Jan 1, 1994
The Superfund program, established in 1980 to clean up the nation's worst inactive hazardous waste sites, has been criticized for generating excessive transaction costs—costs resulting not from cleanup but from assigning liability for cleanup among the various parties. The Superfund Reform Act of 1994 proposes changing the program's operation to reduce these costs. This report assesses whether reforms such as those embodied in the proposed legislation are likely to be successful. The legislative process is not yet complete as we go to press, and we focus on the reform act as proposed by the Clinton Administration in February 1994 and modified with more complete insurance provisions in April 1994. When appropriate, however, we reference congressional amendments being considered.
We evaluate the provisions of the proposed legislation according to their effect on three central attributes of the Superfund process:
Our evaluation draws on the quantitative findings of prior RAND research as well as on discussions with many Superfund participants conducted during that research. We have supplemented these data with additional telephone interviews with stakeholders in the Superfund process, including representatives of potentially responsible parties (PRPs), insurers, reinsurers, the Environmental Protection Agency (EPA), the Department of the Treasury, environmental groups, and research organizations. We make no attempt to present the views of different stakeholders systematically. Rather, we use the reactions of the various stakeholders to inform the analysis and assessment.
Previous RAND work suggests that the Superfund program has generated substantial transaction costs. RAND estimated that the overall transaction-cost share was 32 percent of total expenditures by PRPs through 1991 at all sites on the National Priorities List. Transaction-cost shares for insurers were even higher—88 percent of outlays between 1986 and 1989. These estimates imply that 36 percent of the approximately $11.3 billion spent by the private sector at Superfund sites through 1991 went to transaction costs rather than to cleanup. What the costs will be when cleanup is completed is very uncertain. We estimate the final private-sector transaction-cost share between 23 and 31 percent.
The Superfund process generates transaction costs because it creates a complex set of interactions among the different players and provides many opportunities for the parties to contest the amount of their liability. We focus on five key interactions.
Throughout the Superfund process, each of the key players implicitly or explicitly weighs the costs and benefits of cooperating with the process or contesting it. The incentives to contest rather than to cooperate in large part determine the size of the transaction costs.
The proposed legislation seeks to lower transaction costs by reducing the incentives of PRPs and their insurers to contest liability and by simplifying the process and making more efficient task assignments to the participants. The legislation's provisions focus on the key aspects of the interactions among the players: allocating liability, negotiating settlements, recovering costs from nonparticipating PRPs, determining cleanup standards and selecting remedies, involving the community, and negotiating with insurers and reinsurers.
Figure S.1 summarizes our assessment of how the proposed legislation would affect transaction costs. The key interactions appear as the column heads, and aspects of the Superfund process addressed by the proposed legislation are the row labels.
|Aspect of Cleanup Process||Interaction|
|PRP-PRP||PRP-Govt||PRP-Insurer||Comm.-PRP & -Govt||Insurer-Reinsurer|
|Allocation||↓Likely decrease in transaction costs||↑Likely increase in transaction costs||–No significant effect||–No significant effect||–No significant effect|
|Settlement||↓Likely decrease in transaction costs||↓Likely decrease in transaction costs||–No significant effect||–No significant effect||–No significant effect|
|Cost recovery||↓Likely decrease in transaction costs||↑Likely increase in transaction costs||–No significant effect||–No significant effect||–No significant effect|
|Cleanup standards and remedy selection||–No significant effect||UEffect uncertain||–No significant effect||–No significant effect||–No significant effect|
|Community participation||–No significant effect||UEffect uncertain||–No significant effect||↑Likely increase in transaction costs||–No significant effect|
|Insurance||–No significant effect||–No significant effect||↓Likely decrease in transaction costs||–No significant effect||↓Likely decrease in transaction costs|
Key: ↓ Likely decrease in transaction costs; ↑ Likely increase in transaction costs; U Effect uncertain; – No significant effect
NOTE: The similar size of all the boxes does not mean that the increase or decrease in transaction costs represented by each box is equal.
The proposed legislation seems likely to reduce transaction costs generated by the major interactions among PRPs. But the overall effect on transaction costs generated from the PRP-government interaction is unclear. On the one hand, settlement terms appear much more favorable to PRPs, who thus should have less incentive to fight liability. On the other hand, some of the transaction-cost savings among PRPs come at the price of increased transaction costs for EPA, and many critical issues in setting cleanup standards and selecting remedies remain unresolved. If the insurance provisions of the legislation pass an early vote by PRPs, the legislation should simplify the interactions between PRPs and their insurers and reduce PRP incentives to pursue insurers in court. The legislation would probably also reduce insurer-reinsurer transaction costs to the extent that it reduces PRP-insurer transaction costs. But the enhanced community involvement provided for by the legislation may increase transaction costs between the community and the government because of the increased complexity of the decisionmaking process, combined with the likely persistence of divisions within the community. The proposed legislation may also increase transaction costs between the community and PRPs if increased information about the site generates more bodily injury and property damage cases.
The Superfund Reform Act of 1994 would fundamentally change the way the Superfund program operates. Although the process would remain complex, important components would be simpler and more standardized. The legislation reassigns several tasks to parties better qualified to perform them, and it appears to give PRPs far greater incentives to cooperate rather than to contest settlements. Some troubling features remain, but the legislation takes a significant step toward reducing the resources devoted to litigation rather than to cleanup.
PRP-Insurer and Insurer-Reinsurer Interactions
Community-Government and Community-PRP Interactions
Derivation of Estimated Overall Private-Sector Transaction-Cost Share