Milton Russell

For two days, participants in the National Research Council's conference on valuing health risks, benefits, and costs for environmental decision making considered and debated the ways in which information developed in this process is presented, compared, and evaluated, and how it is to be used or ignored in making environmental decisions.

This debate was an extraordinarily important undertaking. It may be especially useful for those on the "firing line" in agencies such as the Environmental Protection Agency (EPA)—individuals who are in a position to make decisions, to advise those who do, or to prepare the analytic underpinnings that inform the decisions that are made. I know from experience that people in such positions need to set aside time from their day-to-day activities to think critically about the premises underlying their actions. They also need different perspectives, especially the ideas of those who have the opportunity and the inclination to reflect on fundamental issues of environmental decision making. Otherwise, in the press of hour-by-hour activities, they run the risk of relying on rules of thumb and on unexamined value premises of their own—or of others. This conference was designed in part to help those in government, such as myself, carefully consider and critically examine value premises.

My perspective is that of someone who has until very recently been inside the maelstrom—I am not with EPA now—but who has been outside it perhaps long enough to have established some distance. Indeed, this is the third time I have made the journey from academia to government and back. For me, there has been one constant in each of these trips: the way certain operational issues that are so dreadfully important while in government seem a ludicrous waste of energy a month later, and the increased significance, after more thought, of issues that at first appear to be of less importance. Among the latter are the deeper questions that are at the root of the issue we are examining in this volume: What is the principled basis for government decisions that affect the vital interests of citizens, and from what source do those who make these decisions gain their legitimacy? These questions have engaged some of civilization's best minds for centuries, and I have no illusions that totally satisfactory answers will emerge here. Still, environmental decision making offers a particularly thorny set of issues that must be grappled with, and it is through this grappling that the answers to these questions can be approached.

My experience at EPA revealed that the agency deals with extremely complex problems whose potential solutions have serious and far-reaching implications. I found that an explicit decision framework to sort out pros and cons, benefits and costs, was absolutely essential to any reasonable possibility of using the agency's immense power to do good rather than ill. The implications of decisions were simply too numerous and too diverse to be kept in mind without an explicit mechanism and, to the extent possible, a common metric or standard of measurement to keep score among the trade-offs that had to be made. I also concluded that what was true of the executive branch (to the extent it had discretion under the law) was true as well of the legislative branch as it formulated the statutes. Indeed, as Congress provides more and more detail in the environmental legislation it passes, it faces ever more difficulty in understanding the full implications of its actions—and ever more responsibility to do so.

It is also true, however, that only very seldom does the decision itself leap out of the analysis—that is, unless analysis is broadened so much as to lose its commonsense meaning. For example, the specifics of particular situations, the dictates of protecting a sound decision process, and the implicit signals about what sort of society should be fostered all play a role in producing a decision. A man once remarked to me that the British Navy lost its soul during World War II by issuing the perfectly rational order that convoys were not to stop to pick up survivors of submarine attacks. This policy so violated the tradition of the sea and the honor and respect a great power owed its men as to shake the national resolve. I doubt that this armchair rumination really explains, as this man suggested, the eventual loss of the British Empire, but the point was well taken—decisions that are smart may not be wise.

Nevertheless, the beginning of wisdom in environmental decision making is first to be smart, which, in my view, implies careful, explicit analysis in a structured framework.

Structured in this case does not mean that the elements within such a framework should reflect a static or overly narrow system. For example, an environmental regulation will change the situation to which it applies; consequences follow that need to be considered. Thus, new regulation yields new incentives for technological and managerial improvements that will almost certainly lead to ultimate compliance costs that are lower than those estimated. Or again, the simple fact that processes have to be rethought could overturn established ways of operating and also lead to improvements. The experience of U.S. energy use is relevant here. When the energy price shocks of the 1970s occurred and energy use in this country was examined, most firms and individuals found that they had never bothered to take actions that were well justified even at preshock prices. The same process seems to be under way today in environmental matters and can be seen in the new attention being given to safer disposal and lower production levels of hazardous wastes, the reduced use of pesticides, designs for chemical processes involving the risk of release of toxic chemicals that can handle a broader range of conditions and problems, and so forth. Any analysis of proposed environmental protection actions must take these dynamic effects into account and must also consider the second and higher order effects that follow from the initial perturbation. Often, the result of such consideration is greater risk reduction at substantially lower costs than previously anticipated—although, of course, there may be offsetting problems as well. A careful analytic effort within an explicit framework will help in anticipating these effects.

Another thing I discovered at EPA, though, was that, when it comes to the environment, many people in and out of government are opposed to the use of an explicit framework, especially one cast in benefit-cost terms. There are many reasons for this opposition that I will not detail here. It is my view, however, that one of the most important among them is an almost visceral reaction against the open consideration of any trade-offs regarding human health and the environment, even though such trade-offs are implicit in every decision.

