The 25th anniversary of Earth Day came and went last year, with little fanfare and no public demand for more environmental laws. The new Republican Congress tried, and mostly failed, to enact reforms designed to lessen the burden of environmental regulation. Behind the scenes and in public forums, various schools of environmental reform debated and discussed. They talked cost-benefit analysis and "takings" compensation, emissions trading and "win-win" environmentalism. They disagreed about many things, including basic principles. But there was general consensus about two ideas: that environmental goals are important, and that the current structure of regulation isn't that great.

Environmental policy is finally growing up. But to make genuine improvements, rather than merely tinker around the edges, we first need to understand where the demand for environmental regulation comes from, and where it went wrong. And we need a vision of how environmental policy might be set right--of the general principles and concepts that might guide a new environmentalism.

Environmentalism is not, as its critics sometimes portray it, simply a New Age ideology foisted upon an unwilling public. The environmental movement has important ideological components, but the demand for cleaner air and water or for wilderness and species preservation is not that different from the demand for any other good. As living standards rise, people want to buy more environmental "goods." Pollution is as old as human activity, but only recently have we been rich enough to worry about it.

Looking across countries, University of Chicago economist Don Coursey finds a clear correlation between increased wealth, measured by per capita GDP, and increased allegiance to environmental protection. As incomes rise, per capita expenditures on pollution control increase--a phenomenon Coursey observes in most advanced industrialized nations. The amount of land set aside for protection also rises with GDP. Green groups may decry economic growth, but it is growth itself that makes environmental protection possible and popular.

Coursey's work also points up a fact often forgotten in public discussions: "The environment" is not an all-or-nothing good, but a bundle of different goods. In surveys, he asks people to indicate how much they'd be willing to spend to preserve different species. The results are wildly varied. Animals like the bald eagle and grizzly bear consistently rank high, while spiders, beetles, snakes, and snails are barely valued. The varied costs of real-world regulations reflect this distinction: Coursey calculates the amount spent to preserve a single Florida panther at $4.8 million, compared to a mere $1.17 to preserve a single Painted Snake Coil Forest Snail.

Political maneuvering may produce such disparate results, but the law does not actually recognize such distinctions, or the implicit tradeoffs they express. It declares species protection, like many other environmental goods, an absolute. Early environmentalist thinking--influential to this day--did not recognize environmental values as some goods among many but rather proclaimed them preeminent: Earth First! A California regulator describes his state's water policy this way: "If Mother Nature didn't put it in, you had to take it out--everything--that was the goal. This drove us to rigid, grossly exuberant attempts at clean up."

This absolutism suggests one way that environmental policy went wrong. It did not recognize that quality of life resides in pursuit of multiple values. People seek shelter, nourishment, health, security, learning, fairness, companionship, freedom, and personal comfort together with environmental protection. They even seek many, sometimes competing, environmental goals. They don't agree on how to marshal their resources (and time) in pursuit of these many goals. And it is often difficult for outsiders--or even individuals themselves--to know in advance how they would prefer to trade off among different values.

Competing values are not unique to the environmental arena. In fact, people make such tradeoffs every day. They also deal with another conundrum of environmental policy: the "knowledge problem."

On the one hand, environmental problems involve matters of "general" knowledge, scientific knowledge of facts that are constant across time and space. In some cases, the general knowledge is a matter of settled understanding: the boiling point of water, for instance, or the bonding patterns of chemicals. In others, it is a matter of ongoing research and scientific contention. General knowledge includes such still-controversial issues as the health hazards of various substances or the effects of CFCs on the ozone layer. Much environmental debate takes place over issues of general knowledge, and these questions are important. But they aren't the whole story.

Environmental problems and problem solving also often involve "specific" knowledge--the knowledge of time, place, and experience described by Nobel laureate F.A. Hayek in "The Use of Knowledge in Society." This information varies by circumstance and location and may change over time. Specific knowledge is decentralized--it resides on the factory floor, at a particular Superfund site, or on a specific farm.

