market in Los Angeles, the Regional Clean Air Incentives Market (RECLAIM), has separate markets for permits for nitrogen oxides and sulfur oxides. Although this approach will not work for hazardous waste because too few facilities would participate in the market for many types of waste, the government might establish a single market for all wastes, with the required number of permits varying with the hazards of the substance. For example, the least hazardous substances could require only one permit for a one-pound release, but substances deemed twice as dangerous would require two permits for a one-pound release.

However, taxes would be preferable to this tiered-market system because they offer a better chance of getting the prices right. Only taxes can guarantee the price polluters face and thus send facilities the correct signals about environmental costs. Under the permit system, the government would have to issue just enough permits that the prices accurately reflect the true environmental costs of increases in releases. This task would require complete knowledge of the demand and supply of permits over time, a difficult proposition.(14)

The advantage of marketable permits is that they might be politically easier to adopt. The government can give permits away to polluters; when bought off in this way, polluters may not fight the environmental policy. However, it may not be necessary to go this far to recruit polluters' support. The reform proposed here--replacing burdensome regulations with taxes--should reduce the overall costs of the program to the affected industries, which would then have strong incentives to support a tax policy as an alternative to the current sysstem.

The tax system would have several benefits relative to current regulations. First, it gives waste management facilities flexibility in how they control their pollution. Current standards for treatment, storage, and disposal facilities often specify technologies that facilities must adopt. Taxes would allow them to choose alternative controls that better fit the attributes of their facility and the wastes they manage. Current regulations also set many standards that are several steps removed from environmental consequences, such as the requirement that incinerators destroy a certain percentage of organic materials. Taxes would focus facilities' resources on reducing releases to the environment rather than on such indirect goals.

Second, taxes would result in pollution control tailored to the facilities' costs of reducing pollution. Those facilities with low-cost opportunities to control releases would have the incentive to exploit all such opportunities to reduce their tax burden. Facilities without opportunities that compare favorably with the tax would choose to pay rather than reduce their releases. Thus, if the taxes were set at the right levels, they would help eliminate both the undercontrol and overcontrol of pollutants that plague the current policy. Even if taxes were set too high or too low, they would at least assure that whatever level of pollution reduction occurred was achieved with the minimum cost.

Third, the proposed tax policy would address RCRA's failure to prioritize risks. If environmental releases (such as air emissions from incinerators and groundwater contamination from landfills) were taxed at appropriate rates, then market signals would cause waste generators to substitute away from more environmentally costly waste management methods. With this market influence, the EPA would not need to establish arbitrary priorities for waste management methods, such as the land disposal restrictions.

Finally, this policy would adapt better over time than the current regulations. The relative environmental consequences of different management methods may shift over time as technologies for waste management improve and as the mix of waste needing disposal changes. Regulators cannot continuously adjust standards to adapt to these changes. There would be no need to shift taxes on environmental releases, however, if they were based on the environmental cost of the release. Shifting market signals would keep incentives for waste generators and management facilities in step with current technological conditions.

Although the federal hazardous waste program does not currently use taxes, the United States does have some experience with taxes on hazardous waste. Many states levy taxes on hazardous waste generation, management, or disposal. Indeed, these taxes provide one of the broadest and longest experiences with green taxes anywhere in the world. As discussed above, these taxes are not ideal because they do not provide incentives for reducing pollution releases at the waste management facility. Nonetheless, experience with these taxes can provide information about the success of price-based environmental policy in altering polluters' behavior.

Recent research suggests that these taxes successfully change behavior. I conducted a study of the generation and management of a major group of wastes--spent solvents from cleaning and degreasing metal products.(15) According to my research, state taxes reduce the amount of these wastes generated by manufacturing facilities. The reduction is small, between 5 and 12 percent, but the taxes are typically only $10 to $20 per ton in comparison with management costs of more than $100 a ton (see table 1). Most states tax land disposal more heavily than other forms of waste management. My research suggests that state taxes do shift waste away from land disposal, causing an 11 to 44 percent decline, depending on the estimate. These figures, however, date from the early days of the land disposal restrictions; more recent state taxes may have had a much smaller effect because federal regulations preclude most land disposal.