I have some sympathy with this view. It may be that to confront the reality that life has a price, however high, undermines the foundation of a society in which we would want to live. This certainly is the view reflected in the comment on the British Navy convoy policy. It may also be that, when basic values are in conflict, it may seem worthwhile at times to foster the comforting myth of their successful accommodation.

I reject in principle, however, the elitist view that the public cannot be trusted to accept responsibility for cruel choices and that its leaders should instead feed it comforting bromides while making those choices on its behalf. Moreover, to obfuscate inevitable choices is to violate the premise of a government based on the consent of the governed, which to me is the ultimate source of governmental legitimacy in this country.

Besides, there are practical consequences to being less than open and explicit about trade-offs. An explicit analytic decision framework, with quantification to the extent possible, can be critical as a communication device and as a source of discipline for decision makers to prevent their usurpation of power that is not rightly theirs. As a communication device, an explicit decision framework makes obvious at least some of the effects of alternative actions and thereby brings to the surface the bases of decisions. As a result, others may be informed more fully and can make their judgments known. As discipline, it makes it harder for decision makers to hide behind a verbal ''fast shuffle" if they seek to impose their own views of the good society on the public without its informed consent.

Analysis of the sort that meets these requirements can take many forms, and EPA uses a rich array of techniques. Analyses range from data-based but ultimately judgmental comparative risk efforts to risk-risk comparisons, cost-effectiveness estimates, and, finally, full-blown, formal benefit-cost studies.

I welcome the discussion of benefit-cost analysis at this conference because of the important issues involved in its use: its value predicates, its unexamined assumptions, its static bias, its demands for data, and, certainly, the opportunity for manipulation of results by unscrupulous practitioners. Yet I hope that the fact that formal benefit-cost analyses have limitations and that their results can be manipulated or overinterpreted does not lead to rejection of the idea that lies behind the motive for using them. That idea proposes that what really counts is to understand what is gained and at what cost from alternative courses of action, and then to make decisions based on the balance that is cast. I have not found a better basis for decisions, or, indeed, in some deeper sense, that there is any other bass—at least when you are in a position in which you really must decide what is actually to be done.

These are strong statements. In their support, let me offer their predicates as related to environmental protection.

The first predicate is that resources are ultimately limited. There is only so much that can be done, although that amount can be made larger if people work smarter and resources are used more efficiently. The resource "pie" can also expand over time, and resources can be shifted to environmental protection so that the size of that slice of the pie may grow. Yet at any given time, to demand more than exists is an exercise in deluding others; to expect to get it is an exercise in deluding oneself.

The second predicate is that the selection of any action simultaneously rejects others. At the most basic level, labor, materials, and skills devoted to one task cannot be used for another, although it may be impractical to identify the other uses to which the resources would be put.

The third predicate is the corollary of the second. When a regulatory agency rejects one action or technology, it promotes others. For example, to forbid the use of one pesticide promotes the use of alternatives. To prevent sewage sludge from being dumped in the ocean encourages land disposal or incineration.

The final predicate is that decisions are made. Ex post facto, there is an array of goods and services produced, health risks that are borne or avoided, and environmental insults that are imposed or turned aside. There are also patterns of individual behavior that are rewarded and those that are penalized, together with social goals that are enshrined and those that are denied.

Given these predicates, the task of public policy is to make the tradeoffs that lead to a set of outcomes that are the best possible. And there's the rub: how to decide which are "best."

One view, to which I subscribe and which I think is enshrined in the American system of government, is that what is "best" depends on the values of those to whom government officials are responsible; that is, those now living. This approach does not imply a decision framework that turns its back on the past or one that ignores future generations. Nor does it mean the selfish sacrifice of other elements of planetary life for instant, narrowly human, gratification. Rather, it means leaving those choices to the citizens as a whole, working through established political institutions, instead of allowing some few who feel they know best to arrogate the decision making role. It is individual citizens who have the responsibility to consider future generations. It is up to them to reflect in their choices the long-term continuity of the natural systems on which they and future generations will depend, and which they cherish. It is up to them to attempt to convince others to adopt their values on these and other matters. Government officials and political leaders have the dual role of first participating in the education and persuasion process and then reflecting in action the goals that are selected.

With respect to whether these goals properly account for the future and for non-human health outcomes, I can only note that decisions are made by humans and that they are being made today. The only issues are which humans, working through what institutions, and reflecting which values. While neither ducks nor those yet unborn may vote, I, along with Jefferson, know of no safe repository of the power to decide other than with those who do.

The U.S. political system, therefore, mediates between the citizens, whose values are to be served, and their agents, whose decisions and actions yield the trade-offs I discussed earlier. This process brings us back to the function of analysis, which is to illuminate the ramifications of choices. It also brings us to the practical necessity of a formal analytic mechanism or set of conventions to collect, organize, summarize, and present information about alternative sets of outcomes to decision makers and the public. The practical questions that follow are how broadly should the net be cast for effects of consequence, what effects caught in that net are relevant, how are they to be valued, and how are they to be presented. As a contribution to the context for a discussion of these issues, I want to provide an illustration of the way some of these questions are presented, and demonstrate why I find an explicit decision framework essential when it comes to protecting the environment

Municipal sewage plants produce sludge that must be disposed of—on or under the ground, in the air through incineration, or in the ocean. There are irreducible risks in any of these choices.