The impacts of a landfill in a desert will differ from those near the Florida Everglades. Emitting effluent into a fast-moving stream is different from emitting waste into a pond. Using two coats of paint will have different effects than using just one coat. Resource use and emissions associated with cloth and disposable diapers will depend on how many are used each day, what kinds of disposal systems are available, and where those systems are located. Nor are environmental effects the only important specific knowledge relevant to environmental issues--tradeoffs between environmental goods and other goods may vary from situation to situation. These kinds of details do not reside in the minds of bureaucrats in Washington. Yet it is this kind of knowledge that is often most relevant to understanding environmental problems and possible remedies.

Again, there is nothing unique to environmental issues about the knowledge problem. It is a fundamental aspect of human life, something people cope with all the time.

If, for instance, an environment-loving outdoorsman wishes to equip himself for mountain climbing, he may need to buy a special jacket (among much other gear). Deciding what to buy requires tradeoffs: First, he's devoting his financial resources to a jacket, rather than to something else, and his free time to mountain climbing. Then he must decide what characteristics he wants: how much insulation, how much durability, what sort of pockets or hood, and so on--all keeping in mind the continuing tradeoff between money he spends on the jacket and money he could spend on something else. He looks at brand names and reputations. He finds out what sort of return policies and customer service the jacket maker has.

The outdoorsman has specific knowledge. He knows what mountains he wishes to climb, and at what time of year. He knows whether he's an occasional climber, or a devotee. He knows whether he wants a lot of help from a sales clerk, or simply the best price. He knows his size and his favorite color. The jacket maker, too, has specific knowledge: of supplies, of tradeoffs between design features and costs, of distribution channels, of changing patterns of demand. No one centrally decides a single jacket policy for all U.S. mountain climbers.

But behind all these considerations lie a host of social and business institutions, from the Uniform Commercial Code and product liability law to standard sizes, specialty magazines, and outdoor-equipment shops. Overcoming the knowledge and values problems is not easy or cost-free, even in well-developed markets. Companies spend millions of dollars on marketing in an attempt to better understand and address consumer values. They draw on general knowledge--on, for instance, scientific research on physiology or the properties of jacket materials. They wear-test products and adjust to ever-changing information about suppliers. They respond to the unexpected: a cold snap that sends jacket demand up, a supplier's zipper redesign, a sudden change of fashion. Throughout all these processes, they receive constant feedback, direct and indirect, from customers and suppliers.

For "ordinary" goods like jackets, we have hundreds of years of institutional evolution to make markets work, to address the problems of values and knowledge. Some of those institutions involve government policy. Many do not.

Environmental goods present special challenges because of the characteristics of these goods. But meeting those challenges has been unnecessarily difficult. For too long environmental policy has been shaped by people who demanded that environmental values trump all other considerations and who assumed that a regulatory elite possessed all necessary knowledge. Rather than figuring out how to perfect or create institutions that would allow a market for environmental goods to develop and flourish, they have been bent on opposing and destroying markets. They have seen markets not as processes for addressing values and conveying knowledge but as symbols of base commercialism and greed. This moralistic approach is finally fading. We can now begin to examine what sorts of institutions different environmental goods require--to explore a new environmental vision.

For 20 years, some of the most effective proponents of "free market environmentalism" have come out of the New Resource Economics school of thought. These economists, mostly based in the Western United States, focus on the environmental goods most easily incorporated into traditional institutions of property rights--and, perhaps not coincidentally, the oldest areas of U.S. environmental policy, dating back to the Progressive Era. Their starting premise is that individuals are predominantly self-interested. New Resource Economics then explores how different ownership settings and decision-making institutions shape incentives for stewardship.

"Actual or potential owners have incentives to use their resources efficiently," write economists Richard Stroup and John Baden, neatly summing up their thesis. Owners enjoy the fruits of efficient resource use and land stewardship through enhanced returns on their investment and maintenance of their property's value over time.

