New Approaches to Environmental Regulation
Transforming Environmental Regulation
New, more flexible approaches show great promise, but barriers remain.
The new Bush administration has within its reach the tools to implement a new environmental agenda: one that will address serious problems beyond the reach of traditional regulatory programs and will reduce the costs of the nation's continuing environmental progress. Christine Todd Whitman could be the Environmental Protection Agency (EPA) administrator who will transform regulatory programs and the agency itself for the 21st century.
Doing so will require continuing the shift away from end-of-the pipe technology requirements and toward whole-facility environmental management and permitting; expanding cap-and-trade systems to drive down pollution and pollution prevention costs; and implementing performance requirements for facilities, whole watersheds, and even states. The hallmark of the new approach is the creation of incentives for technological innovation, for civic involvement and collaboration, and for place-specific solutions.
Whitman and the EPA do not have to invent these approaches from scratch. Innovators within the EPA and the states--including Whitman's home state of New Jersey--have been pushing the frontier forward for a decade or more. Some of those innovations have proved themselves, demonstrating that the nation will be able to make progress against some of its most daunting environmental problems, including nonpoint water pollution, smog, and climate change. Traditional regulatory programs will not be able to solve those problems. Transforming environmental protection is a prerequisite for delivering the kind of environment that Americans want.
Improving the environment is one of the issues on which President Bush could indeed show himself to be uniter. Environmental policy was deadlocked in partisan wrangling for most of the 1990s. It need be no longer. In her first formal remarks to the Senate Environment and Public Works Committee as part of her confirmation hearing in January 2001, Whitman began to frame an agenda that could gather bipartisan support. The agenda is also consistent with many of the central recommendations in the National Academy of Public Administration's (NAPA's) recent report, Environment.Gov. The report was based on a three-year evaluation by a distinguished NAPA panel of the most promising innovations in environmental protection at the local, state, and federal level.
Whitman told the Senate committee that the Bush administration "will maintain a strong federal role, but we will provide flexibility to the states and to local communities. . . . [W]e will continue to set high standards and will make clear our expectations. To meet and exceed those goals, we will place greater emphasis on market-based incentives. . . . [W]e will work to promote effective compliance with environmental standards without weakening our commitment to vigorous enforcement of tough laws and regulations."
Whitman's framework for action is sound. Her emphasis on flexibility and the use of market-based tools makes sense, but only because she has coupled it with the promise of maintaining and enforcing strong federal standards and enhancing environmental monitoring. Whitman described her environmental accomplishments in New Jersey not in terms of the dollars she had spent or the number of violators she had prosecuted, but in terms of specific reductions in ozone levels, increases in the shad population, and the expansion of areas open to shellfish harvesting. She asserted a need for more of the kind of monitoring and measurement that allowed her to make such claims: "Only by measuring the quality of the environment--the purity of the water, the cleanliness of the air, the protection afforded the land--can we measure the success of our efforts," she said.
Without improved monitoring, more flexible approaches to regulation will be technically flawed and politically unworkable. (Democrats and environmentalists won't buy them.) Without more flexibility, however, new reductions in pollution levels will appear to be too expensive. (Republicans and business interests won't buy them.) Progress will depend on Whitman's ability to persuade Congress and the rest of the United States that her vision of regulatory reform will improve the environment. A significantly enhanced monitoring capacity and the institutional resources to gather, analyze, and disseminate the results to the public must be integral parts of the reform agenda.
Changing the basis of regulation
Whitman's list of principals, like her predecessor's mantra of "cleaner, cheaper, smarter," lays out the challenge: finding ways to improve the environment by reducing the constraints on regulated entities. The key is shifting the basis of the relationship between the regulator and the regulated from static technology-based permits to dynamic agreements that reward improving environmental performance and hence inspire pollution prevention and technological innovation.
The EPA and state and local environmental organizations have been experimenting with various regulatory reforms intended to achieve this shift. Some have demonstrated their potential; others have shown how difficult it is to shift to a performance focus within EPA's existing statutory framework. Among the approaches studied by the NAPA project's 17 independent research teams, the most promising include a self-certification program in Massachusetts; whole-facility permitting, pioneered in New Jersey and now being adapted by several states; emissions caps, also widely used; and allowance trading systems, which have demonstrated their effectiveness with several air pollutants and could be deployed to reduce nutrients in watersheds. The EPA and Congress should take steps to remove institutional and statutory barriers to their broader implementation.
