Superfund Transaction Costs: A Critical Perspective On the Superfund Liability Scheme

William N. Hedeman, Jonathan Z. Cannon, and David M. Friedland

Editors' Summary: The Superfund liability scheme has been heavily criticized for creating needless litigation among potentially responsible parties (PRPs) and for generating inordinately high costs and extended delays in cleaning up hazardous waste sites. This Article discusses the transaction costs and delays arising at each step in the Superfund process from initial listing of a site on the CERCLA national priorities list, through the search for PRPs, to the formation and maintenance of PRP committees and the remedial investigation/feasibility study processes. The Article describes the millions of dollars in administrative, legal, engineering, consulting, and other management costs that governmental and private parties spend to prepare for and participate in litigation designed to raise money for cleanups. These transaction costs are an unavoidable result of the existing Superfund liability system and do not directly contribute to cleaning up hazardous waste sites. The Article concludes that various settlement tools — de minimis settlements, mixed funding, nonbinding allocations of responsibility, model consent decrees, alternative dispute resolution, and covenants not to sue — may incrementally reduce transaction costs and delays but do not address the Superfund statute's underlying tendency to force parties to litigate or prepare for litigation at every step in the process. The adversarial nature of the process remains unchanged, with EPA extracting funds on a site-by-site basis and PRPs exercising legal options to resist EPA and to impose cleanup costs on other parties. Further, the government's resources and ability to control the process are limited, even under an aggressive enforcement-first policy. Thus, in the current public debate over whether to change Superfund's liability scheme, the Article contributes to the necessary first step of finding out why the system is not working.

The authors are environmental attorneys with the Washington, D.C., office of Beveridge and Diamond. Mr. Hedeman is a partner with the firm and served as Director of the Superfund Program for the U.S. EPA from 1981 through 1985. Mr. Cannon is also a partner with the firm, is a former Deputy Assistant Administrator of the U.S. EPA Office of Solid Waste and Emergency Response (OSWER), and served as Acting Assistant Administrator of OSWER from March through November 1989. Mr. Friedland is an associate with the firm and was Director of the Legal Writing Program and an adjunct professor at Vermont Law School from 1985 through 1987. This Article is a modified version of a Report the authors prepared for the American International Group.

[21 ELR 10414]

I. Introduction

The Superfund law¹ is grounded in the principle that the U.S. Environmental Protection Agency (EPA) should sue potentially responsible parties (PRPs) to force them either to clean up hazardous waste sites in the first instance, or later to reimburse EPA for cleaning up sites. In either event, enormous sums of money are placed at risk for parties ranging from industrial companies to municipalities, lenders, and insurers. This Article describes how the high-stakes Superfund liability system breeds protracted negotiation and litigation, which, in turn, entail significant costs unrelated to cleaning up the site. We refer to these as "transaction costs."

These transaction costs impose substantial economic costs on various parties and may divert resources from cleanups to uses that are not environmentally responsive, such as lawyers' and consultants' fees. Transaction costs also directly affect whether the Superfund program can achieve its statutory goals. The inevitable result of the activities that create transaction costs — litigation and negotiation, for example — is significant delays in cleaning up sites. In other words, not only are resources spent on noncleanup activities, but the cleanup itself may be deferred while parties wrangle over liability issues.

This Article details transaction costs at each step in the Superfund process. Virtually all transaction costs relate to the search for parties to help shoulder site cleanup costs and to the allocation of those costs among PRPs, between PRPs and the Superfund, and between PRPs and their insurers. In identifying PRPs, EPA incurs substantial upfront costs. In turn, these PRPs spend more money on lawyers and consultants to find other PRPs. The universe of PRPs then spends huge sums on lawyers and consultants to negotiate who should pay what, a process that results in numerous studies and data searches as companies try to document their own and others' contributions to the waste. More money flows to consultants who compile studies to challenge the studies performed by EPA-hired contractors.

Through it all, one overriding reality looms: the constant threat of, and preparation for, litigation with EPA, states, insurers, and nonsettling PRPs. The possibility of litigation causes PRPs and their lawyers to challenge any decision or factual finding that could pose a potential problem in a future lawsuit. Each challenge can further delay the ultimate cleanup. For example, if EPA is preparing a feasibility study for a site, PRPs will often prepare a shadow study to ensure that EPA has selected appropriate remedies, and to prepare for litigation concerning the selection of the remedy. In the end, these dueling studies must be reconciled, a process taking months or years, and sometimes resulting in yet more studies.

EPA, for its part, recognizing the possibility of having to sue PRPs, leaves no stone unturned in analyzing possible remedies so it can defend itself against charges that its decisions were arbitrary and capricious. EPA must take similar precautions in documentation of government-incurred cleanup costs. As a result, EPA may enter into contracts for studies and documentation that are more extensive, detailed, and costly than necessary.

PRPs can challenge an EPA-selected cleanup remedy only at the point when the Agency seeks to recover its costs, and they can rely only on what is in the administrative record up to that point. Thus, PRPs must create a record on anything that would be relevant in a future lawsuit, a costly and time-consuming necessity. Indeed, no serious negotiations are held without prior scrutiny aimed toward ensuring that the record is of litigation quality.

PRPs also turn to their insurers for reimbursement of cleanup costs. This litigation is complex and costly, with insurance claims reaching billions of dollars. Hundreds of attorneys argue over both substantive liability questions and procedural questions such as which state's law applies. Often, the insurance litigation and the cleanup litigation involve two different sets of attorneys. In fact, insurers and insureds spend approximately $ 500 million a year on Superfund litigation involving insurance coverage² — apart from related nonlitigation expenses.

EPA and the business community recognize that the Superfund program entails substantial transaction costs. The Agency has attempted to ease delays and to cut cleanup costs through such mechanisms as de minimis settlements, mixed funding, nonbinding allocations of responsibility, model consent decrees, improved PRP searches, and negotiated timetables. However, even assuming the success of these reforms, negotiation and litigation will continue between EPA and PRPs, among PRPs, and between PRPs and their insurers. The use of these tools does not alter the underlying liability-based funding system.