Environmental problems arise, in this model, when resources are unowned. This is the famous "tragedy of the commons," in which resources are "owned" by everyone and thus effectively by no one, because they can be used indiscriminately by everyone. Each person has an incentive to consume as much as possible, as fast as possible, rather than to preserve and protect resources for future use. According to this view, institutions, not perverse people, are the genesis of environmental problems.

"The same people who nearly destroyed the buffalo population," write Stroup and Baden, "posed no threat to the more valuable beef cattle raised on the western range. In that instance more clearly defined property rights resulted in the proliferation of beef cattle while the imperfect, if not altogether absent, property rights to the buffalo led to its near elimination."

Translated into policy, the insights of these economists encourage experiments in market creation. Environmental writer Tom Wolf describes a "Ranching for Wildlife" program in Colorado that borrows from the ideas of New Resource Economists. In the Sangre de Cristo Mountains, elk on public and private lands compete with cattle for forage. The challenge, suggests Wolf, is "to figure out how ranchers can capture value from the elk."

Hunters will pay as much as $8,000 for a license to hunt a trophy elk. But until Colorado passed its Ranching for Wildlife program, landowners were unable to take part in potential revenue from sale of hunting licenses. To cattle ranchers, then, elk were merely pests. Under Ranching for Wildlife, the state still "owns" the wildlife, but owners of large ranches can auction off, at whatever price the market will bear, a designated number of hunting licenses that guarantee trophy elk. The program, in effect, has partially privatized elk-hunting rights. Participating ranch owners now manage their lands to provide suitable habitat for elk. They also guard vigilantly against poaching of young bull elk.

One of the advantages of "privatizing" resource and land-use decisions through various property rights arrangements is that these arrangements reduce the need for consensus. Goals, such as wilderness preservation, can be pursued through private land purchases that, unlike public preservation activities, do not require majority voter approval.

Wedding their work to the pioneering work of Nobel laureate economist James Buchanan and Gordon Tullock on "public choice theory," the New Resource Economists are also able to explain some of the perversities that keep surfacing in public lands management. The U.S. Forest Service, for instance, has incentives enshrined in law that encourage it to allow large-scale timber cutting even when the logging costs more than the Forest Service takes in. Because bureaucrats have no rights to the resources they manage, Stroup and Baden argue, "Even when [they] are highly trained, competent, and well intended, the information and incentives they face do not encourage either sensitive or efficient resource management."

Most environmental issues are not, however, quite that simple. Incentives and self-interest are always present, but they are not the whole story.

Markets work for jackets and, with a little work, for elk because these goods have certain characteristics that make transactions relatively simple. It's possible to clearly specify who owns what, to identify buyers and sellers, and to convey all the necessary information for trades to take place. As a result, the market operates as a discovery process to address the knowledge and value problems--and to encourage improvements over time, such as better jacket brands or improved elk forage.

In other cases, however, things aren't so clearly defined. There are frictions: hard-to-divide goods, parties too numerous or scattered to be identified, vital information that isn't easily shared or easily known, blurry property lines. Institutions must evolve to deal with these hard cases.

One such institution is the common law. The common-law approach asks, "What happens when one person's sphere of activity conflicts with another person's?" The result is a focus on the concepts of liability, nuisance, and trespass; the role of courts in evaluating harms and benefits; and their role in resolving conflicts by clarifying the scope of different intersecting rights.

This blurry realm is not confined to a few difficult air pollution problems (and, in fact, common law may not work well for air pollution). Anywhere people congregate, conflicts emerge over sights, smells, and physical invasions that include everything from factory smoke to ugly houses to one neighbor's leaves falling on another's yard. They also include potentially big nuisances such as toxic air emissions or discharges of waste into water bodies.

The common-law framework offers a means of further clarifying rights and refining just what "enjoyment and use" of one's property means. It is thus both a mechanism for conflict resolution and a means of discovering the scope and limit of rights.