The Massachusetts Environmental Results Program (ERP) has begun to make progress in reducing the environmental impacts of small businesses in a way that appears to be cost-effective and transferable to other states. Small businesses such as small farmers and other sources of nonpoint pollution have proved extremely difficult to regulate with traditional permits. There are too many, and each is too small to warrant the kind of time-intensive applications, reviews, and inspections that accompany most traditional environmental permits. The Massachusetts Department of Environmental Protection (DEP) sought a way to bring small operations into its regulatory system without permits and to drive improvements in their environmental performance without protracted litigation. It has succeeded.
Susan April and Tim Greiner of the consulting firm Kerr, Greiner, Anderson, and April evaluated the program for NAPA and concluded that ERP has greatly increased the number of small businesses in three sectors (printing, dry cleaning, and photo processing) that are on record with the state's regulatory system and thus are likely to be responsive to state requirements. ERP requires an individual in each firm to certify in writing each year that his or her business is in compliance with a comprehensive set of environmental regulations. The department has provided businesses in each sector with workbooks to guide managers through the steps needed to achieve compliance. In some cases, self-certification replaces state environmental permits. To ensure that participants take the self-certification seriously, the DEP enforcement staff inspects a percentage of the participating firms.
Most of the facilities in the three business sectors involved in ERP had been virtually invisible to the department. As part of the process of creating the workbooks and certification plans, however, DEP engaged trade associations and other stakeholders in an extensive process of technical collaboration and negotiations. The trade associations helped DEP build a registry of their members. Before ERP, the state was aware of only 250 printers; through ERP, it identified 850 more. Dry cleaners on record with DEP expanded from 30 to 600; the number of photo processors grew from 100 to 500.
DEP estimates that because of ERP, printers have eliminated the release of about 168 tons of volatile organic compounds statewide each year, and dry cleaners have reduced their aggregate emissions of perchloroethylene, a hazardous air pollutant, by some 500 tons per year. Photo processors were expected to reduce their discharges of silver-contaminated wastewater.
The DEP was sufficiently pleased with ERP's success in the three initial sectors that it was moving ahead last year with the development of a certification program for some 8,000 dischargers of industrial wastewater, for thousands of gas stations responsible for operating pumps with vapor-recovery systems, and for thousands of other firms installing or modifying boilers. Massachusetts and Rhode Island were jointly developing regulations and workbooks to apply to auto body shops in both states.
ERP could be adopted on a broader scale in many states to bring tens of thousands of firms into compliance with state standards. The approach could even be modified to reduce agricultural sources of nutrient runoff, where part of the regulatory challenge is finding a way to bring many relatively small operations into a management program or trading system without creating huge new transaction costs.
ERP also demonstrated one of the challenges facing Whitman and others as they seek flexible yet enforceable programs. When Massachusetts attempted to tweak the requirements for dry cleaners in a way that would have conflicted with recordkeeping requirements in the federal Clean Air Act, the EPA and the state found themselves at loggerheads. This seemed to be just the kind of problem that the EPA had in mind when it started a regulatory reinvention program called Project XL, which was intended to encourage innovation by rewarding excellent environmental performance with greater flexibility. Massachusetts and the EPA signed an agreement making ERP an XL pilot, and many within the EPA were enthusiastic supporters of the state's specific proposal. But the EPA ultimately decided that it lacked statutory authority to alter the recordkeeping requirements and quashed the state's alternative approach. As a result, the state now applies ERP only to operations that require no federal permits.
Although Whitman pledged to give states more flexibility in designing and managing programs, the ERP case demonstrates that doing so in any comprehensive way will require congressional authorization. Whitman and Congress should move quickly to secure more discretion for the administrator to approve state experiments in regulatory reform.
Focusing on performance
New Jersey's facility-wide permitting program (FWP) ran through most of the 1990s and demonstrated some of the challenges and opportunities inherent in trying to regulate large facilities in a comprehensive multimedia approach. Those lessons help inform the latest efforts underway in states: the development of performance-track agreements.