As long as EPA must look to private parties to fund cleanups, enormous transaction costs will arise from the unavoidable, contentious process of allocating liability. The fault does not lie with EPA or PRPs. Both act in a rational, responsible way given the demands the process of allocating liability imposes. This Article raises the fundamental question of whether, in the final analysis, Superfund's liability-based funding mechanism is the most efficient and effective means for cleaning up our nation's uncontrolled waste sites and for achieving the overriding goal of protecting human health and the environment.

II. Background

A. Superfund Funding Mechanisms

Operating under a general policy that the polluter must pay, Congress enacted the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA),³ as amended by the Superfund Amendments and Reauthorization Act of 1986 (SARA)⁴ (together CERCLA or Superfund), to clean up uncontrolled waste sites containing hazardous substances. Cleanups are funded in two ways. First, CERCLA, as amended, authorized EPA to spend $ 10.1 billion from a Trust Fund through 1991 to clean up these sites. This money is raised by taxes on feedstock [21 ELR 10415] chemicals and petroleum, with a broad-based corporate environmental income tax, and a small contribution from general revenues.⁵

Second, by enacting the liability-based litigation scheme discussed in this Article, Congress created a fundraising role for EPA. EPA may either clean up a site itself, using money from the Trust Fund, and then bring a lawsuit seeking recovery of its costs from PRPs at a later date,⁶ or it may issue an enforcement order under CERCLA § 106, forcing PRPs to pay for the cleanup in the first instance, with no expenditure of money from the Trust Fund.⁷ If EPA does the work, and issues an enforcement order under CERCLA § 106, the statute allows the Agency to pursue recovery of its cleanup costs at three times the amount it spent, and to impose civil penalties of $ 25,000 per day until the cleanup is completed.⁸ The EPA Administrator, in recent testimony before Congress, affirmed the Superfund enforcement-first policy, in which EPA forces PRPs to perform the majority of cleanups.⁹

B. Liability Under Superfund

Courts have interpreted PRP liability under Superfund to be strict, joint and several, and retroactive.¹⁰ This means that a PRP may be liable for the entire expense of cleaning up a site even though the party's actions were legal and reasonable at the time, even though the company may have disposed only a small amount of hazardous substances at the site. PRPs range from large corporations to small businesses, banks, states, municipalities, hospitals, universities, and the federal government.¹¹ CERCLA imposes liability on the present owner or operator of a site — regardless of whether any hazardous substances were disposed there during the period of ownership, the past owner or operator of the site at the time hazardous substances were disposed there, transporters of hazardous substances to the site, generators of hazardous substances disposed at the site, and others who arranged for disposal of hazardous substances at the site.¹² Liability for today's cleanup costs arises, in most cases, out of disposal that took place many years ago.

III. Overview of Costs Associated with the Liability-Based CERCLA Scheme

A. What Are Transaction Costs?

Criticisms of virtually every aspect of the Superfund program have been leveled by Congress, environmental groups, industry, and even EPA.¹³ These criticisms have attacked the pace, scope, and nature of cleanups, inefficiencies and inequities in the program, and many other issues. This Article focuses in depth on one central concern embedded in the existing adversarial Superfund statutory scheme: transaction costs and delays. Transaction costs are the millions of dollars in administrative, legal, engineering, consulting, and other management costs borne by EPA, private industry, and local governments that do not directly result in cleaning up a waste site, but are nonetheless inevitable in establishing liability for cleaning up the site. Transaction costs have a broad impact both financially and environmentally; they are measurable in terms of dollars spent and in terms of environmental cleanup delays.

Transaction costs are an unavoidable result of the Superfund liability system and have a significant impact on the program. Indeed, in a recent report, the Congressional Office of Technology Assessment (OTA) expressed far-reaching concerns about the inefficiency of the Superfund program, resulting from, among other things, the adversarial nature of the program, which OTA calls "Superfund Syndrome."¹⁴

B. The Superfund Process

To understand the transaction costs associated with Superfund cleanups, one must first understand the framework. Table 1 summarizes the process from site identification to the beginning of the actual cleanup.

[21 ELR 10416]

Table 1: Superfund Cleanup and Enforcement Process [SEE ILLUSTRATION IN ORIGINAL]

To dramatically oversimplify the process discussed in depth in the next section, EPA first becomes aware of a site. Then, the Agency does a preliminary assessment todetermine how the site was used and what wastes are there. Next, EPA does a preliminary search for PRPs, followed by a more formal search. EPA then evaluates site conditions according to hazard ranking criteria to determine whether the site should be placed on the CERCLA national priorities list (NPL)¹⁵ and also whether interim response actions — referred to as removal actions — should be undertaken to abate any immediate or potential threats to public health. If the site is placed on the NPL, EPA evaluates its options for cleanup.¹⁶

In a "fund-lead" site cleanup, EPA will next hire contractors to evaluate potential remedies. The Agency will perform the cleanup and wait to sue the PRPs to recover its costs at a later date. By contrast, in an "enforcement-lead" site cleanup, EPA issues orders to or sues the PRPs, forcing the PRPs to conduct the response action themselves, with EPA retaining a supervisory role. Whether fund-lead or enforcement-lead, once the site is placed on the NPL and PRPs have been notified, the next step is a formal study of the conditions and options for remedying the environmental degradation — known as a remedial investigation/feasibility study (RI/FS). The RI/FS often entails protracted negotiation. It is followed by a record of decision (ROD), which documents the broad remedy chosen. Negotiations then commence over a remedial design/remedial action (RD/RA) plan to flesh out the specifics of the remedy identified in the ROD. Note that if the liability-based system were eliminated, EPA would simply identify sites, list the sites on the NPL, fund the RI/FS, prepare the RD/RA and the ROD, and then commence cleanup. The next section of this Article evaluates the significant transaction costs and delays at each of these steps.

