Superfund Reform Needs Drastic Simplification

Norman W. Bernstein

In 1983, Mr. Bernstein led the Enviro-Chem site Superfund settlement with EPA — the first large multiparty Superfund settlement with the Agency after the scandals that forced the Administrator from office. In 1984 and 1985, he led the first small generator cash outs anywhere in the country at the Chem-Dyne and Conservation Chemical sites. In 1988, he led the settlement of the New York landfills litigation — the first major landfills settlement in the country. In 1992, he argued United States v. Alcan Aluminum Corp., 964 F.2d 252 (3d Cir. 1992), the first case ever lost by the U.S. Department of Justice at the federal appellate level on the issue of Superfund joint and several liability. Mr. Bernstein is a cum laude graduate of the Columbia University School of Law and a former editor of the Columbia Law Review. He chairs several Superfund site steering committees, and he recently established his own law firm in Washington, D.C.

[25 ELR 10008]

Congress will be returning to Washington about the time this Dialogue is published. The Superfund¹ reauthorization bill that did not pass in the last Congress is the natural starting point for the renewed discussion in this Congress of Superfund's reauthorization. This Dialogue is about that upcoming reauthorization and has two purposes.

The first purpose is to educate the public and the environmental bar about some of the fundamental problems in the small print of S. 1834 and H.R. 4916, the most recent Senate and House bills.² The bills are complex; the Senate version runs approximately 500 pages, the House version runs 300 pages. Many of the assumptions about the proposed legislation's content and effect are incorrect. Part of the problem is that the legislation worsened as it went through the committee process. The committee process produced legislation that if enacted, will increase costs and complexity, decrease fairness, and slow cleanups.

The second purpose of this Dialogue is to propose a far simpler legislative approach toward the liability issues, and to (a) suggest basic corrections to cleanup standards and remedy selection, and to (b) propose methods for improving the actual implementation of cleanups once they are decided upon.

As explained below, the simplified approach should reduce the transaction costs of those parties that are incurring excessive costs and will take advantage of the common-law precedents worked out over 100 years of dealing with polluting tanneries, mines, mills, and other facilities. The proposed approach also addresses one of the core problems with implementing cleanups that is not addressed by the Administration's bill. There is a need to improve the operational efficiency of the U.S. Environmental Protection Agency (EPA) in carrying out cleanups. The approach proposed in this Dialogue is to elevate dramatically the authority, qualifications, and compensation of EPA's remedial project managers (also sometimes referred to as on-scene coordinators), while reducing the role of EPA Regional staffs and their paperwork requirements. The staffs need to become a useful resource for cleanups rather than a bureaucratic bottleneck.

Finally, to keep matters simple and speed Superfund reauthorization, Congress should handle legislation on insurance issues separately. Nearly everyone agrees that pollution insurance reform is highly desirable. The national resolution of contract disputes between insurance carriers and companies, however, requires controversial mechanisms to fund and implement what amounts to a mandatory buy-out of claims. Insurance reform has nothing to do with solving such Superfund problems as what risk levels to use for evaluating contaminated groundwater and soil, whether state or federal cleanup standards control, whether the liability of potentially responsible parties (PRPs) is joint and several or only several, how costs should be allocated among those that must pay, and how to speed up cleanups. This Dialogue therefore addresses only the Superfund issues.

The Legislation That Came Out of Committee

Billions of Dollars of Potential Liability Transferred From the Public Sector to the Private Sector by a Timing Change

The original concept in the Administration's bill was that the potential liability of municipalities that generate hazardous waste for the cleanup of municipal landfills would be capped at 10 percent per municipal landfill, even if the municipalities' fair share is much larger.³ Although the Administration intended the reauthorized Superfund to pick [25 ELR 10009] up the resulting shortfall as an "orphan share," that is not exactly how it turned out.

The municipal liability cap is in the House and Senate bills' liability section and is immediately effective.⁴ The orphan share provision, however, is in the allocation section and is mandatory only for sites involving records of decision adopted after February 1994.⁵ Thus, for all sites with a record of decision adopted between 1980 and 1994, municipal liability will be capped at 10 percent with no Superfund pickup of the orphan share. Private industry will apparently have to pick up this large orphan share.

Landfill cleanups are among the most expensive in the Superfund program and typically run in the $ 40 to $ 50 million per landfill range.⁶ At least one landfill in New Jersey has exceeded $ 100 million in capital costs alone, plus an estimated $ 50 to $ 60 million in future operating and maintenance costs.⁷ In 1991, there were 231 municipal landfills on the national priorities list (NPL).⁸ At a very conservative $ 40 million per landfill, that is over $ 9 billion in cleanup costs. If municipal liability is capped at 10 percent, the private sector will pay an estimated $ 3.6 billion in orphan share costs⁹ because of the gap between the time the 10 percent cap becomes effective and the time the Superfund picks up the mandated orphan share.

Unreasonably Costly Cleanup Standards May Drive Remedy Selection

The goal of every cleanup under the reauthorized Superfund would be the "reasonable certainty of no harm" from exposure to carcinogens or noncarcinogens, expressed as a single numerical health risk level (e.g., 10<-6>).¹⁰ If it is technically infeasible or unreasonably costly to achieve that cleanup goal at a site, a less protective concentration level could be selected.¹¹ On the other hand, even where the "national goal" cleanup standard is unreasonably costly, it could still be "selected" because the cleanup standard may be "selected" irrespective of costs.¹²

Consideration of costs is only required in choosing an appropriate remedy that will achieve the site's selected cleanup standard.¹³ Clearly, if the selected cleanup standard can only be achieved through expensive or unreasonably expensive remedies, the "appropriate remedy" will be expensive indeed.