In addition, recent research suggests that the state taxes alter where generators manage their waste. Arik Levinson has studied the influence of state taxes on the extensive interstate shipment of waste.(16) He finds that high taxes significantly deter waste management. Whether this effect is desirable depends on the determinants of tax rates. Waste management has especially low environmental costs in some states, such as those with low population density and arid conditions. If those states choose low taxes, interstate shipment in response to differentiated taxes reduces the environmental costs of waste management. If politics rather than environmental costs drive the pattern of tax rates, however, this response to taxes is costly. Nonetheless, the evidence does illustrate that waste generators respond to price signals in ways that a good tax policy could harness.

Adopt a Deposit/Refund System

Neither of the two reforms proposed above addresses RCRA's enforcement problems. A deposit/refund approach to hazardous waste management might solve these problems. Deposit/refunds are familiar from "bottle bills," the deposits on beverage containers that many states collect from consumers and rebate when the consumer returns the container to a collection center. Because consumers who return their containers receive the full deposit back, the deposit/refund is like a tax on containers that the consumer disposes. A deposit/refund, however, does not have the incentives for illegal disposal that a direct tax on disposal would have; rather than penalizing undesirable behavior, it rewards desirable behavior.

For industrial hazardous waste, the government could charge a deposit on precursors of waste. Many wastes, such as spent solvents, begin as commercial products, so levying the deposit is not difficult. Other wastes, particularly those generated by chemical plants, have more distant precursors, so choosing the appropriate deposit is more complex. The deposit should apply to an input or mix of inputs from which the firm cannot substitute away without also reducing its waste generation.

The choice of refunds depends on the goal of the policy. If the goal is to encourage recycling in place of all waste management, then the refund might apply only to recycling. If the goal is to influence the choice among management methods, however, the government could establish differentiated refunds. For example, the facility might receive a full refund for recycled wastes and a partial refund for incineration, thus encouraging facilities to choose incineration rather than other methods such as disposal. In this way, the deposit/refund system could accomplish the same incentives as taxes on waste management, without strengthening incentives for illegal disposal. This policy would be a good candidate to replace the current land disposal restrictions: it would discourage land disposal where desirable and not encourage environmentally harmful illegal alternatives.

Like taxes on hazardous waste, however, this deposit/refund would not replace the current standards for treatment, storage, and disposal facilities. To replace those standards would require incentives to reduce releases at the waste management facility. A modified deposit/refund could accomplish this goal, while maintaining the deposit/refunds' desirable effects on enforcement. Under this modified policy, the government would charge a "deposit" but rebate it indirectly through subsidies to waste management facilities per unit they reduced their environmental releases. Don Fullerton has presented the arguments for this general approach to environmental policy, which he terms a "two-part instrument."(17)

In this plan, the subsidies provide similar incentives to taxes: if the facility can reduce pollution for less than the subsidy, it will but will forgo more costly reductions. As a result, the government should set subsidies for all releases equal to the per-unit levels of the ideal tax. With these levels, the subsidies will give rise to all the benefits of taxes touted earlier.

Unlike a tax, however, subsidies encourage waste generation because they lower the costs of waste management. That is where the deposit comes in. The deposit generally discourages waste generation by taxing the precursors to waste and thus undoes the perverse incentives created by the subsidy. At the same time, this combination encourages legal waste management rather than illegal dumping since the costs of legal management decline relative to illegal alternatives. Thus this policy has all the advantages of the taxes but is easier to enforce.

If this deposit-subsidy approach is not politically feasible, the government could still adopt a deposit/refund to address enforcement alone. This deposit/refund could supplement either the current policy or the proposed taxes on environmental releases. Under this enforcement-only deposit/refund, waste generators would pay a deposit on waste precursors as above but receive a refund for all documented legal waste management. This approach, by changing the incentives in the choice between legal and illegal waste management, would curtail illegal disposal and reduce the enforcement burden on the EPA, without the need for fundamental reform.

CONCLUSIONS

Hazardous waste policies, like other environmental policies, create many different incentives for polluters. Often these incentives are by-products of the program's direct activities. For example, current standards for waste management facilities encourage waste generators to reduce their wastes, extending the program's environmental benefits. At the same time, they create incentives for illegal disposal, which undermine the regulations by encouraging an environmentally harmful activity. Thus, these policies should be designed with attention to both the direct and the indirect incentives they create.