[21 ELR 10417]

IV. Description of the Transaction Costs and Delays at Each Step in the Superfund Process

A. Initial Steps — Preliminary Investigation Through Listing on the NPL

EPA takes the early steps in the evolution of a Superfund cleanup. These include site identification; preliminary assessment of site conditions; inspections to determine possible public exposure (actual or potential) to site hazards, including groundwater contamination and other forms of environmental damage; and the formal hazard-ranking process that determines whether a site will be listed on the NPL. The Rand Report notes that an average of 43 months elapses between the time EPA is made aware of a site and the time a site reaches the NPL.¹⁷

Large transaction costs arise for PRPs even at this early stage in the Superfund process. First, some PRPs will argue that a particular site should not be listed on the NPL because it does not meet the NPL hazard ranking criteria. In fact, in the latest revisions to the NPL, EPA withdrew a proposal to list eight sites on the NPL, and recategorized three sites as awaiting deletion.¹⁸ If a site is not removed, and a PRP still believes it should be removed, the PRP may litigate the matter in federal court.¹⁹ Second, EPA may proceed with an RI/FS at a site even before the site is listed on the NPL.²⁰ Consequently, PRPs will incur any costs relating to shadow RI/FSs and the like, as discussed throughout this Article.

B. PRP Identification

Next, EPA identifies PRPs. This entails significant costs for the Agency, because it must locate and hire contractors and civil investigators to search for PRPs. EPA has even resorted to taking out ads in local newspapers in its search for PRPs.

EPA's guidance documents suggest that a PRP search should be complete by the time a site is proposed for inclusion on the NPL and before any obligation of federal funds for the RI/FS is proposed.²¹ Theoretically, EPA should identify all potentially liable parties at the site. In practice, EPA often will identify only enough parties to commence an action. For example, in the past year, EPA has increased its use of unilateral administrative orders under CERCLA § 106,²² compelling PRPs to perform work at a site. Although detailed information is not yet available, some PRPs have observed that EPA issues § 106 orders to relatively few parties — fewer than 10 — even though the sites ultimately may involve dozens or hundreds of PRPs.²³ This practice imposes substantial transaction costs on the named PRPs because those parties will have a great interest in ensuring that all potentially liable parties are joined in the enforcement action.

Searching for other PRPs is a difficult and time-consuming process. PRPs often will file Freedom of Information Act (FOIA)²⁴ requests with EPA, seeking all information relevant to the site so that other parties can be found and joined. Under FOIA, EPA must respond to these requests by a tight deadline.²⁵ Responding to these FOIA requests requires the Agency to expend resources in determining whether any documents requested are "enforcement sensitive" and thereforenot subject to release under FOIA. Named PRPs will also hire private investigators to find other parties. Presently, several firms specialize in finding PRPs. Just as the named PRPs have a strong incentive to find other potentially liable parties, those parties have a strong an incentive to resist being named as PRPs at the site.

Disputes also arise between named PRPs that have small liability shares and those having large liability shares. To reduce their proportionate liability share, PRPs with a larger share often want to identify as many additional PRPs as possible, while PRPs with smaller shares want to curb site investigation expenses, which they often bear pro rata. Moreover, EPA's failure or inability to identify all or substantially all PRPs may delay the rapid cleanup of the site, because the named PRPs will not want to agree to a cleanup plan until they are certain of the number of parties that will ultimately contribute to the cleanup costs.

One example illustrates the transaction costs incurred as a result of the inefficient PRP identification process. For cleanup at the Lone Pine landfill in New Jersey, more than 300 parties have been identified as PRPs to date.²⁶ However, in July 1982, EPA sent letters to only 14 companies identifying them as PRPs. It was not until the autumn of 1984, more than two years after EPA identified the first PRPs, that EPA, through the efforts of the initial PRPs, notified an additional 142 parties of their potential liability.²⁷

C. Response to CERCLA § 104(e) Information Requests and the Formation of PRP Committees

Once EPA has identified some PRPs, it notifies them of their potential liability. Sometimes EPA will send a general notice, stating simply that a company may be a PRP. EPA may also require that all notified PRPs attend a meeting to discuss potential remedial action at the site. EPA often will first send an information request under § 104(e) of CERCLA,²⁸ which directs the company to determine the amount of hazardous substances that were shipped to the site. The § 104(e) request may be general or extremely detailed, and [21 ELR 10418] may require the company to interview present or former employees and do extensive searches of company files. Response to a broad § 104(e) request is expensive and time-consuming. EPA often takes the position that all present and former employees having knowledge of the hazardous substances disposed at the site must be interviewed. Frequently, former employees are difficult to locate. Many companies retain outside counsel at this point to begin protecting their interests in what will be a long and complicated process.²⁹

Once a sufficient nucleus of identifiable PRPs is established for the site, EPA encourages the parties to form PRP organizations. These organizations may become mini-bureaucracies that hire lawyers and technical consultants to evaluate liability and perform other tasks. As described in one handbook commonly used by PRP organizations — the PRP Organization Handbook³⁰ — the PRP group will create working committees to aid in the negotiation process. Typically, a steering committee is responsible for making all decisions for the group. The steering committee may hire one counsel — known as common counsel — to speak for the group in negotiations with EPA, but each company represented on the steering committee generally will also retain its own lawyers. The PRP Organization Handbook details the duties of the steering committee as follows:

Negotiating with the government and other parties regarding settlement and other matters;

Retaining and directing the activities of outside consultants and common counsel;

Appointing members of other committees;

Recommending to the PRP group a method of allocating costs;

Recommending to the PRP group a de minimis buyout proposal, if appropriate; and

Referring settlement matters to the group.³¹

Numerous additional committees are generally formed: an executive committee to handle administrative and financial matters; a technical committee to oversee technical consultants and negotiate technical issues with the government; an allocation committee to recommend a method to allocate shared costs among the group; and a de minimis committee to recommend the terms of a de minimis settlement, if appropriate.³² Table 2 reflects the complicated structure of a typical PRP organization agreement.

Table 2: Committee Structure Created by Initial PRP Organization Agreement [SEE ILLUSTRATION IN ORIGINAL]

[21 ELR 10419]

The creation and maintenance of a PRP organization often entails high costs. Hundreds of lawyers and consultants may be involved at a site, and participation by company personnel in the negotiation and litigation process leads to further costs to the detriment of other productive activity.

