EACH summer I'm reminded why John Steinbeck thought that Montana would be heaven if it only had an ocean. This summer, the Foundation for Research on Economics and the Environment (FREE) again hosted conferences for federal judges, seminars for environmental writers, bike trips, and research on environmental economics and policy.

But I see changes in Montana. One is congestion and the new rich who've discovered the "last best place." Also, new bureaucratic pathologies have arrived. A dramatic example is the Idaho Pole Superfund site in north Bozeman. Though many cities have Superfund sites - Denver and Salt Lake have three, Seattle has four - they are ripples in the economic ocean of a big city. In Bozeman, they are tidal waves.

Professor Richard Stroup, my former colleague, and Sandra Goodman recently published a report on Superfund, EPA's effort to clean up toxic waste ("Rights vs. Regulation; How to Reform Superfund"). Their report includes a description of Bozeman's Idaho Pole facility and indicates that Idaho Pole exemplifies many Superfund sites where health risks are overstated and bureaucratic power enhanced. This is not an accident, for the EPA has strong incentives to push caution to dangerous levels.

Twenty-five years ago, I worked in the woods and helped send some of the best and largest lodgepole pines to Idaho Pole. They treated these logs with chemicals to make long lasting wooden utility poles. At some point, a chemical used to treat these poles, pentachlorophenol (PCP), was spilled, leaving what the EPA says are potentially harmful concentrations. The EPA has tried to convince the public that the Idaho Pole site is a threat to the community.

But the EPA is probably wrong. EPA calculations of the future risks posed by the Idaho Pole site, as at many other Superfund sites, rely upon shaky assumptions. The EPA assumed that Idaho Pole's property would become a residential trailer park within two years - even though Idaho Pole is profitable - and that residents would drill wells for drinking water even though Bozeman's municipal water supply is already there. The EPA also ignored information from its own Office of Toxic Substances indicating that the concentration of PCP in water decreases by half every 20 to 200 days. Risks that may exist will diminish significantly over time.

Stroup and Goodman see further problems with the EPA's risk assessments. In identifying Superfund sites, the EPA consistently assumes unlikely, risk-enhancing situations and ignores mitigating factors in their Superfund decisions. Such methodological errors undermine the agency's judgment and make the usefulness of its National Priorities List highly suspect. Yet the Superfund law severely limits the accused polluter's right to review EPA claims in court.

Stroup and Goodman do not discuss, but I find more disturbing, Superfund's apparent violations of the spirit of our constitutional prohibition on ex post facto laws - laws that make past actions illegal and punishable. It forces firms to finance cleanups of perfectly legitimate activities when no actual harm has been demonstrated.

If people were engaged in legal, productive activities and we later decide that cleanup is necessary, we the public should pay. It is not just or equitable to impose cleanup costs on someone whose responsibility is tangential and trivial. But with Superfund's joint and several liability, any involvement can lead to complete responsibility as EPA searches for deep pockets.

To a political economist, these problems are predictable consequences of the EPA's incentives. The EPA is blamed when toxic waste harms people directly, but rarely criticized if its regulations harm people in hidden ways. For example, if a highly carcinogenic chemical leached into groundwater, and the EPA failed to act, there would be severe repercussions in Congress.

The EPA takes this logic to the extreme, concocts its farfetched "reasonable maximum exposures" and uses them to guide action. Concurrently, the agency discounts the consequences of its cleanup mandates: disruption of local communities and increasing exposure to a chemical by site disturbance. The incentives are to overstate the risks of toxic waste and to discount the human and financial costs of cleaning up.

Improving Superfund's environmental performance and reducing its ethical and economic costs requires fundamental reform. First, the EPA's farfetched risk assessments should be replaced by more sober science. Second, people should only be liable for past actions if real harms were wrongly imposed on other people. Even then, they should pay in rough proportion to their responsibility. And third, we should remove the near prohibition on accused polluters having their day in court to challenge EPA directives.

Toxic chemicals are not romantic. They lack the appeal of parks or forests. But their lesson is applicable throughout environmental policy: Well-intended government programs often go awry. Successful environmental reform must heed the ecological insight that many actions have unintended consequences. It applies to political economy as well as biology. I explore this theme in future columns.