The same result obtains for surface and groundwater remediation. Standards established under the Safe Drinking Water Act (SDWA) for a hazardous substance, pollutant, or contaminant, must be attained for surface or groundwater that may be used for drinking water.¹⁴ Consideration of costs is not required.¹⁵ If there are no applicable SDWA standards, then a cleanup standard may be selected pursuant to the generic provisions of § 121(d)(1).¹⁶ Only after the cleanup standard has been selected are costs to be considered in selecting the remedy appropriate for the contaminated groundwater.¹⁷

It may well be too late to consider costs in selecting the appropriate remedy after a cleanup standard is selected because selecting a cleanup standard largely predetermines the selection of an appropriate remedy. Since the choice of the appropriate remedy may be only among two or more expensive or unreasonably expensive treatment alternatives, cost considerations become academic.

The Bias Toward Unneeded Treatment Remains

The Administration's original version of its bill was designed to reduce excessive remedy costs by eliminating the bias toward treatment in all cases. Treatment was to be the preferred remedy only for "hot spots."¹⁸ Interestingly, the first nonpartisan critique of the Administration's proposed legislation, issued by the Rand Corporation's Institute for Civil Justice (Rand Study),¹⁹ warned that if national standards turn out to be very strict "or hot spots to be broadly defined, remedies may remain quite expensive."²⁰ Keeping remedies expensive gives PRPs cause to continue contesting remedies and "maintain[s] the gap between proposed [25 ELR 10010] cleanup costs and what the PRPs think they can convince the public, Congress, or regulators is reasonable."²¹

Unfortunately, the Rand Study's warning went unheeded. During the legislative process the definition of a "hot spot" expanded to include not just "hazardous substances," but also "pollutants and contaminants."²² Virtually anything is a pollutant or contaminant.²³ S. 1834 also provides a "higher threshold" for evaluating the reasonableness of costs in selecting a remedy for its expanded definition of "hot spots."²⁴ With potentially whole sites as "hot spots" and a "higher threshold" for evaluating the reasonableness of costs, Congress' proposed legislation fails to accomplish the Administration's original objective of eliminating the bias toward treatment.

Cleanups Will Be Delayed and Made More Costly

At the heart of the proposed legislation's remedy provisions is the National Risk Protocol (Protocol).²⁵ EPA will draft and use the Protocol for conducting risk assessments to determine the need for remedial action, for establishing cleanup standards, and for evaluating remedial alternatives.²⁶ Data collected from a site is "imputed" into the Protocol. The Protocol will tell EPA (1) whether remediation is needed, (2) if remediation is needed, how much cleaning up is required, and (3) what remedial alternatives will achieve the cleanup goal.²⁷

This Protocol will require time to develop and will not work for every site. EPA must propose the Protocol not later than 12 months after enactment, and promulgate it 6 months after that.²⁸ No disincentive is provided for normal Agency delay. Even assuming EPA complies with this schedule, however, the Protocol, as promulgated, may be subject to legal challenges by industry groups believing it to be too stringent, and by environmental groups believing it to insufficiently stringent.²⁹ EPA's development of the Protocol and the judicial resolution of these challenges will likely delay new remedial actions.³⁰ The delay could easily last two to three years.

Remedial actions under the new system will likely be more expensive. The original Administration bill proposed a uniform federal cleanup standard. A state could impose more stringent and therefore more expensive standards only if the state pays "the incremental costs of the stricter standard — and [the state] cannot pass those costs on to either the federal government or to the PRPs."³¹ The requirement that the state pay for the more expensive cleanup, however, was dropped in committee.³²

What Congress left in the legislation is a complex of federal cleanup standards selected irrespective of cost and based on "pollutants and contaminants" that as applied to any one site, will always be "trumped" by more stringent and expensive state regulatory standards.³³ The state environmental regulatory agencies need only: (i) adopt the regulations under applicable state administrative procedure requirements; (ii) intend that they apply to Superfund cleanups; and (iii) manifest an intention to apply them to state-directed cleanups.³⁴ No state legislation or state money is needed to impose more stringent cleanup standards. What state regulator will be able to resist imposing more stringent cleanup standards at no cost to the state?

Incredibly, the proposed legislation may also subject trustees that are retained by PRP groups to clean up Superfund sites to unlimited personal liability. The House bill treats them as if they were owners or operators of the site.³⁵ These are not the people who caused the problem; these are the people hired to clean it up. Who will want to be a trustee under those conditions? What happens to cleanups if no one will want to conduct them? To paraphrase Alexander Hamilton: Creating unlimited personal liability for cleanup trustees reflects a zeal for liability more ardent than enlightened.³⁶

Fairness Is Hurt, Not Improved, in the Proposed Legislation

If all of these remedy problems were at least offset by improved fairness of the process, some might think that the tradeoff was worthwhile. That simply is not the case. Under existing law, a company is subject to fines and potential treble damages if it fails "without sufficient cause" to comply with an EPA § 106 cleanup order.³⁷ Although courts have strictly defined "sufficient cause," it was always possible that an EPA effort to impose grossly disproportionate liability on a small generator might be "sufficient."³⁸ That will no longer be the case.