D. One Early Contentious Issue — Cost Allocation Among PRPs

Allocating responsibility for costs at the site is difficult and expensive. Many scientific, technical, and policy issues arise. For generators of waste transported to the site, liability is often allocated based on a volumetric share of the total waste disposed at the site.³³ However, this allocation scheme is difficult to apply to site owners, operators, and transporters because these parties did not generate a percentage of waste at the site. Nonetheless, liability must be allocated to all parties, a process that entails negotiating delays as each party seeks to reduce its proportionate share.

1. Volumetric Shares. Generators have argued that the volumetric share concept may be unfair if certain wastes are more toxic than others or have a greater tendency to migrate into the groundwater. If the hazardous constituents discovered in sampling groundwater can be traced to the wastes of a particular PRP, other PRPs will argue that their proportionate shares should be reduced accordingly. At larger sites, PRPs may create an allocation committee to evaluate all of these arguments.

2. Incomplete Data. Another factor that complicates cost allocation is incomplete data. For older sites in particular, companies will have little or conflicting data on the amount of wastes shipped to the site. At the time the wastes were shipped, there may have been no reason to keep records. Second is the problem of so-called orphan shares — shares of liability of companies that have gone out of business or cannot otherwise be reached. PRPs must negotiate an allocation method to distribute the costs of the orphan shares among themselves.

3. Municipal Liability. Related to the orphan share question is the liability of municipalities as PRPs. EPA takes the position that municipalities generally are not liable as generators at Superfund sites as long as only household wastes collected in the normal course of business were disposed at the site.³⁴ Increasingly, however, private party PRPs concerned about the large amounts of municipal wastes at hazardous waste sites are challenging EPA's determination. They are attempting through actual or threatened lawsuits for contribution under CERCLA §§ 107 and 113(f)³⁵ to force municipalities to share greater costs at Superfund sites. For example, PRPs for the Operating Industries site in California have sued 27 towns located near the site. The towns have spent more than $ 1 million defending themselves against the lawsuit brought in March 1990, and they expect to spend several million dollars more.³⁶ Additionally, federal courts in California and Connecticut have ruled that the Superfund statute does not expressly exempt household wastes from liability if those wastes contain hazardous substances.³⁷

4. Building the Record. Finally, overlaying cost allocation issues, as well as all of the other actions taken by PRPs, is the tension between litigation and settlement. At an enforcement-lead site, PRPs recognize that if they cannot reach an agreement with EPA on aproposed remedy, they may be sued by the Agency or issued a CERCLA § 106 order with the possibility of treble damages. This threat gives each PRP an incentive to proceed cautiously and to challenge at every juncture any EPA or PRP decisions that may set precedents in a future lawsuit. PRPs can challenge a remedy selected by EPA only at the cost recovery stage and can rely only on the administrative record to support a claim that EPA's action is arbitrary and capricious. PRPs therefore pay close attention to ensuring that the record contains anything and everything that may possibly be relevant in a future lawsuit. Consequently, PRPs often will duplicate EPA's efforts, particularly on risk assessments. This constant eye toward future litigation causes both PRPs and EPA³⁸ to ensure that all technical analyses are of litigation quality. Predictably, the result is significant costs and delays.

E. The RI/FS

The remedial investigation (RI) is a study of the conditions at the site. Groundwater and soil samples are collected for analysis. The RI may require between 18 months and 3 years to complete. The feasibility study (FS) is an engineering analysis of methods to clean up the site. The FS typically takes between three and six months. Many factors complicate the RI/FS, including the complex mixture of hazardous substances found at most sites, the lack of adequate historical records relating to contamination, and the scientific uncertainties inherent in groundwater monitoring and remedial approaches.³⁹ According to EPA, the costs of an RI/FS are likely to average approximately $ 1.3 million per site.⁴⁰

In a fund-lead cleanup, EPA or its contractor prepares the RI/FS. PRPs will often do a shadow RI/FS to ensure that the record contains all information needed to defend against a subsequent enforcement action.⁴¹ At an enforce [21 ELR 10420] ment-lead site, EPA is required under CERCLA § 104(a)(1) to oversee the RI/FS prepared by the PRPs.⁴² EPA officials have also testified that EPA will do risk-endangerment assessments for all future PRP lead RI/FSs.⁴³ The shadowing process is obviously inefficient and expensive, but it is an inevitable result of the litigation-or-preparation-for-litigation relationship between EPA and PRPs created by the Superfund statute.

PRPs must also keep a close eye on the contents and costs of the RI/FS for future allocation purposes. EPA must track its direct and indirect costs, which it may seek to recover from PRPs, and which PRPs will then audit before they agree to pay.

F. The ROD

At an enforcement-lead site, once a ROD is issued, EPA again sends special notice letters to PRPs. The parties have 60 days to submit a good faith offer to perform the remedy described in the ROD. If EPA receives an acceptable offer, another 60 days are allowed to negotiate a draft consent decree embodying the agreement.⁴⁴ The draft consent decree undergoes review by EPA and the U.S. Department of Justice (DOJ) headquarters' officials and, if deemed acceptable, it is lodged with the appropriate federal district court. Public comment is received for 30 days, after which the decree is finalized and formally entered by the court. EPA, the DOJ, and PRPs expend considerable resources on attorney, consultant, and engineering fees in the ROD/consent decree process.

G. The RD/RA

Once the RI/FS has been completed and a selected remedy incorporated into the ROD, EPA then develops a more specific engineering plan — the remedial design (RD), and implements the selected remedial action (RA). If a decision is made to undertake the RA with Trust Fund monies that would subsequently be recovered from PRPs in a cost recovery action, the state in which the site is located must pay 10 percent of the RA costs, or at least 50 percent of the costs if the site is or was owned or operated by a local government. Because states are often strapped for money, EPA is further motivated to aggressively pursue an enforcement-first policy against PRPs to avoid the inherent delays that might occur at the RA state if the state contribution becomes necessary.⁴⁵

Concerns have arisen about the length of time spent on the RD/RA process.⁴⁶ One source of delay has been negotiation between EPA and PRPs. The U.S. General Accounting Office asked regional EPA project managers and attorneys what factors accounted for the protracted RD/RA negotiations, and their responses are reflected in Table 3.