Consider an apartment building with a noisy air-conditioning unit that disturbs a neighbor. Asked for injunctive relief, one tool of common law, the court may emphasize the neighbor's rights, requiring the apartment owner to eliminate the air conditioner noise--unless the neighbor agrees to some other arrangement. Or it may assign the owner the rights, declaring that the neighbor must put up with the noise unless he can make a deal with the owner. In actual conflicts, injunctive relief usually balances the two interests. "The law of nuisance," says Chicago economist Coursey, "actually would tend to use a rule that looked like a combination...: the apartment owner may make noise with impunity up to some critical level, and, if the apartment owner makes more than the critical level of noise, the single family may obtain an order of the court directing the apartment owner to reduce the noise down to the critical level."

In his famous "Coase theorem," Nobel laureate Ronald Coase developed the theoretical underpinnings of this sort of institution. In a friction-free world, where there are no "transaction costs," he argued, it doesn't matter which side is given the rights, because the two parties can always make a deal. (This calculation covers only the question of whether activities will occur, not issues of fairness, where distribution of rights does matter.) If your air conditioner is making noise you have every right to make, I can pay you to stop; if, on the other hand, you have no right to a noisy air conditioner without my permission, you can pay me to grant that permission. Problems arise in the real world, however, because transactions aren't free and the way rights are defined affects the cost of reaching agreement. So social institutions, such as courts, step in and assign rights to minimize conflict and maximize wealth, at least in theory.

Common-law tradition embodies a discovery process that clarifies and refines rights boundaries and obligations in those blurry realms where different sets of rights intersect. Common law tends to follow precedent, and precedent can only be disturbed by private parties bringing new cases with slightly different circumstances or new arguments. As the law gets better and better at maximizing the welfare of the parties in a particular kind of case, fewer and fewer such cases will be brought. As a result, the common law tends to settle on fairly efficient rules--those that make the value pie larger. Private parties then bargain "in the shadow" of settled law, dividing a larger pie than they would under rules not tested over time.

Common law, write economists Bruce Yandle and Roger Meiners, "continues to evolve. Changing preferences and improved understanding of pollution problems continuously enter the arena of law." And, unlike statutes, common law takes into account the particular circumstances of specific situations.

The common law can work to mediate disputes between discrete, identifiable parties. But, concede Yandle and Meiners, "It is hard to imagine how common law could address urban auto emission control, ozone layer problems, and global warming, to the extent that the science of those problems becomes more settled." In such cases, it is much too difficult to identify a clear-cut "polluter" and a clear-cut "plaintiff." Either most people fall simultaneously into both categories or the cost of dividing the environmental good--clean air, an undisturbed ozone layer, etc.--is much too high (sometimes approaching infinity).

Yet the problems of knowledge and values remain, and so does the demand for environmental goods. The trick for environmental reformers is to develop a vision of evolving institutions that permits different sorts of institutions to address different kinds of issues, and to do so at the appropriate decision-making level. It helps to think of this challenge as a sequence of interrelated questions, a decision tree based on the characteristics of the particular environmental goods involved.

In some cases, what is needed is not political rule making but business institutions, analogous to standard sizes and outdoor-equipment trade shows. In these cases, environmental goods are divisible, rights are assignable, and we are in the ordinary realm of markets, where entrepreneurs are rewarded for finding ways to address the knowledge and value problems effectively. Here the only issue is allowing time for institutions to evolve on their own.

Markets for some recyclables, for instance, are hampered because buyers and sellers sometimes lack information about available supplies and demand, and uniform quality is not guaranteed. These problems resemble those of many farm commodities in the 1800s. One remedy is to mimic the experience of corn farmers a century ago: Establish a coordinated process for trading in recyclables. The recent creation of electronic listing of some recyclables with the Chicago Board of Trade is a first step in this direction.