The legislation Congress proposed provides that "sufficient cause" means either that a company has no involvement with a site at all, or that it obtains "a determination, [25 ELR 10011] by a Federal court that has jurisdiction over the order in an enforcement action, that the action to be performed pursuant to the order is inconsistent with the national contingency plan."³⁹ Section 113(h)(2), however, provides that no federal court has suchjurisdiction until after EPA sues for treble damages or penalties. Thus, when a company has to decide whether to comply with an EPA order, no matter how unfair and no matter how much it costs, its only basis for refusing to comply is that it had no involvement with the problem at all.

"Free Riding" Is Encouraged

Under existing law, a company sued under Superfund can bring a contribution action "during" or after the government's claim against it.⁴⁰ In the Superfund Amendments and Reauthorization Act of 1986 (SARA), Congress explicitly rejected a proposal pushed by EPA and the U.S. Department of Justice (DOJ) to restrain PRPs from filing contribution actions until "after" the conclusion of the government's case against the few companies it had targeted.⁴¹ Congressional concern for abuse of potential joint and several liability by the government in targeting only a handful of "deep-pocket" PRPs precluded that approach.⁴²

The proposed 1994 legislation, however, vivisects Superfund's right of contribution and eliminates the present statute's preservation of common-law contribution rights.⁴³ Worse, under the proposed allocation rules in S. 1834, if the government does not itself target a company, those companies that the government does target cannot bring a contribution action against those the government missed even though the companies the government missed may have created a significant share of the problem.⁴⁴

The government's chronic failure to locate PRPs once it has found a few deep pockets is a critical and overlooked defect in the proposed allocation system.⁴⁵ Fewer PRPs mean much larger costs for each PRP EPA targets. The cost difference can be dramatic.

For example, at one site in West Virginia, after interviewing the owner/operator, EPA only found six generator PRPs. Since 3 of the 6 generators EPA located were "deep pockets," they wound up paying for the entire $ 3 million cleanup. The three generators brought a contribution suit against the owner/operator. His records were obtained through normal judicial discovery, and more than 20 deep-pocket generator PRPs were found and named in an amended complaint. Most of these PRPs were major utilities that EPA could easily have found if it had acted to force the owner/operator to disclose his records. Within 11 months, 21 of the 24 additional PRPs found by the private parties settled by repaying approximately $ 2 million of the $ 3 million dollars spent by the 3 PRPs EPA targeted. Several of the remaining generators and the owner/operator settled about 10 months later. Under the proposed legislation, however, all of the 24 generator PRPs that EPA missed would have gotten a "free ride" and the few deep-pocket PRPs it had targeted would have had to pay most, if not all, of the cleanup costs.⁴⁶

In this context, it is important to note that the liability of PRPs that EPA misses are not part of any orphan share to be absorbed by the Superfund. Their liability is not accounted for in the allocation process.⁴⁷ Only PRPs already identifiedby EPA can be pursued for contribution, giving many companies a completely free ride.⁴⁸ In theory, a neutral allocator will have the power to request government assistance in locating PRPs, but if the government is unable or unwilling to do it for itself, how effectively will it assist a neutral allocator? Moreover, EPA recovers the same amount regardless of how many PRPs are found.

By emasculating any defense to an EPA § 106 cleanup order and drastically reducing the contribution rights of [25 ELR 10012] companies EPA chooses to target, transaction costs are said to be reduced.⁴⁹ But unfairness and "free riding" will be increased.

What Is to Be Done?

Cleanup Standards and Remedy Issues

Remedy problems can be fixed by restoring some of the Administration's original concepts, by drastically simplifying the procedures for determining cleanup standards and remedy selection, and, at least as importantly, by greatly improving the operational efficiency with which remedies are implemented in the field. A series of steps is needed to accomplish these objectives.

First, Congress must restore the Administration's original concept of a single federal cleanup standard. If a state insists on a more stringent standard it should pay for it and not be able to pass the cost back to the federal government or the PRPs. Second, Congress must restore the Administration's original concept of concentrating on hot spots and eliminate the bias against containment for non-hot spots. That means that Congress should restrict the definition of hot spots to "hazardous substances" as set forth in the original bill and not expand the definition to include "pollutants and contaminants." Congress should also strike the proposed legislation's gratuitous reference to a "higher threshold" for cost considerations in hot spot remedy selection. Third, the cleanup standard selected at every site, including standards for surface water and groundwater, must be subject to a "reality" check. The standard selected must be technically and economically reasonable. EPA should be required to consider costs in selecting a cleanup standard for a specific site. Congress must also eliminate hopelessly elaborate tests keyed only to "technical feasibility" regardless of cost. Finally, remaining tests should be kept simple. Lastly, Congress must restrain its urge to micro-manage the Superfund program while still addressing the key challenge of improving the program's operational efficiency.

Those who have actually participated in the cleanup of a Superfund site know that EPA over administers even the smallest activity. After an elaborate remedial investigation and feasibility study (RI/FS), which can take years to complete, EPA requires submittal of a preliminary design report, typically a 30 or 60 percent engineering design, to implement the remedy that EPA has chosen. The PRPs must also prepare detailed specifications on the management structure for the cleanup, the quality assurance program for testing and laboratory work, the air monitoring program, and the health and safety program. In addition, extraordinarily detailed designs of the remedy itself are required. These designs and specifications can run hundreds of pages even for a relatively small cleanup.⁵⁰

Moreover, EPA's Regional office staffs subject these designs and specifications to extensive reviews based on a series of criteria. One staff reviews quality assurance, primarily for sampling and laboratory work; another staff reviews health and safety plans and specifications; and yet another staff reviews air monitoring. EPA's own engineering contractor also reviews the cleanup plans and drawings as well as the above specifications. Additionally, state approval is typically sought at each stage of review. Matters get worse under the proposed legislation since the "community" also becomes a more active player in remedy selection. Unfortunately, no one person is really in charge or given the authority to make prompt decisions.