Table 3: Factors Contributing to Lengthy RD/RA Negotiations

(Responses in Percent)

Somewhat

Number of Greatly to to not

Factors respondents totally Moderately at all

Number/complexity

of issues to be

settled (too many,

too complex 110 67 21 12

Number of PRPs

(too many) and/or

PRPs were not

organized 108 56 18 26

Disagreements

among PRPs 108 48 30 22

Size of EPA staff

work loads 109 42 38 20

Total cost of the

cleanup remedy

(too costly) 102 41 29 30

PRPs pursue tactics

to delay the

progress

negotiations 109 38 38 24

Note: The number of respondents represents the subset of respondents who have a basis to judge from among the universe of regional project managers and attorneys who indicated that they had moderate to extensive experience in RD/RA negotiations (N-115). The percent of respondents is for this subset of respondents having a basis to judge.

Source: U.S. General Accounting Office, Superfund, A More Vigorous and Better Managed Enforcement Program Is Needed, 41 (Dec. 1989).

The most prominent reasons for lengthy negotiations — the number and complexity of issues to be settled, the number of PRPs and/or the fact that PRPs were not organized, and disagreements among PRPs — are a product of the inexorable costs and delays that result from EPA's mandate under Superfund to pursue PRPs for cleanup costs through negotiation or enforcement actions.

In sum, while a definitive study quantifying Superfund transaction costs has yet to be performed, those costs are likely to be substantial. From the initial listing of a site on the NPL, through the search for PRPs, to the formation and maintenance of PRP committees and the RI/FS and RD/RA processes, both EPA and private parties incur these transaction costs as an inevitable result of the need to prepare for and defend litigation designed to raise money for cleanups.

V. Transaction Costs of Secondary Litigation

In addition to the disputes described in the Article's previous section between EPA and PRPs over identification of responsible parties and selection of remedies, many other complicated lawsuits will inevitably arise at virtually every site as PRPs struggle to spread their costs to their insurers and to other PRPs. This Article refers to these lawsuits as "secondary litigation."

[21 ELR 10421]

A. Intra-PRP Litigation Expenses

Further costs and delays at many sites are caused by lawsuits brought by PRPs against other PRPs. The issues vary. Some PRPs may have gone out of business and cannot be located. Others may file for bankruptcy, and individual PRPs often may participate in the bankruptcy proceedings, hoping ultimately to recover some money from the bankrupt person or entity.

Some companies not named by EPA as PRPs will keep a low profile, hoping to avoid being dragged into the enforcement action. This forces the named PRPs to engage in extensive research and investigation costs, with the ultimate goal of bringing the recalcitrant PRPs into the process.

One example of the complexity of intra-PRP litigation involves the Ludlow Landfill site in New York. The State of New York filed suit against the landfill company and its owners, and these PRPs then brought into the litigation five other companies. In 1990, two of these companies agreed to pay approximately $ 4 million toward the cleanup. The companies then filed a contribution action under CERCLA §§ 107 and 113(f)⁴⁷ against 600 companies and government entities that they claimed should contribute to the costs of cleaning up the site.⁴⁸ The waste generators and transporters named included transportation authorities, post offices, banks, hospitals, fast food restaurants, building contractors, the local Elks Club, and veterinarians.⁴⁹ The 44 municipal defendants sued included cities, school districts, and fire departments. Similarly, at the Cannons Engineering site in Massachusetts, EPA initially identified as PRPs, among other parties, high schools, local laundries, courthouses, and a department store.⁵⁰

Other PRPs may dispute their liability under CERCLA, leading to further litigation. For example, numerous cases deal with the question of when a party is "arranging for disposal of a hazardous substance" as contrasted with selling a commercially valuable product, albeit one that contains hazardous substances. Courts have held that the sale of asbestos-containing building insulation material to a contractor, the sale of PCB-laden transformers to a utility, and the sale of creosote to a lumber and paper manufacturing facility do not constitute arranging for disposal of a hazardous substance under CERCLA.⁵¹ By contrast, courts have held that the sale of waste oil to a racetrack for dust suppression purposes and the sale of used caustic solvents to a hazardous waste treatment facility for use as fuel do constitute arranging for disposal of a hazardous substance and would incur CERCLA liability.⁵² PRPs continue to litigate their liability in the contexts of these cases.

Other key issues for determining the scope of PRP liability remain unsettled:

the extent to which a bank or other lending institution may be deemed a PRP as a result of foreclosure on a loan;⁵³

the extent to which municipalities may and should be liable as PRPs;⁵⁴

the scope of liability for corporate successors, parents, and subsidiaries;⁵⁵ and

the breadth of the so-called "innocent landowner" defense to Superfund liability.⁵⁶

All of these issues arise from the inherent incentives to litigate created by the site-by-site funding/liability mechanism of Superfund. Although it is true that many issues surrounding Superfund — such as joint and several liability — have been resolved, complex factual scenarios and new legal issues remain — such as the scope of municipal and lender liability — that will ensure continued litigation for the foreseeable future.

B. PRPs Versus Their Insurers

As PRPs are notified of their potential liabilities or are assessed their share of cleanup costs, many turn to their insurers seeking reimbursement. In several cases, insurers have refused to pay these claims, resulting in complicated [21 ELR 10422] litigation.⁵⁷ Litigation between a PRP and its insurers for purposes of assessing transaction costs is usually pursued by a new group of attorneys, experts, and consultants rather than the same professionals employed by the PRP to negotiate with EPA regarding site cleanup.⁵⁸

As with cleanup actions, the stakes in the insurance litigation are high, with claims reaching millions and billions of dollars.⁵⁹ The litigation is complex, both legally and factually. Common legal issues relating to insurance coverage addressed by courts are:

Whether liability for cleanup constitutes "damages," as that term is used in the insuring agreement;⁶⁰

Whether property damage or bodily injury caused by disposal of hazardous wastes was expected or intended. If such actions were expected or intended, coverage is denied;⁶¹ and