Land-based disposal carries risks mostly for humans; ocean-based disposal modes primarily affect marine organisms. Some of these risks are incurred immediately: emissions from an incinerator are breathed at the time of disposal, and dietary risk from cropland disposal follows within months. Other risks are incurred in the future: disposal in landfills may lead to the leaching of toxic substances into groundwater, which, even in the event the water were drunk, would not bring exposure for some time. Another distinction among risks is the certainty of the exposure. The air will be breathed, but the water may not be drunk. On yet another dimension, emissions from an incinerator may expose a sizable population to risk, although of a very small level, whereas groundwater risks may be greater but would affect only the limited population that someday might draw water from an untested, contaminated well.

In addition, all of these disposal options require resources that could be used to satisfy other needs. The amount spent will vary among options and also within each option with respect to what precautions and controls are imposed. Therefore, the costs to be incurred influence the probability and magnitude of the residual risk.

Furthermore, as noted earlier, any choice that is made and enacted will affect the system as a whole. Dynamic adjustments occur that will often—although not always—yield less perturbation than a static analysis might suggest. Thus, costs are likely to be lower, as are environmental impacts, as systems rebound and defend themselves against stress.

In this hypothetical (although relevant) example of deciding where to put sludge, protecting fish has to be balanced against protecting humans. Is the probability of avoiding one excess premature death worth reducing the risk to fish in one cubic mile of ocean? A hundred cubic miles? The North Atlantic? Or, with respect to timing, is avoiding one probable excess premature death now worth as much as avoiding one next year? Or avoiding 500, let us say, 1,000 years from now?

Furthermore, how much is avoiding that excess premature death this year worth? A million dollars worth of other desired expenditures or programs? A billion dollars? The gross national product of the state of New York? What life extension as well as life enhancement would those other allocations yield?

On what basis is a lower level of risk to many to be traded off against a higher level of risk to a few? Is each person of equal concern? If so, is it just the number of health effects that is to be minimized, or is the relationship more complex than this? What about disabling or even merely uncomfortable health effects? How are they to be reckoned when the alternative is the possibility of an excess premature death? What is the rate of exchange between colds and cancer?

For those in public service, the temptation is to say that these are, in principle, unanswerable questions and that they cannot be considered together in one decision. Another response might be that the choices are too cruel to have morally acceptable answers; therefore, they should never be presented in stark terms that require individuals to face them—and face themselves after they have chosen. Rather, a veil should be cast over such choices so that the public is not exposed to them and made both uneasy and a knowing party to an essentially immoral decision.

It must be remembered, however, that the sludge has to go somewhere. When it gets there, the fabric of consequences are real, and the trade-offs will have been made. Human lives may have been exchanged for fish; current lives may have been traded off for lives in the future; one array of goods and services and risks will have been experienced while others will not; health risks of one sort will have been distributed in a particular way to a particular population. A set of values will have been summarized in an explicit decision—and somebody made that decision for the rest of the nation's citizens. In making it, that person or persons had to choose among options that exhibited different kinds of goods and bads. In the process, a common basis of comparison was used—whether it was conscious or unconscious, freely admitted or kept secret. Apples and oranges cannot be added, but how many of one must be given up to get how many more of the other can—and I believe should—be reckoned consciously, before a decision is made.

I noted earlier the resistance to a decision process that openly confronts such trade-offs, a process that, however gingerly, puts a "price" on health effects or ecological damage. Again, I can sympathize with that kind of resistance. The rhetoric that supports it strikes a primitive chord and appeals to our childlike longing for a world in which every problem has a solution and that solution is an unalloyed good. Nevertheless, cruel choices have got to be made, and it seems to me surely irresponsible in a public official to make those choices without tracing their consequences to the fullest extent possible. Formal analysis must be brought in for this task. The values to be placed on different ecological, health, economic, social, and personal outcomes at different times are supplied by the decision maker, who is responsible to the political process. The decision can be made on the basis of the balance of the apparent advantage of one option over another. This response is my incomplete and still unsatisfactory answer to the question posed at the beginning of this paper regarding the nature of a "principled basis" for making environmental decisions.

In terms of the second question posed earlier, I believe legitimacy flows from an acceptance of the decision, or at least of the decision process, by those affected. If this belief is valid, it can be achieved only when the bases of decisions are made explicit and open so that citizens also experience the reality of cruel choices, a policy that offers the possibility of true accountability, should citizens choose to exercise their oversight potential. The opportunity to confer or remove authority is essential to a free and democratic society, and providing the information on which such action may be based is essential in sustaining legitimacy. Moreover, in my judgment, being explicit and open about controversial choices is an exercise in leadership. It leads to a successful, lasting resolution of the case in point and also develops among citizens greater sophistication and understanding to make other decisions in the future.