EPA's remedial project managers, who supposedly have that responsibility, all too frequently assemble and pass along the disparate comments of each staff and of EPA's outside contractor. On occasion, the project manager attempts to mediate between the staffs and the PRPs. After the staffs have made their comments, usually consisting of pages of single-spaced minutia, the PRPs respond. Each staff then advises the project manager as to which of the PRPs' responses are acceptable or need clarification. The PRPs again reply to the staffs' responses. The entire process of submittals, comments, responses, and approvals is typically repeated again at the 90-percent design stage. The process of preparing, commenting on and revising the endless plans, specifications and drawings — even for a relatively simple remedial action — increases costs dramatically and can take one year or more to complete, not including the several years spent on the RI/FS.

Moreover, many Superfund sites involve conditions for which limited information is available. Notwithstanding testing during the RI process, it is hard to know, for example, what is really underground at an old solvent recycling site until you dig. The best of plans worked out in excruciating detail, literally down to the size of every nut and bolt, may have to be modified once the remedial work is started. Regional staffs review and comment on the revisions to plans and specifications too. Matters become even worse if a more technically advanced remedy, such as vapor extraction, is proposed.

The underlying problem is a lack of trust in EPA's project managers to use good engineering judgment in the field. That lack of trust may not be unjustified under present conditions. Project managers are underpaid, and are underqualified. Typically, project managers earn in the $ 40,000 to $ 50,000 range, and have little more than a college degree. Some are recent college graduates with little practical experience. Others are old-line EPA employees that have not qualified for management positions. They are charged with running numerous complex engineering projects simultaneously. Each project may have a value in the tens ofmillions of dollars, and in some cases a value in excess of $ 100 million.

Congressional action is needed to drastically alter this situation. First, all EPA project managers should be paid on a par with DOJ honors program lawyers who can earn $ 60,000 to $ 80,000 a year. Second, all EPA project managers should be required, at a minimum, to have: (a) a college degree in one of the basic scientific disciplines relevant to a Superfund cleanup, e.g., chemical engineering, geology, hydrogeology, mechanical engineering, or chemistry from a program approved by the American Chemical Society,⁵¹ and; (b) a completed master's degree in environmental [25 ELR 10013] engineering. Third, all EPA project managers should report directly to the Regional Administrator or the Assistant Regional Administrator responsible for implementing the Superfund Program in that region, with no intervening management layers. That will assure that the project manager has direct access to top Regional management and ensures that top Regional management knows what is going on and can back up the project manager if necessary.

EPA Regional staffs that review detailed plans and specifications for management, health and safety, quality assurance, drawings, air monitoring etc., should be available to provide assistance to the project manager when the project manager wants their help. But the project manager must have the absolute right to override staffs. Placing centralized responsibility for decisionmaking with a highly educated, highly paid, elite corps of project managers will dramatically cut the massive over engineering and endless paper reviews. This will be good for the Agency, good for the PRPs that ultimately have to clean up the sites, and good for the environment.

Liability Issues

Pressure for reforming Superfund's liability structure comes primarily from two sources. First, a sense of unfairness arises from the government's abusive use of alleged "joint and several" liability. The problem occurs when the government targets only a few PRPs at a site when it knows, or could find out with reasonable investigation, that many others contributed to the problem.⁵² Second, there is widespread sense that transaction costs, including legal fees and engineering support expenses, are disproportionately high.

To try to remedy these Superfund shortcomings, the proposed legislation creates out of whole cloth, an elaborate system of "neutral" allocators. With the help of the government, these allocators will somehow acquire information about all of the PRPs even though the government is incapable of doing it for itself, and allocate the costs in Solomon-like fashion among the known parties in a way that most will believe fair and reasonable, while saving money in the process. This elaborate system, springing forth as it does like Gargantua out of Pantagruel's ear,⁵³ will not likely work and, moreover, is unnecessary to achieve the stated objectives of the legislation.

The neutral allocator system will not work because the allocator must rely on the government to locate many of the PRPs and it creates another duplicative bureaucratic layer in resolving allocation issues. If the PRPs targeted by EPA are to find the PRPs that EPA has missed (as they now can in contribution actions), they will need the authority to issue subpoenas and take depositions during the neutral allocation process. Congress' proposed legislation does not provide either of these powers to the PRPs.

The legislation provides that the neutral allocator will turn for help to the DOJ. The DOJ may, in turn, obtain and seek to enforce subpoenas to locate additional PRPs that EPA has missed. But the DOJ has only limited resources, and if it has not been of much help to EPA in finding additional PRPs, how aggressive will it be in helping a neutral allocator find additional PRPs? And even if a neutral allocator does somehow manage with the help of the PRPs and the DOJ to locate other parties, they will spend as much money as the government and others have always spent in hunting recalcitrant PRPs. Finally, in the allocation function, the neutral allocators will duplicate the functions of PRP steering committees that in many cases resolve allocation issues on a voluntary basis, and of federal magistrates and judges who are already there to decide exactly those issues if the PRPs cannot decide for themselves.