Each of New Jersey's 12 completed FWPs consolidated between 12 and 100 air, water, and waste permits into a single FWP. Previously, some factories had separate permits for each of dozens of air pollution sources. The facility-wide permit first aggregated those sources into separate industrial processes within the facility, and then generally set an air emissions cap on each process. Those caps allow firms to "trade" reductions within their facilities. Ten of the 12 FWP facilities reported that the program's biggest benefit was operational flexibility: Authorization was no longer needed to install new equipment or change processes, provided that the changes did not increase the waste stream or exceed permitted emission levels.
Susan Helms and colleagues at the Tellus Institute evaluated the program for NAPA and found that the intensive review required to prepare the facility-wide permits improved both the regulators' and plant managers' understanding of the plants and their systems. Indeed, it was this learning process--and not necessarily the consolidation of air, water, and waste programs into one new permit--that allowed participating facilities to reduce their emissions. Working with Department of Environmental Protection staff, facility managers in virtually every firm discovered at least one air pollution source that lacked a required permit. "Environmental managers saw their facilities, often for the first time, as a series of connections and materials flows, rather than as a checklist of point sources," Helms concluded.
At least seven states, including Oregon, Wisconsin, and New Jersey, as well as the EPA itself, have been trying to build a performance-track program that would couple some of the facility-wide approaches explored in the FWP with some of the enforcement strategies of the Massachusetts ERP. The states and the EPA are trying to establish two- or three-tier regulatory systems that reward higher-performing firms with greater regulatory flexibility.
The Wisconsin and Oregon programs offer firms a chance to propose an alternative set of performance requirements that would enhance both the environment and the firm's bottom line. Both programs recognize that each facility is unique and that imposing the most effective and efficient set of environmental conditions on each firm requires judgments about the tradeoffs between established regulatory requirements and new opportunities for environmental gain. The programs assume that regulatory flexibility--and public recognition as an environmental leader--will inspire firms to make the kind of systematic review of their pollution reduction potential that New Jersey DEP staff had to supervise in the FWP project. After making their performance-enhancing proposals, the firms negotiate a binding permit or contract with state regulators. It remains to be seen, however, how much flexibility the EPA will allow the states in approving those agreements.
The nation's environmental statutes discourage flexible multimedia permitting, reported Jerry Speir of Tulane Law School, who reviewed the projects for NAPA. The EPA's insistence on the enforceability of permits requires compliance with the letter of the law, not the spirit. The specificity of EPA's regulations is intended to paint bright lines between compliance and noncompliance, eliminating the need for plant managers, permit writers, or enforcement officers to make judgments about the effectiveness of the overall system as it applies to an individual facility. The same constraints, of course, make it unlikely that plant managers or regulators will maximize the effectiveness of a plant's systems.
A regulatory system capable of recognizing high-performing firms and then essentially leaving them alone is an ideal worth striving for. The state and federal experiments with performance-track systems may result in powerful economic incentives for firms to minimize their environmental impacts and thus qualify for maximum freedom. The EPA should encourage those experiments as an investment in long-term change, and Congress should authorize the EPA administrator to approve site-specific performance agreements that would not otherwise comply with existing laws or regulations.
One of the problems with performance-track proposals is that they still require intensive site-by-site review of facilities. Each permit or agreement is customized and thus fairly resource-intensive. Emissions caps, on the other hand, offer firms some of the benefits of greater flexibility and lead naturally to a more efficient and dynamic system of allowance trading among many firms operating under a single regional or national emissions cap.
Facility-level emissions caps are not yet routine, but they are far less controversial today than they were just five years ago when Intel and the EPA used Project XL as a framework to agree on one for a chip-making plant in Arizona. Generally, a regulatory agency sets a limit for one or more pollutants, above which a firm may not emit. In most cases, regulators then allow the firm to determine how best to stay under the cap, allowing it to make process changes without the traditional preapproval through the permitting process. The degree of flexibility varies, as do associated reporting requirements.
The EPA has been experimenting with flexible facility-wide caps and permits through Project XL and through so-called P4 permits (Pollution Prevention in Permitting Project). In January 1997, for example, EPA signed an XL agreement with a Merck Pharmaceuticals plant in Stonewall, Virginia. The agreement sets permanent facility-level caps on several air pollutants and requires increasingly detailed and frequent environmental reports as emissions approach those caps. As long as emissions are low, reporting requirements are minimal. In exchange, Merck spent $10 million to convert its boiler at the facility from coal to natural gas, achieving a 94 percent reduction in sulfur dioxide (SO2) emissions, an 87 percent reduction in nitrous oxide (NOx) emissions, and a 65 percent decrease in hazardous air pollutant emissions, compared to baseline levels.