This approach differs markedly from political activists' calls to mandate recycled content in products. Those proposed mandates simply override the specific knowledge of circumstances so critical to efficient resource use. For example, mandating high levels of recycled content in certain paperboard products can require adding extra virgin (nonwaste) fiber to maintain adequate strength of the paperboard. The result is a heavier product that uses more total fiber.

Often, however, environmental goods are indivisible and present challenges to ordinary markets. Faced with these "market failures," the traditional response from the green movement has been to substitute government coercion for individual choice. Yet absolutist regulation that suppresses knowledge and imposes a single value hierarchy is not the only way to achieve such goals as clean air. It is possible to create evolvable institutions that, while they are not as simple or politically neutral as traditional markets, capture much relevant information about knowledge and values.

Traditional regulations, such as technology prescriptions and resource-use mandates, ignore the location-specific and ever-changing information critical to all production and consumption decisions. Performance standards, by contrast, allow individuals and firms to figure out how best to achieve the stated standards: to lower overall air pollution to a certain level, for instance. This is the central insight of economists who have articulated the case for market-oriented regulations like the tradeable permits scheme set forth in the 1990 Clean Air Act Amendments. (See "Selling Air Pollution," page 32.)

Tradeable permit schemes and pollution charges provide flexibility to producers (and consumers, in the case of vehicle emission charges) that should, in the long run, result in more efficient responses to air pollution problems. These approaches still require top-down goal-setting, and they are still therefore subject to political pressures that don't affect the market for jackets. In setting such standards, general scientific knowledge is critical, and often a matter of dispute.

But the problem of indivisibility--especially in the case of air pollution--makes some sort of collective goal-setting inevitable. The number of affected parties makes common-law approaches or voluntary bargaining cumbersome, given today's technologies. What is attractive about tradeable permit schemes is their potential to prepare the groundwork for creating enforceable "clean air rights" over time. For this to happen, however, legislators need to eliminate all the current language that insists these pollution credits are not rights--language that renders investment in such credits uncertain.

Tradeable permits and pollution charges are promising mechanisms. But even if we decide they are the right mechanisms to address a certain environmental problem, we still must ask who the affected parties are--who is breathing the air in question--and where, then, the goal-setting ought to be done. For three decades, we've taken for granted the idea that there should be one single environmental standard for the whole nation. But understanding the roles of knowledge and values in defining and pricing environmental goods suggests that that may not be the case. For some problems, impacts are strictly local and narrowly circumscribed. Other environmental problems may impose regional, or even global, impacts. The locus of impact should help determine where decision-making authority resides.

If most relevant knowledge is location-specific and dynamic, decisions about "how clean is clean" and what remedies to use should take place closest to where the problem occurs. For air-emission problems, that might mean a local air basin. For decisions about siting a hazardous waste facility, that might mean bargaining between landowners adjacent to the site and the site owner.

The rationale for using decentralized bargaining approaches to address environmental problems lies not with any mystical faith that small is always beautiful, nor with the now-faddish notion that all that is good must come from communities. The rationale for these approaches builds, instead, on two premises. One is the importance of decentralized information in understanding and remedying environmental problems. The other is the importance of finding ways for real people affected by real conflicts of social space to undertake their own balancing act among competing values.

Several years ago, when the Mobro garbage barge out of Islip, New York, roamed the high seas in search of a place to unload its unappetizing cargo, nightly newscasters regaled viewers with tales of the NIMBY (not-in-my-backyard) problem. People increasingly don't want landfills--or any other "nuisance" facility--sited in their communities. But people do want to be able to throw out their waste, and the trash must go somewhere.

Political edicts directing communities to site landfills are one possible remedy. But such edicts simply override the concerns and preferences of affected individuals. Bargaining between would-be landfill operators and local communities offers another option--one already used by waste management companies. Sometimes called YIMBY-FAP (yes, in my backyard, for a price), these arrangements involve negotiations in which landfill operators offer a package of protections and benefits, including compensation, to affected landowners--or sometimes entire communities--in exchange for permission to site a landfill. The costs of the landfill are thus borne by all its customers, rather than only the property owners in the surrounding area. (The same principle applies in legal reforms aimed at requiring government "takings" compensation to landowners who bear the cost of such policies as wetlands protection. Society as a whole is buying an environmental good, and all the "purchasers" should bear the cost.)