Congress should let the system work instead of attempting to invent a new system. The courts have already taken steps to check government abuses of alleged joint and several liability. In United States v. Alcan Aluminum Corporation⁵⁴ and In re Bell Petroleum Services, Inc.,⁵⁵ the Third and Fifth Circuits noted that nothing in the Superfund statute requires the imposition of joint and several liability. On the contrary, Congress struck all references to joint and several liability from the statute when it was adopted in 1980 and the matter was left to evolving principles of common law.⁵⁶ Congress reaffirmed that position in the 1986 amendments and made no change in the statute.⁵⁷

There is no joint and several liability under common-law principles if a defendant can show distinct harms or establish a reasonable basis for apportioning a single harm.⁵⁸ There is no especially high evidentiary threshold. It is simply a matter of weighing all of the testimony, including expert testimony, under normal evidentiary rules.⁵⁹ Moreover, although there is some division among the Federal Circuits on the question of timing, the better rule adopted by the Third Circuit is to resolve divisibility up front.⁶⁰ Similarly, the Third Circuit rebuffed EPA's attempt to collect "oversight" costs as overreaching.⁶¹ In short, Congress should simply note its approval of the Third and Fifth Circuits' decisions, and do what it did in 1980 and 1986 — leave the [25 ELR 10014] joint and several liability issue to evolving principles of common law.

Indeed, the primary problem with Superfund from the beginning has been § 107(a)'s introductory phrase: "Notwithstanding any other law. . . ." The words have never meant what they seemed to suggest, but rather have conveyed the notion that Congress intended to set Superfund apart from the common law. As a result, courts had to reinvent methods of resolving pollution liability problems that the common law had already resolved through more than 100 years of dealing with pollution from mines, tanneries, mills, and slaughter pens. This misinterpretation has provoked endless questioning on whether and to what extent basic common-law principles apply. Replacing § 107(a)'s existing introductory language with the phrase "To the extent consistent with common law . . ." will explicitly reunite Superfund with mainstream common law. The replacement phrase brings in common-law concepts of causation in fact, proximate cause, the duty to mitigate damages,⁶² and contributing to a public nuisance. In short, it restores to the law a sense of fairness and continuity.

Reducing transaction costs is the other matter to address in reforming Superfund. The Rand Study is instructive on this point and it reveals that the two groups paying excessive costs are smaller companies with $ 100 million-a-year and under in total revenues and insurance carriers.

A 1992 study of five very large industrial firms with annual revenues over $ 20 billion found that transaction costs were 19 percent of total expenditures at 49 NPL sites between 1984 and 1989. A 1993 study of 108 smaller PRPs at 18 NPL sites also found that transaction costs were 19 percent of total expenditures between 1981 and 1991 for firms with annual revenues between $ 1 and $ 20 billion, and 15 percent for firms with annual revenues between $ 100 million and $ 1 billion. Transaction-cost shares were much higher for smaller firms. Those with annual revenues less than $ 100 million averaged shares of 60 percent.⁶³

Similarly, insurance carriers expended an extraordinary 88 percent on transaction costs. The Rand Study projects that insurance carriers will expend 69 percent on transaction costs by the time the cleanup of the selected sites is completed.⁶⁴

The only group of PRPs that have paid excessive transaction costs, however, are small companies. Congress correctly identified and provided a solution for this problem in 1986 when it enacted the de minimis settlement provisions contained in § 122. Fearing criticism, EPA has dragged its feet in implementing these provisions. EPA's failure to implement this portion of § 122 has nothing to do with protecting the environment or protecting the public. It represents a classic case of bureaucracy protecting itself. When EPA promulgated its guidance for de minimis settlements, it provided that "as a general rule, de minimis settlements should not be concluded . . . prior to such time as the Agency is confident that adequate information about the extent of each settling party's waste contribution to the site has been discovered."⁶⁵ Moreover, de minimis settlements involving a full release for de minimis parties "generally should not be pursued until the Agency is able to estimate, with a reasonable degree of confidence, the total response costs associated with cleaning up the subject site, including oversight and operation and maintenance costs."⁶⁶ Most of the small player transaction costs intended to be saved, however, have already been incurred by the time those conditions can be met.

Even in the context of de minimis settlements, the Agency insisted on delays and reopeners that eliminated the value of any early settlement. EPA's June 19, 1987 Interim Guidance on Settlements with De Minimis Waste Contributors under SARA § 122(g) states:

In order to protect the Agency . . . settlements should, in most cases, also include a reservation of rights which would allow the Government to seek further relief from any settling party if information not known to the Government at the time of settlement is discovered. . . . In addition to the natural resource damage reservation and the reservation for new information . . . two further . . . 'reopeners' may be required depending upon the facts of the case and the timing of the settlement. These reopeners protect the Agency against (1) the risk of cost overruns . . . and (2) the risk that further response action will be necessary in addition to the work specified in the ROD.⁶⁷

The phrase "protect the agency" runs like bureaucratic mantra through the guidance.⁶⁸ It is time for Congress, as it did with the Clean Air Act, to take a "stick to EPA" and compel the Agency to implement the 1986 provisions on de minimis buy-out provisions. Vigorous implementation of Congress' de minimis settlement policy was intended to and can provide for the relief that small companies need to "cash out" of most Superfund problems before they incur large transaction costs.

[25 ELR 10015]

Finally, giving EPA leave to "sell" rather than "give" full releases to settling parties will provide EPA with greater flexibility and bargaining leverage in negotiating settlements with all PRPs. Even large PRPs will pay a premium to cash out. The city and state of New York have used this approach to good effect in resolving problems arising from large-scale cleanups of five major New York landfills.⁶⁹ EPA used this approach successfully in early Superfund settlements. Even under the Administration's proposal, an entity settling under the allocator's formula and paying a premium for "cashing out" will not be subject to further liability.