Such caps, including those used in New Jersey's facility-wide permits, can remove perverse incentives that discourage facilities from pursuing the best possible environmental practices. The caps allow facility managers to convert to new, cleaner equipment without going through a slow and expensive permit process. The Tellus Institute's Helms reported that one of the reasons Merck had not previously converted its boilers to gas was that the company would have had to obtain permits for the new boilers, whereas the old boilers remained grandfathered out of the permit requirement.
Merck's emissions cap removed another systemic disincentive to pollution prevention. Most companies usually choose a new piece of equipment that emits right at their permitted limit, in order to avoid having the EPA lower the emissions limit based on the new piece of equipment, Helms reported. Because Merck had an incentive to keep emissions as low as possible and because of the assurance that the EPA would not lower the emissions cap, Merck managers specifically asked the procurement staff to buy the lowest-emitting gas boilers possible with reasonable reliability.
Intel has been instrumental in developing facility-level caps, using the P4 process in Oregon and Project XL in Chandler, Arizona, to negotiate agreements with regulators. Intel has replicated those agreements in several other states, including Texas and Massachusetts. All of the permits rely on mass-balance estimates of emissions; all require Intel to publish more information about actual emissions and environmental performance than most statutes require. None of the permits subsequent to Chandler has invoked Project XL, required much federal involvement, or generated much controversy. The Intel and Merck permits would probably not have been politically feasible without the firms' willingness to provide the public with detailed reports on their environmental results.
The proliferation of emissions caps represents a fundamental change in how regulatory agencies relate to pollution sources. Caps invite businesses to apply the same kind of ingenuity to environmental protection as they do to the rest of their business, provided they are in fact free to innovate and are not unduly constrained by technology-based emissions requirements in the Clean Air Act Amendments.
The significance of emissions caps is not the handful of negotiated permits described above, but the potential they demonstrate for the broader application of cap-and-trade systems to reduce emissions. Cap-and-trade systems similar to the familiar SO2 trading system Congress created in 1990 could be used to reduce nutrient loads in watersheds or NOx and volatile organic compounds in airsheds. Experience with effluent trading in water and ozone precursors in air demonstrates the potential for cap-and-trade systems to achieve specified social goals for the environment at a relatively low cost.
Allowance trading systems shift the respective roles of regulator and regulated in ways that improve the effectiveness of both. The regulator's role shifts from identifying how individual firms should control their waste stream to setting the public's environmental goal and then monitoring changing conditions and enforcing trading agreements. The regulated enterprise decides just how best to manage its own waste stream.
The essential rationale for creating trading systems to reduce pollution is that one size does not fit all. Firms--and farms, for that matter--vary in size, location, age, technical sophistication, production processes, and attitude. Those differences make it relatively less expensive for some operations to reduce their environmental impacts and relatively more expensive for others. Trading systems exploit the variances by allowing firms that can reduce their impacts cheaply to generate "emission reduction credits," which they can sell to firms at the other end of the cost spectrum. The high-cost firms buy the credits because it is cheaper than reducing their impacts directly. In short, some firms pay others to meet their environmental responsibilities for them. Their transactions reduce the total amount of pollution released by the participating firms at lower overall costs than would have been possible if regulators had simply asked each firm to install the same piece of control technology or reduce emissions by the same amount.
Cap-and-trade systems do not work in a free market; rather, they all start with government intervention in the market to achieve a broader social goal. The most important key to making a cap-and-trade system work is, of course, the cap itself. A legislature or regulatory agency must impose a pollution-reducing cap on participants: a regulatory driver that creates incentives among participants to reduce their emissions and generate emissions credits to trade. In 1990, for example, Congress required coal-burning utilities to reduce their aggregate emissions of SO2 by 10 million tons.
Reducing nutrients in surface waters
The United States will be unable to end the eutrophication of lakes and estuaries and revive the vast "dead zone" in the Gulf of Mexico unless it reduces the amount of nutrients pouring into surface waters from agricultural operations such as fields and feedlots. Those operations have not been effectively regulated, and trading systems offer one way of bringing agriculture into the environmental era with the least amount of government intrusion and expense.
