Skip to content

About PERC

All Areas of Focus

All Research

Superfund Follies

  • Daniel Benjamin
  • EPA cleanups of superfund sites
    cost an average of $12 billion
    for every cancer case prevented.

    Most people are aware that we live in a world of scarce resources and act accordingly. Not so with the Environmental Protection Agency (EPA). When it comes to cleaning up Superfund sites around the country, recent research suggests that the EPA acts as though costs don’t matter–no matter how high those costs might be.

    In a study actually funded by the EPA, Kip Viscusi and James Hamilton (1999) have found that EPA cleanups of Superfund sites cost an average of almost $12 billion for every cancer case prevented. Even more amazing is that virtually all–99.5 percent–of the cancer cases that will be averted by EPA efforts are prevented by the first 5 percent of the agency’s expenditures. The remaining 95 percent of expenditures avert only 0.5 percent of the cancer cases–at a cost per case of an astonishing $200 billion.

    Although economists have previously scrutinized the Superfund program, which is designed to clean up hazardous waste sites, the Viscusi-Hamilton study is an improvement on several fronts. First, they use geographic information systems and detailed census data to permit the most precise assessment yet of costs and benefits of cleanups. Second, by examining how cleanup decisions are made, they uncover new evidence regarding biases in EPA responses to risks. And finally, they isolate the role of political factors in influencing cleanup decisions. Together, these features help us understand both the huge cost of Superfund cleanups and the enormous variation in those costs across sites.

    Cleanup at Superfund sites targets chemical pathways. These are the specific ways in which people are exposed to particular chemicals–such as breathing in contaminated dust blown from a slag heap. When the pathways pose a high risk, cleanup is mandatory, while with low-risk pathways cleanup is at the discretion of local EPA officials. Viscusi and Hamilton find that the forces pushing the extent of cleanup often vary markedly between these two types of sites.

    To cite just one example of EPA’s inconsistency: In high-risk settings, the agency sets more stringent cleanup standards the greater the population density, a policy that seems sensible enough. But in low-risk settings, greater population density leads the EPA to choose less stringent standards–an outcome that neither the authors nor I can justify on any sensible grounds.

    Overall, Viscusi and Hamilton find that “Superfund site [cleanup] decisions do not follow the expected pattern for efficient risk management.” This will come as little surprise to many readers, because Congress directs the EPA to make Superfund decisions without also ordering the agency to consider costs. What is disconcerting is which factors replace cost-effectiveness in guiding EPA decisions: misplaced risk perceptions and political influence.

    For example, a key ingredient in determining EPA cleanup stringency is the public notoriety of the chemicals at the site. Even after controlling for the known risks of the site, Viscusi and Hamilton found that the more times a chemical was mentioned in the popular press, the more stringent was the target (or permissible) risk chosen by the EPA. Thus, instead of cleaning the most dangerous sites, the EPA is cleaning up the sites that might get bad press.

    Sadly, the EPA does not seem to care whether the cleanup costs it incurs will actually benefit real people. That is, cleanup decisions generally are unaffected by whether the risks of the site are borne by people who live there today, or are hypothetically borne by people who might someday, under a worst-case scenario, live near the site. (Often the EPA assumes that today’s Superfund sites will someday become residential communities teeming with children.) Thus, many cancer cases “prevented” by EPA cleanups are purely hypothetical–benefits likely to materialize only in the minds of EPA employees.

    The average cost per cancer case averted by the EPA expenditures–$11.7 billion per case–masks enormous variation from site to site. At the most efficiently cleaned-up site, the cost per cancer case averted was but $20,000. At the other end of the spectrum, the cost was $961 billion. Now, the EPA is not actually spending $961 billion; indeed, the largest amount spent on any one site was only $134 million. The problem is that the hundreds of millions of dollars poured into the least efficient cleanups had so little impact that they were essentially a complete waste to society.

    What could possibly lead to such abysmal decision-making by the EPA? The answer, it seems, is plain old politics. Viscusi and Hamilton use local voter turnout as their proxy for political pressure. They find that higher turnout pushes the EPA into more stringent cleanups–and does so in the worst possible manner. For sites with cost-effectiveness at the median or better, political forces actually have little effect. But at the most inefficient sites, where costs per cancer case averted are in the billions, political factors have their strongest effect. Thus, in answer to the question: does politics matter in determining EPA policy, the answer is “yes”–by inducing local EPA managers to pursue ridiculously costly cleanups. For anyone who doubts that Superfund ranks with the worst of Congress’s policy choices, I can only hope that this study will end their skepticism.

    Viscusi, W. Kip, and James T. Hamilton. 1999. Are Risk Regulators Rational? Evidence from Hazardous Waste Cleanup Decisions. American Economic Review 89(4): 1010-27.

    Daniel K. Benjamin is a PERC senior associate and professor of economics at Clemson University. His regular column, “Tangents-Where Research and Policy Meet,” investigates policy implications of recent academic research. He can be reached at:

    Written By
    Related Content