If EPA is willing to trust itself to set such premiums in a settlement arranged by a "neutral allocator," why is it unwilling to trust its own staff to set premiums in negotiations with PRPs? SARA's provisions allowing the Agency to provide a cash out — a full release with no reopener — shouldbe expressly extended to all, not just de minimis PRPs, provided they pay a reasonable premium to cover uncertainties left in a given case. Congress should strike the provisions restricting such releases to "special circumstances." Expanding the scope of the cash out would break open the log jam on settlements and give EPA's staff something to sell. The cash out would provide the "carrot" to supplement EPA's enforcement "stick."

Returning to common-law principles, compelling EPA to carry out § 122's de minimis settlement provisions, and freeing EPA to "sell" full releases without reopeners to all PRPs, will restore fairness, reduce transaction costs, and speed up settlements dramatically.

Congress was wise not to adopt a 500-page bill that would have created more problems than it would have solved. The steps needed to fix Superfund are straightforward. If Congress wants to reestablish credibility with the American people, it can use Superfund's reauthorization to usher in a new era of short, simple, plain English legislation that does what needs to be done.

1. "Superfund" is the common name for the hazardous substances cleanup program under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), 42 U.S.C. §§ 9601-9675, ELR STAT. CERCLA §§ 101-405.

2. S. 1834 and H.R. 4916, 103d Cong., 2d Sess. (1994) (hereinafter S. 1834 and H.R. 4916). The bills differ in a number of details but have the same basic structure and contain many of the same problems. Differences in the bills relevant to this Dialogue are noted. For convenience, this Dialogue primarily refers to the Senate version while addressing some of the more glaring problems with the House legislation.

3. Statement of Carol M. Browner, Administrator, EPA, Before the Subcommittee on Water Resources and Environment Committee on Public Works and Transportation, U.S. House of Representatives (June 9, 1994) (EPA Press Release).

4. S. 1834, § 403, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 107(n)(2)); H.R. 4916, § 403, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 107(a)(7)).

5. Id., § 409, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 129(a)(1) and (2)).

6. The Lipari Landfill in New Jersey will cost more than $ 100 million before the cleanup is complete. Jon Nordheimer, Residue of Fear Remains as Toxins Are Removed, N.Y. TIMES, Nov. 6, 1994, at A56. The remediation of the GEMS site, also in New Jersey, is estimated to cost in excess of $ 50 million. The city and state of New York estimate that it will cost approximately $ 250 to $ 450 million to clean up five city landfills, or $ 50 to $ 90 million per landfill. These cost estimates are actually understated since they do not include government enforcement costs.

7. The Helen Kramer Landfill in New Jersey was number four on the national priorities list in 1991. OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE, EPA, FACT BOOK: NATIONAL PRIORITIES LIST UNDER THE ORIGINAL HAZARD RANKING SYSTEM, 1981-1991 (Oct. 1993).

8. Id. There are currently 238 landfills on the NPL, and 1 currently proposed for inclusion. RCRA/Superfund Hotline, November 29, 1994 Search of Booze, Allen & Hamilton CERCLA Database.

9. Size of landfill cap, size of wastewater treatment plant, grading costs and methane gas collection costs are almost entirely driven by the vast quantities of municipal waste at municipal landfill sites. Assuming conservatively that municipalities' fair share is approximately 50 percent of costs, the effect of the 10 percent cap is to transfer 40 percent of costs to the "orphan share." 40 percent X $ 9 billion = $ 3.6 billion.

10. S. 1834, § 501, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 121(d)(1)(A))).

11. Id. (amending CERCLA § 121(d)(1)(B)) (emphasis added).

12. Id.

13. Id., § 502, 103d Cong., 2d Sess. (1994) (amending CERCLA § 121(b)(3)(A)).

14. 42 U.S.C. §§ 300f-300j-26, ELR STAT. SDWA §§ 1401-1465. Groundwater that may be used for drinking water is defined in S. 1834, § 501, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 121(d)(8)(E)). All groundwater may be used for drinking water unless: (i) it contains more than 10,000 milligrams per liter total dissolved solids from naturally occurring sources; (ii) it is so contaminated by naturally occurring conditions, or by the effects of broad-scale human activity that restoration to SDWA levels is impracticable; or (iii) there is not enough of it to yield 150 gallons per day to a well or spring without adverse environmental effects.

15. S. 1834, § 501, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 121(d)(8)(A)).

16. Id. (amending CERCLA to add § 121(d)(8)(A)(i)).

17. Id. (amending CERCLA to add § 121(d)(8)(A)(iii)).

18. Although designed to reduce excessive remedy costs, as discussed above, S. 1834 would retain a de facto preference for treatment.

19. INSTITUTE FOR CIVIL JUSTICE, RAND CORPORATION, FIXING SUPERFUND, THE EFFECT OF THE PROPOSED SUPERFUND REFORM ACT OF 1994 ON TRANSACTION COSTS (1994) (hereinafter the RAND STUDY).

20. Id. at 34.

21. Id.

22. S. 1834, § 502, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 121(b)(3)(B)(i)).

23. See 42 U.S.C. § 9601(33), ELR STAT. CERCLA § 101(33).

24. S. 1834, § 502, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 121(b)(3)(B)(ii)).

25. Id., § 501, 103d Cong., 2d Sess. (1994) (amending CERCLA § 121(d)(2)).

26. Id. (amending CERCLA § 121(d)(2)(A)).

27. See generally id. (amending CERCLA § 121(d)).

28. Id. (amending CERCLA to add § 121(d)(6)).

29. 42 U.S.C. § 9613(a), ELR STAT. CERCLA § 113(a) (judicial review of regulations promulgated under CERCLA available in the Circuit Court of Appeals for the District of Columbia).