Whether traditional policy exclusions, such as the pollution exclusion, act to bar all or a portion of coverage, as well as the broader question of whether a government-ordered cleanup is an occurrence within the terms of the policy.⁶²

Examples of factual issues that arise are:

What kind of operations were conducted on the site and what were the disposal practices?⁶³

Did conditions at the site cause property damage?⁶⁴

Was timely and adequate notice of a claim or an occurrence provided by the policyholder to the insurer?⁶⁵

As a preliminary matter, insurance litigation is governed by state, not federal law. As with any legal issue, the laws of some states tend to favor policyholders, while others favor insurers. This leads insurance companies and PRPs to search for the jurisdiction most favorable to their claims. Depending on the law of the particular state, a PRP may wish to argue that the appropriate forum for the lawsuit is the state where the site is located, the state in which either the PRP or the insurance company is headquartered, the state whose laws are set forth in the insurance contract, or another state. Even when diversity jurisdiction exists, federal courts will apply state law to resolve the dispute. These jurisdictional disputes often take years to settle.⁶⁶

Extensive discovery of the drafting of the insurance policy, involving depositions and interrogatories, often is conducted to determine the intent ofthe parties to the insurance contract and to satisfy state common law rules governing contract interpretation. Insurers often will pursue broad discovery to probe the policyholders' past waste management practices.

In addition, even after a court determines that there is insurance coverage, many issues remain. The site is likely to involve multiple insurers with overlapping policies. Sometimes these insurers may be joined in the original lawsuit; other times separate lawsuits must be filed. If coverage exists, the court may have to apportion liability among the various insurance policies and carriers.

Court rulings both for and against insurers make outcomes unpredictable. Trials are expensive for both the insurer and the PRP, in part because large fees are paid to lawyers, engineers, and other experts necessary to the litigation process. Many [21 ELR 10423] cases are appealed, resulting in further expense and delay. All of this litigation, of course, does nothing to clean up Superfund sites. One former long-time staffer for OTA estimates that insurers and insureds spend $ 500 million annually on Superfund litigation involving coverage issues.⁶⁷

VI. The Root of the Problem — Overall Costs and Delays

Having described the nature and scope of transaction costs at each stage in the Superfund process, this Article turns briefly to a more global discussion of the overall costs and delays occasioned by the statutory scheme.

A. Costs

EPA estimates that the direct cost of remedying the problems at sites currently on the NPL is likely to be about $ 30 billion, with half the work paid by the Trust Fund and half by private parties. The Agency also estimates that it will take 13 years to begin construction on all of those sites, and that between 75 and 100 sites will likely be added to the list each year.⁶⁸ The NPL will, therefore, grow from the current approximately 1,200 sites to 2,100 by the year 2000. EPA's former Assistant Administrator for Enforcement, James Strock, recently testified that the average cost of construction per site is $ 29 million. He further noted that this estimate did not even include, among other things, enforcement costs.⁶⁹

In response to questions from members of Congress, Mr. Strock estimated that 10 percent of all costs obligated by the government for the Superfund program go to litigation and enforcement efforts.⁷⁰ EPA obligated $ 1.6 billion for the Superfund program in 1990,⁷¹ so litigation costs for that year would be $ 160 million under Mr. Strock's estimate. Transaction costs for private industry are significantly higher than EPA's because PRPs typically engage in both intra-PRP litigation, litigation with their insurers, and also litigation with state agencies and the U.S. Department of Justice.

No one knows the precise amount of transaction costs that accompany these cleanup costs, but several parties have offered estimates. OTA, for example, has stated that no precise quantitative analysis of the linkage between litigation and negotiations and spending is possible.⁷² OTA estimates that 20 to 40 percent of total spending for the Superfund program is inefficient because of prolonged negotiations and litigation between EPA and PRPs. Moreover, this figure applies only to EPA spending. Private sector spending, OTA estimates, is likely to be just as inefficient because of litigation between PRPs and other PRPs, PRPs and their insurers, and PRPs and citizen groups:

Responsible parties perform about half of current site studies and cleanups and many of their activities and problems mirror EPA's. For example, they also bear high administrative, management, and transaction costs. But the mix of private sector spending in the latter area is probably different than for EPA. Responsible parties are probably spending much more, proportionately, on litigation than on studies, administration, and management. In addition to negotiation and litigation with the government, responsible parties are in negotiation and litigation with other responsible parties, insurance companies, and private citizens and community groups. One recent review of Superfund concluded that "of the total funds spent since 1980 … something between 30 and 60 percent has gone for legal expenses." Of course, not all legal expenses are unnecessarily high or unavoidable, but here too, OTA believes it is fair to estimate that a significant portion of legal spending is unnecessarily high or avoidable and, therefore, inefficient.⁷³

It is easy to see why the Superfund structure breeds such litigiousness, for the stakes are so high — millions of dollars per site — and it is a zero-sum game — that is, to the extent that one party pays less for cleanup, another party pays more. All parties have a strong incentive to assure that someone else pays. This incentive, combined with the constant threat of enforcement litigation by EPA, necessarily results in extremely cautious, contentious, time-consuming, and expensive lawyering by each PRP. This litigation diverts time and money that could be spent actually cleaning up sites.

B. Delays

Many reports have chronicled the delays in cleaning up sites that have occurred throughout the life of the Superfund program.⁷⁴ The Rand Report, for example, states that on average it takes more than eight years from the time a site comes to EPA's attention to the time definitive cleanup work begins.⁷⁵ Table 4, from the Rand Report, describes the average time between principal steps in the Superfund process.

Table 4 [SEE ILLUSTRATION IN ORIGINAL]

Source: Rand Report at 16.⁷⁶

[21 ELR 10424]

Further, the Sasser Report notes that only four percent of the sites on the NPL have been cleaned up.⁷⁷ The Sasser Report also notes that many thousands of sites are not even in the EPA system.⁷⁸ Finally, the Sasser Report also chronicles delays at specific sites:

The site ranked by EPA as the number one toxic waste hazard in the country languished on the Superfund list for eight years before substantive cleanup work began. Almost immediately, cleanup work was halted due to problems with the main subcontractor. Only now, in the summer of 1990, is substantial cleanup work beginning.

