IF YOU ARE responsible for ensuring health and safety in the work place, what do you do about a substance that is known to cause cancer, but for which the quantitative relationships -- how much is dangerous and over what period of time -- are unknown and will likely remain unknown for many years? Do you set the lowest exposure standard that is technically and economically feasible -- even if meeting that standard is quite costly? or, in the face of so much uncertainty, do you leave in place a higher -- and riskier -- but generally accepted standard while the search for better evidence countinues? If you choose to impose the more stringent standard, must you be able to demonstrate a numerical relationship between th cost of meeting the standard and the number of lives that are likely to be saved? Or is it enough to know that lowering exposure will save lives enve if you don't know how many? In short, how do you balance cost, lives and uncertainty?

There are the questions with which the Supreme Court must grapple in its consideration of a case that could have profound implications for the ways in which this country adapts itself to living with toxic chemicals. It is a case that rests heavily on the testimony and knowledge of scientific experts and crystallizes the dilemma facing scientists who are thrust into policty debates. For a scientist faced with so much uncertainty, the course of action is clear: he does not publish. His experiment is not over until further research reduces the uncertainty to an insignificant level.He goes back to the lab and collects the answers he needs.But the policy-maker, and the scientists he depends on for advice, have no such luxury. They have responsibilities, mandated by law, which force them to act.

The case now before the court concerns a standard that the Occupational Safety and Health Administration believes should be set to further limit the exposure of workers in the petrochemical and other industries to the widely used chemical benzene. Benzene has long been known to be toxic, and its allowed exposure level has dropped over the years from 100 parts per million to 10 ppm. On the basis of more recent evidence linking benzene with leukemia, a usually fatal form of cancer, OSHA believes it should now be dropped to 1 ppm. Like most other cancers, leukemia has a long latency period -- perhaps 20 years or longer -- between exposure and the appearance of the disease. This long latency, combined with the relatively recent discovery of the relationship between the chemical and the disease, means that the evidence is incomplete, but thre is now little doubt that benzene is a carcinogen.

The lower court that heard this case agreed that benzene causes cancer and other diseases. It agreed that exposure to this chemical at 10 ppm poses a health risk and that reduction of the exposure to 1 ppm would likely yield health benefits. It also accepted OSHA's contention that it is not now possible to establish a precise relationship between amount of exposure and degree of risk. And it acknowledged the widespread -- though by no means universal -- scientific consensus that there is no safe level of exposure to a carcinogen. But the court nevertheless decided that OSHA must "regulate on the basis of knowledge rather than on the unknown" and must be able to provide evidence of the "benefits" that would be gained by lowering the allowed exposure from 10 ppm to 1 ppm. This would then allow OSHA and the courts to decide whether these benefits -- saves lives -- bear a "reasonable relationship" to the costs of meeting the new standard.

The Supreme Court justices must now decide whether this particular cost-benefit calculus is necessary or even permissible. In doing so, they will privately have to wonder how -- if the benefit's were, as if by magic, suddenly known for certain -- each of them would solve the equation: how many lives for how many dollars?