Indeed, good policy often strips away all but the incentives. Previous research has emphasized that incentive-based policies decentralize environmental decision making, placing the responsibility for finding the least costly ways to reduce pollution on polluters themselves rather than on regulators. But incentive policies also make the priorities of the policy transparent and flexible. For example, by taxing environmental releases from waste management facilities, the government chooses tax rates for individual types of environmental damage. As a result, the policy encourages and discourages different waste management methods on the basis of the extent of their environmental releases and the relative tax weight given to each release. Through the choice of tax rates, this policy links the priorities to environmental goals, even as technology changes over time. Under conventional policies, the priori!ties are often either obscure, as when they result from technology standards that effectively set arbitrary priorities, or inflexible, as when the government invokes the goal of eliminating land disposal regardless of costs and benefits.

It is important that any policy, whether incentive based or conventional, target the harms caused by regulated pollutants. Setting policy targets, such as reduced land disposal, that differ from the true environmental goal can result in unintended effects and unnecessary costs. Focusing on the environmental ends can help us choose policies that achieve environmental benefits while remaining consistent with other social goals.

TABLES

TABLE 1.
Commercial Hazardous Waste Management Prices (in 1996 dollars per metric ton, except transportation)

TABLE 2.
Estimated Costs and Health Effects of Land Disposal Rules

¹ Paul R. Portney, "The Economics of Hazardous Waste Regulation," mimeo, Resources for the Future, Washington, D.C., no date.

² Marc K. Landy, Marc J. Roberts, and Stephen R. Thomas, in The Environmental Protection Agency: Asking the Wrong Questions (New York: Oxford University Press, 1990), narrate the tortuous process of writing these first RCRA regulations.

Hilary Sigman, "Public Policies for Hazardous Waste and Toxic Substances," in P. R. Portney and R. N. Stavins, eds., Public Policies for Environmental Protection, 2d ed. (Washington, D.C.: Resources for the Future, forthcoming).

• W. Kip Viscusi, "The Value of Risks to Life and Health," Journal of Economic Literature 31, no. 4 (1993): 1912-46.

• U.S. EPA, Office of Air Quality Planning and Standards, Cancer Risk from Outdoor Exposure to Air Toxics, vol. 1 (Research Triangle Park, N.C.: U.S. EPA, 1990).

• Josh Barbanel, "Elaborate Sting Operation Brings Arrests in Illegal Dumping of Toxic Wastes by Businesses," New York Times, May 13, 1992, B5.

• Hilary Sigman, "Midnight Dumping: Public Policies and Illegal Disposal of Used Oil," RAND Journal of Economics 29 (1998): 157-78.

• U.S. General Accounting Office (GAO), Environmental Enforcement: Penalties May Not Recover Economic Benefits Gained by Violators (Washington, D.C.: U.S. GAO, 1991).

• Abt Associates, National Small Quantity Hazardous Waste Generator Survey (Washington, D.C.: U.S. EPA, 1985).

• U.S. Environmental Protection Agency, Enforcement and Compliance Assurance Accomplishments, FY 1995 (Washington, D.C.: U.S. EPA, 1996).

• Lorene Russell and Emy Meiorin, The Disposal of Hazardous Wastes by Small Quantity Generators: Magnitude of the Problem (Oakland, Calif.: Association of Bay Area Governments, 1985).

• U.S. General Accounting Office, Illegal Disposal of Hazardous Waste: Difficult to Detect or Deter (Washington, D.C.: U.S. GAO, 1985).

• U.S. Environmental Protection Agency, Office of Solid Waste, Draft Regulatory Impact Analysis for the Final Rulemaking on Corrective Action for Solid Waste Management Units (Washington, D.C.: U.S. EPA, 1993).

• By contrast, a marketable permit system guarantees the total quantity of releases but not the their price. Sometimes this quantity assurance is important, such as when ecological damage becomes much more severe if the pollution levels exceed some critical threshold. With the tiered permit system proposed here, however, the quantity of permits does not guarantee the amount of any given pollutant since permits trade across pollutants. Neither does it specify the level of stress on any one environmental resource because management facilities that release contaminants into different watersheds and aquifers can trade pollutants. Thus, control over quantities is not likely to be especially helpful in this context.

• Hilary Sigman, "The Effects of Hazardous Waste Taxes on Waste Generation and Disposal," Journal of Environmental Economics and Management 30 (1996): 199-217.

• Arik Levinson, "NIMBY Taxes Matter: State Taxes and Interstate Hazardous Waste Shipments," National Bureau of Economic Research Working Paper 6314, 1997.

• Don Fullerton and Ann Wolverton, "The Case for a Two-Part Instrument: Presumptive Tax and Environmental Subsidy," National Bureau of Economic Research Working Paper 5993, 1997.