30. Pending the Protocol's promulgation, the President may (not shall) select remedies relying on then current regulations and guidance. S. 1834 § 506(b), 103d Cong., 2d Sess. (1994). The legislation makes no provision for remedy selection pending judicial review of the Protocol.

31. RAND STUDY, supra note 9, at 30.

32. H.R. REP. NO. 103-582, 103d Cong., 2d Sess., pt. 1, at 202 (1994).

33. S. 1834, § 501, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 121(d)(7)).

34. Id. (amending CERCLA to add § 121(d)((7)(A)).

35. H.R. 4916, § 606, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 101(20)(E)(i)). S. 1834 appears to limit trustee liability, but appearances are deceiving. All of the "safe harbor" provisions of S. 1834 for which a trustee is not subject to personal liability are subject to a due care standard, breach of which subjects a trustee to unlimited personal liability. Section 411, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 132(d)).

36. THE FEDERALIST NO. 26 (Alexander Hamilton).

37. 42 U.S.C. § 9606(b)(1), ELR STAT. CERCLA § 106(b)(1) (as to fines); 42 U.S.C. § 9607(c)(3), ELR STAT. CERCLA § 107(c)(3) (as to treble damages).

38. See United States v. Reilly Tar & Chemical Corp., 606 F. Supp. 412, 420, 15 ELR 20348, 20351 (D. Minn. 1985). See also Wagner Electric Corp. v. Thomas, 612 F. Supp. 736, 15 ELR 20977 (D. Kan. 1985); Solid State Circuits v. U.S. Environmental Protection Agency, 812 F.2d 383, 17 ELR 20453 (8th Cir. 1987).

39. S. 1834, § 402, amends CERCLA § 106(b)(1) to add a new subsection (B). The new subsection (B) states that sufficient cause for purposes of fines and compliance "means — (i) an objectively reasonable belief by the person to whom the order is issued that the person is not liable for any response costs under section 107; or a determination, by a Federal court that has jurisdiction over the order in an enforcement action, that the action to be performed pursuant to the order is inconsistent with the national contingency plan." S. 1834, § 402, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 106(b)(1)(B)) (emphasis added). Since the term "sufficient cause" is defined in the new proposed subsection (B) for purposes of compliance as well as fines, and the same phrase (sufficient cause) is used as the sole defense to treble damages for noncompliance in § 107(c)(3), courts will construe the same words to have the same meaning in the two related statutory provisions. Thus, as to both treble damages for noncompliance and fines for noncompliance, the only criteria are the two discussed above. The House version is identical to the Senate version.

40. 42 U.S.C. § 9613(f)(1), ELR STAT. CERCLA § 113(f)(1).

41. See H.R. REP. NO. 99-253, 99th Cong., 1st Sess., pt. 1, at 80 (1985); H.R. REP. NO. 99-962, 99th Cong., 2d Sess., at 222 (1986); 131 CONG. REC. S11857 (remarks by Senator De Concini, daily ed. Sept. 20, (1985).

42. Superfund Improvement Act of 1985: Hearings Before the Committee on the Judiciary, 99th Cong., 1st Sess. at 65, 77-78, 147-51, 452-53, 512-14 (1985).

43. S. 1834, § 406, 103d Cong., 2d Sess. (1994) (amending CERCLA § 113(f) (no statutory right of contribution in the case of a § 106 order; preservation clause regarding common-law rights of contribution also eliminated).

44. Id., § 409, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 129(p)(3)).

45. In its review of the strengths and weaknesses of the proposed new system, the Rand Study missed the point completely. Rand assumed that all PRPs fell into two groups — those that participate in a proposed settlement and those that do not. It was unaware that many key PRPs will fall into neither category, i.e., those that are simply missed by the government or waste allocator, and therefore are not part of the allocation process at all. (See RAND STUDY, supra note 19, at 14-23.) As noted in the text of this Dialogue, private parties, because of the economic incentive to share the costs with others, have proved vastly superior to EPA in locating other private parties and governmental agencies that have contributed to the problem at a site. See supra note 43.

46. Another example is a large landfill (capital cleanup costs allegedly in excess of $ 100 million) located in New Jersey. EPA brought suit against only about 25 PRPs, about 20 of whom were deep pockets. It included no municipalities even though the site had been the principal disposal location for years of one of the largest cities on the eastern seaboard. The few parties targeted by EPA located about 200 other industrial generator and transporter PRPs and upwards of 50 municipalities and sewage sludge generators whose waste constituted the overwhelming bulk of the site. EPA also overlooked vast quantities of CERCLA hazardous incinerator ash sent to the site by one of the municipalities. All of those municipal generators and industrial generators would also have gotten a "free ride" under the new legislation.

47. S. 1834, § 409, 103d Cong., 2d Sess. (1994) (amending CERCLA to add § 129(h)(4)(B)).

48. As part of the price of settling, companies in the allocation process give up all contribution rights against other PRPs regardless of whether they have been identified by EPA at the site.

49. Statement of Carol M. Browner, supra note 3.

50. For example, at one site in Indiana, preliminary site preparation and removal steps in 1993 (consisting of grading, installation of drainage ditches, removal of previously emptied tanks, and the dismantling of a small shed) required several feet of plans and specifications (not including draft versions commented on by EPA).