A New Jersey site … was listed on the NPL in 1982. Substantive cleanup did not start until 1990 and was temporarily halted shortly thereafter, due to discovery that the primary subcontractor lacked adequate insurance.

Cleanup decisions were made in 1986 on a Delaware site, listed second on the NPL in 1981. But EPA now expects the site will not be completely cleaned up until 1995, 14 years after being placed on the NPL.⁷⁹

Obviously not all of the delay is attributable to the time spent determining and allocating liability under CERCLA. However, significant delay at each stage of the Superfund process is the inevitable result of litigation, or negotiation to avoid litigation, necessitated by the statutory framework.⁸⁰

VII. Prospects for the Reduction of Transaction Costs at Superfund Sites

The transaction costs and delays set forth in this Article are known to Congress, to EPA, and to the public. Numerous attempts have been made to introduce mechanisms into the Superfund process to eliminate the problems. This section of the Article discusses these reforms, and concludes that although they may reduce transaction costs and delays to someextent, such costs and delays are inherent in the current statutory structure. Transaction costs will continue to be a substantial percentage of overall program costs, despite good faith efforts to control them, as long as EPA is charged with raising funds from PRPs for cleanup.

A. De Minimis Settlements

SARA explicitly authorized de minimis settlements — early settlements with minor contributors of waste at a site — in 1986.⁸¹ Congress hoped that small contributors would cash out early in the process and avoid lengthy litigation. A General Accounting Office (GAO) Report of December 1989, noted, however, that EPA has entered into only a few de minimis settlements, primarily because of limited staff training and experience, limited financial resources, and the lower priority accorded this settlement mechanism by Agency personnel.⁸²

In its 90-Day Report,⁸³ EPA responded to these and other criticisms. The Agency stated that it would continue to encourage the use of de minimis settlements and other settlement tools by providing additional assistance and specialized training to its attorneys and project managers, by developing an incentive system that provides additional support for regional offices to use de minimis and other settlement tools, and by establishing specific goals for the use of de minimis settlements and other settlement tools.⁸⁴

The prospects are slim, however, that de minimis settlements will significantly reduce transaction costs. First, de minimis actions are a strain on Agency resources. To litigate and settle with de minimis PRPs requires a substantial effort by Agency personnel; EPA will usually still have to negotiate, litigate, and settle separately with at least one other group of major PRPs.⁸⁵

Second, although de minimis settlements eliminate the smaller contributors, they do not fundamentally speed up the site cleanup process. Troublesome issues regarding cost allocation, insurance litigation, and liability of the major parties will remain unresolved. Consequently, protracted problems related to negotiation, litigation, and other activities that create transaction costs will persist.

B. Mixed Funding

SARA also explicitly authorized mixed funding — that is, joint EPA/PRP financing — of cleanups to expedite the settlement process.⁸⁶ As with de minimis settlements, GAO noted that this tool has also been used less frequently than contemplated for reasons such as limited training and experience, limited resources, and low priority.⁸⁷ EPA officials [21 ELR 10425] queried by GAO also stated that mixed-funding settlements were less frequently used because "(1) settling parties agreed to full settlement and (2) identified PRPs accounted for most waste or had enough resources to clean up sites."⁸⁸ Other EPA enforcement officials and attorneys told GAO that they did not feel comfortable providing PRPs with federal funds to clean up a site for which the officials felt PRPs were responsible.⁸⁹

In other words, EPA officials have argued that mixed funding is inconsistent with Superfund's joint and several liability scheme. EPA's enforcement strategy is to insist on a 100 percent cleanup offer from PRPs. Generally, PRPs themselves must pursue recalcitrants. In addition, there are administrative burdens. EPA must meet federal procurement regulations⁹⁰ that often inhibit use of mixed funding.

EPA stated in the 90-Day Report that it intended to provide additional assistance and specialized training to Superfund enforcement personnel in the use of the mixed funding option.⁹¹ However, once again, these improvements are unlikely to significantly curtail the delays and transaction costs inherent in the Superfund statutory scheme.

C. NBARs

Another settlement tool explicitly authorized by the SARA amendments is a nonbinding allocation of responsibility (NBAR).⁹² An NBAR is a statement from EPA containing the identities of PRPs, the nature and volume of substances contributed by each party, and a ranking of the substances by volume.⁹³ NBARs are presumed to facilitate the settlement process and to reduce transaction costs by eliminating litigation on the apportionment of responsibility.⁹⁴

NBARs appear to have been largely ignored. EPA personnel questioned by GAO stated that they believed NBARs were appropriately used, but also stated that they had little or no experience with them.⁹⁵ The 1989 GAO Report itself discusses EPA's various settlement tools but makes no mention of NBARs.⁹⁶ EPA's 90-Day Report mentions them only in a heading, neither explaining why they have not been used nor outlining procedures to increase future NBAR use.⁹⁷ Of course, NBARs simply shift the burden of data collection and analysis from PRPs to EPA, and they are subject to further debate both within EPA and among PRPs. Moreover, it is not clear that PRPs desire NBARs. However, as long as the liability-based system is in place, allocation will remain a contentious issue and impose significant costs on the system.

D. Model Consent Decrees

Another possible tool to reduce transaction costs is the development of model consent decrees and other model documents. However, drafting model documents that contain terms upon which both PRPs and EPA can agree and rely is nearly impossible. As expected, model documents proposed by EPA contain provisions most favorable to the Agency. In fact, the recent EPA-drafted model RI/FS proposal⁹⁸ is considered by PRPs to be so onerous that many PRPs have stated they will refuse to enter into settlement discussions under the terms of the model. Moreover, all sites raise new and unanticipated issues that, given the nature of the process, will still involve negotiations and litigation among PRPs and EPA.