These bargaining processes have shed light on several important decision-making conundrums.

First, surveys of New York citizens showed that acceptance of benefit packages depends on how the package is offered. Early direct involvement in the bargaining process is important to participants.

Second, perceptions matter. Just as people may "irrationally" want to buy cars with tail fins, so too they may "irrationally" dislike living next to a landfill, even if it poses no health hazards. In our study, "Too Little, Too Late? Host-Community Benefits and Siting Solid Waste Facilities," economist Rodney Fort and I note that "perception costs might be portrayed as distortions of reality, or as the inability of lay people correctly to assess the problem, but individuals will react according to their perceptions....The [perception] costs are real regardless of whether public perception is viewed as correct or rational by policy makers, scientists, and technical experts."

Technocratically minded environmental reformers fear that public perceptions will yield irrational demands, driving up costs to solve environmental problems--whether those problems involve building landfills or cleaning up Superfund sites. Those fears do not appear to be justified in circumstances where citizens face both the risks associated with a facility and enjoy directly the benefits--in the form of lower costs or higher compensation--associated with a particular remedy. Bargaining itself may serve as a discovery process, revealing more accurate information about both risks and benefits.

Unbounded fears may, in fact, be more likely to drive decisions toward "zero risk" in centralized decision processes. There, the costs of decisions are spread over an entire population, while the benefits from pursuing pristine clean-ups are enjoyed by those few near a particular site. The few then have an incentive to invest heavily in lobbying, while the many do not--with obvious results. Experience with many EPA regulations, and the very high cost per year of life saved associated with those regulations, confirms this observation.

The contrast between costly Superfund site clean-ups, which have occurred in a top-down, regulatory framework directed by the EPA, and more recent local remediation of abandoned industrial sites through negotiated settlements offers further testament to this point. Processes that create closer links between those who pay for clean-up costs and those who enjoy the benefits of clean-up offer a discipline missing from traditional top-down approaches. The locally negotiated clean-ups have, in general, been achieved at a fraction of the cost for Superfund sites.

Not all indivisible environmental problems will be well-suited to collective-negotiation processes, however. In a very imperfect world, sometimes national, state, or local restrictions may be the best we can do.

What distinguishes these circumstances? They involve a high degree of public consensus--even among the regulated parties--that limits or "injunctions" against the activity are appropriate. These are those very rare cases in which everyone would be better off if an environmentally damaging practice were ended, but where there will always be incentives to "cheat" unless a restriction can be enforced by law.

Most environmental problems do not share these characteristics, but some problems may come close--as in the use and discharge of acute, well-understood toxins into the air or water. Consider two examples: the case of the "mad hatter" and the problem of cadmium discharges into water in Japan. The term "mad hatter" came from the severe effects of using mercury in the hatter's trade to make felt hats. And in 1950s Japan, mining operations discharged cadmium into water that eventually found its way into the food supply, causing Itai-Itai ("ouch-ouch") disease that results in demineralization of bones. When the source of the disease was pinpointed in 1968, the Japanese government set strict effluent standards for cadmium and prohibited consumption of rice with cadmium concentrations above a specific standard.

In each of these cases, uniform standards on the handling, use, or disposal of these materials might reduce harm, coincide with broad-based public values, and lessen transaction costs associated with case-by-case bargaining or court remedies. In effect, a legal ban enforces a cartel, in which everyone in the industry is able to stop using the dangerous chemicals because they know their competitors will have to stop, too.

Such standards need not always emerge through government actions. When acute problems of this kind become known, the marketplace itself sometimes moves to eliminate use of the offending toxins. For example, after it became clear that vapors from chromium-plating processes resulted in serious health problems for workers, the costs from worker compensation claims drove industries to find ways of safely containing the vapors.