Paul Faeth of the World Resources Institute has published a study that demonstrates how a trading system could work to reduce nutrient loadings in several areas of the upper Midwest. The key requirements of a trading system are present: an identifiable set of actors responsible for nutrient discharges (both point sources and nonpoint sources), reasonably effective techniques to define and verify the generation of credits (including those generated by nonpoint sources), and enormous variations in the price per ton that different actors would have to pay to reduce their contributions of nutrients.
Most of the nutrients in water systems come from nonpoint sources, and because those sources have done so little to control their contributions, enormous gains can now be made relatively cheaply. After modeling nutrient loadings in three watersheds, Faeth concluded that the most cost-effective way to reduce the loadings would be to impose 50 percent of the net reduction on the point sources and 50 percent on farmers. To achieve the former, the point sources would be allowed to trade with one another and with nonpoint sources; to achieve the latter, public funds would subsidize farmers to implement conservation measures. This combination of subsidies and trading would cost approximately $4.36 per ton of phosphorus removed, Faeth estimated, compared with $19.57 per ton under a traditional regulatory approach aimed at point sources.
Creating a trading system for nutrients will probably require congressional authorization. The Clean Water Act does not require point sources to adopt any particular technology, but the technology-based performance standards required in the act tend to be used that way. Firms have a propensity to install the same technologies that regulators used to set the standard. Those practices inhibit technological innovation, as well as the kind of flexibility that trading systems reward.
Kurt Stephenson and his colleagues at Virginia Polytechnic and State University write of the potential for provisions of the Clean Water Act to discourage both the generation and the use of credits by sources with federal permits to discharge wastes into surface waters. The act requires entities with such permits to seek renewals every five years. Regulated entities may fear that if they aggressively control their discharges and sell or bank their allowances, they will signal to regulators that regulators should impose tighter controls at renewal time, which is precisely the problem Helms identified in firms contemplating new air pollution controls. Moreover, the Clean Water Act's antibacksliding provisions prohibit permitted dischargers from purchasing allowances that will enable them to discharge more effluent than the technology-based performance standards will allow. By inhibiting both the generation and purchase of credits, provisions of the Clean Water Act would undermine trading and raise the cost of achieving the environmental goal.
Congress should authorize the EPA to foster cap-and-trade systems to reduce nutrient loadings in watersheds. Such authorization should be coupled with appropriations for expanded water quality monitoring to ensure that trading delivers on its promise.
The success of the national SO2 allowance trading system and of a regional program in southern California suggests that statewide or regional cap-and-trade systems could be an effective way for Eastern states to meet the NOx reductions that the EPA ordered in 1998, under its responsibility to prevent cross-state pollution. The order required 22 states and the District of Columbia to reduce NOx emissions by fixed amounts by 2003 and 2007. The EPA set the reduction quotas at levels intended to help reduce the long-range transport of NOx and ground-level ozone, which is partially responsible for harmful levels of ozone in the eastern United States. Midwestern and Southern states, which generate much of that ozone, had resisted imposing additional NOx controls, but the EPA prevailed in court. Now that the regulations must be implemented, many states are considering using cap-and-trade systems to achieve the specified reductions as efficiently as possible.
The existing system of NOx controls generally requires specific types of large emitters--power plants, industrial boilers, and cement kilns--to meet specific rate-based standards (measured as units of NOx per million units of exhaust volume). The evolution of those specific standards has resulted in a system that treats old and new sources differently and fails to achieve effective and efficient NOx reduction. Byron Swift of the Environmental Law Institute in Washington, D.C., identified some of the problems with today's regulations in a paper published in 2000. The Clean Air Act allows older, largely coal-fired plants to emit NOx at levels of 100 to 630 parts per million (ppm) of exhaust volume, whereas standards applied to new and cleaner gas-fired plants require NOx emissions of no more than 9 ppm, or in some states 3 ppm. The marginal cost of reducing emissions from gas-fired plants to those levels can be $2,500 to $20,000 per ton, compared with marginal costs as low as $300 per ton for coal-burning plants. This cost structure discourages investment in clean technologies.
A cap-and-trade system for NOx reductions would create incentives to invest in the least costly reduction strategies first (adding controls to coal-burning plants) while eliminating some of the disincentives for installing gas-fired turbines and industrial cogeneration facilities to the grid. Allowance trading would also tend to favor reductions in mercury and SO2 from coal-burning plants and in carbon monoxide from gas-fired plants.