51. At many colleges a chemistry degree is little more than a basis for teaching high school. That is why chemistry degrees acceptable for this purpose should be limited to the much more rigorous programs approved by the American Chemical Society.

52. See supra note 34 and accompanying text. This problem has been endemic since the earliest days of Superfund. For example, at the Conservation Chemical site, the government chose to sue only 4 companies when information was readily available to it that some 50 other companies had taken or transported waste to the site. Eventually, approximately 175 companies were named as third-party defendants by the 4 direct defendants. See United States v. Conservation Chemical, 619 F. Supp. 162, 16 ELR 20193 (W.D. Mo. 1985).

53. Rabelais, Francois, Gargantua and Pantagruel, Great Books of the Western World, Encyclopedia Britannica, Inc. (1952).

54. 964 F.2d 252, 22 ELR 21124 (3d Cir. 1992).

55. 3 F.3d 889, 23 ELR 21474 (5th Cir. 1993).

56. See, e.g., United States v. Chem-Dyne Corp., 572 F. Supp. 802, 806-07, 13 ELR 20986, 20987 (S.D. Ohio 1983).

57. H.R. REP. NO. 99-962, 99th Cong., 2d Sess. at 17 (1986).

58. RESTATEMENT (SECOND) OF TORTS, § 433A (1965); Alcan Aluminum Corp., 964 F.2d at 268, 22 ELR at 21134; In re Bell Petroleum Services, Inc., 3 F.3d at 895, 23 ELR at 21475, 21476.

59. In re Bell Petroleum Services, Inc., 3 F.3d at 903, 23 ELR at 21479.

60. See Alcan Aluminum Corp., 964 F.2d at 270, n.29, 22 ELR at 21134 ("[W]e believe that this inquiry [into divisibility] . . . is best resolved at the initial liability phase and not at the contribution phase. . . ."). See also In re Bell Petroleum Services, Inc., 3 F.3d at 901, 23 ELR at 21478 ("With respect to the timing of the 'divisibility' inquiry, we believe that an early resolution is preferable."). But see United States v. Alcan Aluminum Corp., 990 F.2d 711, 723, 23 ELR 20706, 20712 (2d Cir. 1993) ("[T]he choice as to when to address divisibility and apportionment are questions best left to the sound discretion of the trial court. . . .").

61. United States v. Rohm & Haas Co., 2 F.3d 1265, 1278, 23 ELR 21345, 21351 (3d Cir. 1993) (In the absence of clear congressional intent in CERCLA that EPA can recover oversight costs, the normal rule applies. Governmental costs are paid for by Congress and thereby controlled by Congress.).

62. Interestingly, federal courts are presently split as to whether EPA remedies and costs that are unreasonably expensive are nevertheless recoverable by EPA from the PRPs. Cf. United States v. Hardage, 982 F.2d 1436, 1443, 23 ELR 20624, 20627 (10th Cir. 1992), cert. denied, 126 L. Ed. 2d 248 (1993), and United States v. Northeastern Pharmaceutical & Chemical Co., 810 F.2d 726, 748, 17 ELR 20603, 20614 (8th Cir. 1986), cert. denied, 484 U.S. 848 (1987) (holding that once EPA selects a remedy that is not inconsistent with the NCP, all EPA costs are recoverable no matter how unreasonable), with In re Bell Petroleum Services, Inc., 3 F.3d at 907, 23 ELR at 21481 (5th Cir. 1993) (EPA costs incurred in implementing an arbitrary and capricious remedy are unreasonable and may not be recovered). See Barmet Aluminum Corp. v. Reilly, 927 F.2d 289, 294, 21 ELR 20850, 20852 (6th Cir. 1991) (EPA required to consider cost in conducting, not merely in selecting, remedial actions).

63. RAND STUDY, supra note 19, at 4-5.

64. The insurance carriers' problems were in part self-inflicted. They spent extraordinary amounts fighting coverage. There is no question that some overall "buy-out" of the type contemplated by the proposed legislation would cut those transaction costs and reduce the congestion in the courts caused by massive disputes between carriers and clients. But problems in resolving those issues cannot, and should not be allowed to prevent Superfund reform for everyone else.

65. EPA, Memorandum, Interim Guidance on Settlements With De Minimis Waste Contributors Under SARA § 122(g) (June 19, 1987), 52 Fed. Reg. 2433 (June 30, 1987); ELR ADMIN. MAT. I 35067.

66. 52 Fed. Reg. at 2435; ELR ADMIN. MAT. I at 35067.

67. 52 Fed. Reg. at 2435; ELR ADMIN. MAT. I at 35069 (emphasis added).

68. EPA has supplemented, but never supplanted, its 1987 guidance on de minimis waste contributors. See OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE, EPA, METHODOLOGIES FOR IMPLEMENTATION OF CERCLA § 122(g)(1)(A) DE MINIMIS WASTE CONTRIBUTOR SETTLEMENTS, OSWER DIRECTIVE 9837.7-1B (Dec. 1989); EPA, METHODOLOGY FOR EARLY DE MINIMIS WASTE CONTRIBUTOR SETTLEMENTS UNDER § 122(g)(1)(A) OF CERCLA, 57 Fed. Reg. 29312 (July 1, 1992); ELR ADMIN. MAT. I 35515.

69. See Norman W. Bernstein, To Clean Up Landfills, the Leader Should Be Municipalities Using Economic Incentives to Settle, 19 ELR 10012 (Jan. 1989).