The same kind of evolution occurred among dentists in their use of mercury amalgams. The government did not ban the use of such amalgams. However, dentists found ways of minimizing exposures to the mercury vapors created during the preparation of amalgams, and some dentists turned to substitutes. Trade associations, trade unions, professional organizations, and consumer groups often promote these kinds of changes by monitoring safety issues relevant to their members.

Whether uniform standards ought to be considered will be a function of the level of consensus regarding whether some action should be taken; the clarity of knowledge about the causation of a problem; and the level of risk associated with the problem. Where problems are indivisible, risks posed by the problem are extremely high, and causes of those risks are well-understood, public rules offer a plausible solution.

Clearly, an environmental vision based on evolving institutions will not please everyone. It acknowledges tradeoffs among values, and it admits both the necessity and the limits of political decision making--positions guaranteed to upset both traditional environmentalists and free market absolutists. It does not promise a perfect world, merely a slowly improving one. And it faces squarely the underlying problem with current environmental regulations: Centralized, top-down rule making is ill-suited to addressing environmental problems in a complex, dynamic world in which most relevant information is location-specific and different people have very different priorities.

Applying such a vision cannot be a matter of waving a single legislative wand. Three decades of statutes have created layer upon layer of regulations. But Congress could start with a few basic reforms. The National Environmental Policy Institute is exploring ways to craft a single statute that would phase in devolution of most environmental decisions to states. The concept is worth pursuing. Devolution to states does not really go far enough, since, ultimately, what is needed is further decentralization to local communities and, where feasible, privatization of environmental decisions. But devolution to the states is a good place to start in any reform agenda.

Similarly, Congress needs to get the EPA out of the business of prospectively approving state and local environmental protection programs. Under the Clean Air Act, for example, Congress sets air-emission standards and states are delegated responsibility for developing State Implementation Plans. Using computer models and other criteria, the EPA assesses those plans to determine whether they comply with federal law. States get full emission-reduction "credits" when the EPA's computer models show a state program achieving over time some estimated pollution reduction. But this means the EPA's assumptions about everything from population growth to commuting patterns, not actual pollution levels, determine the outcome. The process prevents experimentation. It locks states into using technologies or programs that the EPA thinks--but has not necessarily demonstrated--will work. Eliminating the prospective approval process would give states the latitude to design programs they believe will achieve emission reductions, and to evaluate and adjust those programs based on real-world data.

Above all, what is needed is a fundamental shift away from an approach that is primarily regulatory and punitive to one that emphasizes bargaining, improvement in information flows, and incentives for stewardship. The 1995 House proposals regarding takings compensation are essential to realigning the incentive equation.

Turning environmental policy in the direction of more bargaining is likely to require experimentation and many varied measures. But a place to start is in revising approaches to environmental enforcement. Current approaches have blurred any distinctions between intent and accident; the line between civil and criminal cases is not always clear; sentencing guidelines bear little logical relationship to the scope of alleged crimes.

In recent soul-searching, business leaders and legislators who led the 1995 regulatory reform effort of the 104th Congress opined that their reform vision was the right one. It was, they concluded, only their message that was inadequate. This self-appraisal is too generous.

Their message was inadequate because their vision was not well thought out. Pieces of that vision--such as the need to realign private incentives through takings compensation--were on target. But bills to require cost-benefit analysis and risk assessment really targeted only a symptom--high costs and skewed goals--without looking at the more fundamental problem. Would-be reformers adopted a mirror image of their opponents' technocratic, top-down approach without thinking seriously about how to ensure environmental protection in a world in which environmental goods are widely valued.

The movement toward environmental reform will not, however, disappear. And, in the long run, the past year's setbacks may provide an important opportunity: an occasion to reflect seriously not only on the politics of environmental regulation but on the alternatives to traditional methods. We have a chance to get it right this time, but only if we are willing to invest in developing a dynamic vision.