Eight states in the Northeast, all members of a broader Ozone Transport Commission, have adopted compatible rules establishing the NOx Budget Program, an allowance trading system that went into operation in 1999. It requires 912 NOx sources to reduce their aggregate emissions by 55 to 65 percent from the 1990 baseline. Contrary to industry predictions, sources were able to reduce the emissions without installing expensive end-of-pipe controls. The flexibility provided through allowance trading kept costs down around $1,000 per ton in the first year.
A broader group of 19 states in the East and Midwest are subject to EPA requirements for reducing NOx emissions, and a cap-and-trade approach involving all of them would make economic and environmental sense. However, the EPA lacks specific authorization to implement such a system on a regional basis. A regional cap-and-trade approach could include 392 coal-burning power plants, as well as other large emissions sources that are the primary targets of EPA's rule. Trading at the regional scale would be appropriate, because the pollutant mixes in the atmosphere across regions, and toxic hot spots are not of particular concern with NOx emissions. In the absence of a federally coordinated regional market, individual states could implement their own trading systems. They could also collaborate to build multistate markets, as is happening in the Northeast, though doing so requires a substantial commitment of state resources. The states' other alternative is to use traditional regulatory approaches to meet their emission limits.
When EPA and the states decide to tackle the even more daunting health risks posed by sulfates and other fine particles, they will probably find cap-and-trade systems to be among the best solutions. Sulfates, such as NOx, are generated by many large combustion sources and are transported across broad airsheds. The EPA has established a monitoring network to gather more information on their transport and fate. Data from that system, coupled with the lessons from the NOx trading efforts, should provide the EPA with a foundation for establishing regional cap-and-trade systems for sulfates in the near future. Allowance trading may ultimately be part of a national strategy to control greenhouse gases. Certainly the traditional approach--uniform technology standards imposed on all combustion sources--would be unworkable.
The U.S. experience with cap-and-trade systems demonstrates that they are highly effective approaches for implementing publicly driven pollution reduction goals, provided that the sources of pollution can be identified, monitored, and regulated; that the sources face varying prices for making environmental improvements; and that the pollutants being traded are unlikely to create toxic hot spots. In other words, implementing an effective and efficient trading system requires solving significant technical challenges and overcoming even more daunting legal and political challenges. The Bush administration will need congressional authorization and encouragement to make trading systems work, and it will also need to demonstrate up front that those systems will leave the environment cleaner for nearly everyone and make conditions worse for almost no one.
Successful regulatory reform will require more of the Bush administration and Congress than simply authorizing and implementing the programs described above. The EPA will need to adopt new management approaches and build new organizations, including an independent bureau of environmental information. The kind of regulatory flexibility described above can only work if government agencies have the tools to monitor the overall effectiveness of the system and if individuals throughout the country have access to the same information and find it credible. With the advent of the Internet, we have the potential to make every citizen part of the oversight network that deters firms, communities, and states from damaging the environment or violating specific requirements. For the Internet to become such a tool, however, some institution must provide absolutely reliable, credible information about environmental conditions. That institution must be part of the federal government, though there is no office within the EPA today that can deliver on such a tall order. With better environmental information, the EPA and the states will be better able to base their relationship on performance: a critical step toward establishing priorities, detailing work plans, and assessing the effectiveness of their respective efforts.
The NAPA panel responsible for Environment.Gov concluded the volume with a set of detailed recommendations that lay out a pragmatic agenda for Administrator Whitman, the Bush administration, Congress, and the states. Recommendation 1 urged the administrator to "tackle the big environmental problems": reducing nutrients in watersheds, reducing smog, and preparing to reverse the accumulation of greenhouse gases. The only practical way to achieve these goals will be through new regulatory approaches designed to minimize the cost of environmental improvements while maximizing the American public's understanding of environmental conditions and trends. With that information, as Whitman told the Senate, "we will be able to look and know how far we have come--and how much further we need to go."
Richard A. Minard, Jr. (firstname.lastname@example.org), deputy director of the New Hampshire Center for Public Policy Studies, was associate director of the NAPA project that produced the November 2000 report Environment.Gov: Transforming Environmental Protection for the 21st Century (available at www.napawash.org).