Now that Congress will insist on risk assessment and cost-benefit analysis on all regulations, the Chesapeake Bay can rest easy. Or can it? Unfortunately, managing an ecosystem that supports hundreds of stakeholders across dozens of jurisdictions requires making very difficult decisions. Those charged with making such decisions hope to find an instrument that will process information and produce the answer: an instrument that would essentially make the decisions for them. This would not only save on work, it would also reduce their accountability. If something goes wrong, the risk model is responsible, not the decision maker. Unfortunately, neither risk assessment nor cost-benefit analysis provides such a handy tool.

There are at least three reasons why the proposed use of risk assessment in the regulatory process would be a step in the wrong direction. First, neither risk assessment nor cost benefit analysis are exact sciences. Rather, they are somewhat subjective tools whose predictions will vary depending on their application. Second, they are time and resource intensive. Finally, requiring risk assessment on all regulations is a step backward, shifting our emphasis away from necessary prevention and risk minimization.

First, the lack of information and subjectivity provide major impediments to the application of these tools. Cost-benefit analysis calculates the cost of an improvement (such as an upgrade on a sewage treatment plan) and weighs that cost against the dollar value of the benefit (such as a reduction in oxygen-depleting algal growth in Chesapeake Bay). While the dollar value of the upgrade is estimated simply by what it will cost the treatment plant, the value of the benefit is not as easily determined. How much money would be saved by increasing oxygen levels in the Bay and what is that worth to society? Techniques that have been proposed to answer these questions fall prey to limitations that arise when one attempts to monetize values that are not subject to market forces. Values that are determined are subjective, and difficult to compare to the hard costs of an upgrade or improvement which is market driven.

An environmental improvement is a difficult commodity on which to put a price tag. The cost is simple; however, the dollar value of the benefit is uncertain, even though the benefit itself is very real. The value depends on both the amount of reduction in algal growth and, perhaps more importantly, on the payer's value system. Some may not care about algal growth in the Chesapeake Bay. Others, maybe anglers, may recognize the value in protecting the fishery. Still others may see the Chesapeake as merely a conduit for shipping.

It is no surprise that in fiscal notes on environmental regulations, obvious benefits such as those to the seafood industry or tourism are classified as "Unquantifiable" rather than valued. As long as this is the case, the value will always be subjective, and the environment will never be fairly represented in a cost-benefit analysis. Thus, the analysis will remain a veiled tool through which managers can justify making difficult political decisions.

Risk-assessment techniques have similar shortcomings. First, there simply is not enough information to determine the dangers with any degree of certainty. In addition, the outcome of the assessment is subject to value judgments made by the assessor. The assessor must determine what type of risk, and how much risk, is acceptable. As a result, the process can be highly influenced by political involvement.

Historically, we have not adequately analyzed the hazards posed by chemicals. Although chemicals are responsible for a variety of health problems, we have generally limited risk assessments to only one: cancer. Most risk assessments have not considered effects on reproductive systems, digestive systems or respiratory systems (except for cancers of those organs). Also, we have failed to address effects that chemicals have had on the environment, especially those that are not immediate, but manifest themselves sometime in the future. In some cases, as with asbestos, even our cancer assessments have not been protective.

This is because contrary to what some scientists and managers may suggest, science is not almighty, and we simply do not have all the answers. The case of chlorofluorocarbons - or CFCs - serves as a good example of the fallibility of tying risk assessment to management decisions. Prior to their use, studies showed that CFCs were harmless to humans since they immediately dissipated into the air upon release. With that assessment, CFC production and use became extremely common; disposal could even be characterized as care-free. When recharging CFCs, purging the chemicals directly into the air became common practice. And why not? Science had given CFCs a clean bill of health. The effects of CFCs on the ozone layer, however, were not foreseen. The risk was not adequately assessed, and now society has to pay the costs. Some of those costs, such as increases in skin cancers, are health related; others are business related, such as the costs to the industry of a CFC phase-out.

In some cases, the results are not reversible. Oversights related to the effects of asbestos have sickened thousands of people. The public health community is currently concerned about the effects of estrogen-like chemicals on human and animal reproductive systems. These endpoints generally have not been considered in risk assessment, and even if we do begin to consider them, we will most likely continue to overlook others. Risk assessment simply is not powerful enough to give us all the answers. This brings into question exactly how much of our resources should be invested in it.

And invest we will. In order to carry out the risk assessments that the bill would require, the EPA predicts that it would have to add about 980 positions. That in itself is a multimillion dollar cost to taxpayers. According to Robert Huggett, the EPA's assistant administrator for research and development, this will also place a major additional burden on our research institutions, which are not set up to meet these demands. Rather than streamlining government activity, this measure will significantly increase the amount of paperwork surrounding regulations.

Prior to the 104th Congress, the EPA began to shift its emphasis away from costly bureaucratic measures to more efficient preventative approaches to reducing pollution. In the Chesapeake Bay, the states and the Bay Program all signed on to the Nutrient Strategy and the Toxics Reduction Strategy emphasizing pollution prevention approaches.

Using risk assessment to prioritize issues on which to concentrate, though, does not address the issue of prevention. Alleviating preventable risks should be the very first step. Many of the chemicals that we know are toxic to humans and animals have alternatives that should be pursued before we invest large amounts of resources in such a complicated and subjective prioritization process. Once the available prevention methods are in place, we will be in a better position to assess and prioritize issues. At that time, resources can be concentrated on the most pressing problems that remain.

A distinction must be made between risks that we take as individuals and risks that are subject to management decisions. Risk is the probability that something (generally negative) will happen. When we buy a lottery ticket we know we have a one in a million chance at winning, and for many of us those odds are worth a dollar or two. Those who disagree do not have to play, and nobody is going to force anyone else to drop a dollar for a lottery ticket. It is an individual decision.

With risk assessment, someone else is taking a gamble for us. This raises ethical questions. For example, the chance of an additional person getting cancer from dioxin is thought to be (for the sake of argument) about 1 in a million. This is a threshold that the government says is acceptable, given the industrial need to create dioxin through burning municipal waste and bleaching paper. So dioxin release is determined to be acceptable and one person out of every million people who is not already stricken with cancer will likely be diagnosed as a result. It doesn't sound like a lot, unless you or someone you know is that one person. But what are the chances of that? Based on today's cancer rates, the chances are frightening.

According to the National Cancer Advisory Board, by the year 2000, one in three people will be diagnosed with cancer, and one in five will die from it. This cannot be attributed to lifestyle issues alone since a significant portion of the cancer rate is unexplained by smoking and diet. Therefore, potential exposure to chemicals that are known to cause health effects should be minimized. Some would argue that making large investments in assessment while neglecting available prevention is unethical when human lives are at stake.

It is not enough to "hold the line" on health and environmental impacts. The proposed risk assessment process would move this country away from the needed emphasis on risk minimization and into an era of costly analysis that will ultimately be constrained by subjectivity and lack of scientific information.