E. Alternative Dispute Resolution

Alternative dispute resolution has been used with limited success at Superfund sites, primarily because of the scientific and technical uncertainties involved in evaluating the contamination of groundwater and other issues and EPA's hesitancy to enter into a process that may undercut its enforcement authorities. Neither EPA nor PRPs are comfortable with the prospect of settling for what could ultimately prove to be only a small fraction of the total cleanup costs.⁹⁹

F. Covenants Not to Sue

EPA is authorized to provide any party with a covenant not to sue, immunizing the PRP from CERCLA liability to the United States, if the covenant would expedite the cleanup and meet certain other statutory requirements.¹⁰⁰ This tool has not had much of an impact. Except in limited situations defined by statute, EPA will insist on including reopener provisions in a consent decree if it grants a covenant not to sue, permitting the Agency to reopen the decree in response to conditions unknown at the time the decree was signed. The covenant usually will apply only to "covered matters," and not, typically, to such "noncovered matters" as criminal liability, natural resource damages, liability of the PRP for off-site disposal of wastes, and liability to states.

G. Other Suggested Mechanisms to Reduce Transaction Costs

Several other methods could be brought into play to reduce transaction costs. These include rigid adherence to internal deadlines by which milestones in the cleanup process must be reached;use of a cadre of experienced and knowledgable [21 ELR 10426] lawyers and consultants both in and outside of government to pursue Superfund cleanups; and development of protocols for dispute resolution. These suggested improvements, however, are only marginally helpful and for several reasons they are unlikely to make much of a dent in resolving the enigma surrounding Superfund transaction costs and delays described in this Article.

First and foremost, these incremental reforms do not address the fundamental structure of the Superfund statute — it is a liability-based scheme with extremely high stakes that forces parties to litigate or prepare for litigation at every step in the process.

Second, the process remains adversarial, with EPA extracting cleanup funds on a site-by-site basis. Although the liability scheme is strict, joint, and several, and the enforcement mechanisms EPA has access to are powerful, the PRPs retain legal options both to resist EPA and to seek to impose cleanup costs on other parties. It is unrealistic to conclude that PRPs will not continue vigorously to exercise these options in their own self-interest.

Finally, the government's ability to control the process is limited, even under an aggressive enforcement-first policy. In seeking to discipline the process, the government may risk causing widespread resistance among PRPs. If settlements cannot be reached in a significant number of cases EPA has targeted for enforcement, and if PRPs do not comply voluntarily with cleanup orders issued by the Agency in those cases, the Superfund cleanup system may face substantial difficulties. The Agency's enforcement resources are limited, and not enough money is currently available in the Trust Fund to clean up more than a few sites. Thus, EPA can go only so far in enforcing strict deadlines or dictating the terms of settlement without driving PRPs from the bargaining table in numbers sufficient to undercut the entire enforcement program. The government is even more limited in its ability to control behavior among PRPs themselves, and between PRPs and their insurers.

VIII. Conclusion

Transaction costs and delays are embedded in the Superfund statutory scheme. The costs are substantial and do not directly contribute to the cleanup of Superfund sites. Although EPA has made serious efforts to minimize transaction costs, particularly those causing delays in the cleanup process, these reform measures cannot address underlying problems inherent in the statutory structure. These underlying problems include protracted and complex discovery, negotiation, and litigation involving EPA, state agencies, PRPs, insurers, and municipalities, as well as the necessity for parties to amass documents in an effort to compile an accurate, all-encompassing record. Fundamental reform of the Superfund cleanup system is needed if the problems caused by these costs and delays are to be met head on.

1. Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), 42 U.S.C. §§ 9601-9675, ELR STAT. CERCLA 001-075.

2. Hearings Before the Subcomm. on Policy Research and Insurance of the House Comm. on Banking, Finance and Urban Affairs, 101st Cong., 2d Sess. at 3 (Sept. 27, 1990) (testimony of Dr. Joel S. Hirschhorn) [hereinafter Hirschhorn testimony].

3. Pub. L. No. 96-510, 94 Stat. 2767 (codified as amended at 42 U.S.C. §§ 9601-9675, ELR STAT. CERCLA 001-075).

4. Pub. L. No. 99-499, 100 Stat. 1613 (Oct. 17, 1986).

5. See 26 U.S.C. §§ 59A, 4611, and 4661. The Superfund program was recently reauthorized through September 30, 1994, with an additional Superfund tax of $ 5.6 billion; the Trust Fund was extended through December 31, 1995. Pub. L. No. 101-508, tit. XI, § 11231(d), 104 Stat. 1388-455 (Nov. 5, 1990).

6. See CERCLA §§ 104 and 107, 42 U.S.C. §§ 9604 and 9607, ELR STAT. CERCLA 012 and 024. These actions are known as "fund lead" because the Trust Fund pays the costs of cleanup in the first instance.

7. See CERCLA § 106, 42 U.S.C. § 9606, ELR STAT. CERCLA 024. These actions are known as "enforcement lead" because EPA's efforts to enforce the liability sections of CERCLA compel PRPs to fund the cleanup.

8. Id. See United States v. Carolina Transformer Co., 739 F. Supp. 1030, 20 ELR 20935 (D.N.C. 1989) (defendants jointly and severally liable for three times the response costs incurred by the government at the site).

9. Hearings Before the Senate Comm. on Environment and Public Works, 101st Cong., 1st Sess. at 8 (June 15, 1989) (statement of William K. Reilly, Administrator, U.S. EPA) [hereinafter Reilly Testimony].

10. United States v. Monsanto, 858 F.2d 160, 19 ELR 20085 (4th Cir. 1988) (CERCLA establishes a strict liability scheme; liability is joint and several; and liability may be imposed retroactively without violating due process).

11. See U.S. ENVIRONMENTAL PROTECTION AGENCY, FREQUENCY BY PARTY NAME REPORT (Sept. 10, 1990).

12. CERCLA § 107, 42 U.S.C. § 9607, ELR STAT. CERCLA 024. The same parties may be liable under CERCLA § 106, 42 U.S.C. § 9606, ELR STAT. CERCLA 024.

13. See, e.g., STUDIES AND ANALYSIS GROUP OF THE MAJORITY STAFF OF SENATE COMM. ON THE BUDGET, 101ST CONG., 2D SESS., MANAGEMENT OF SUPERFUND (Aug. 1990) [hereinafter SASSER REPORT]; SENATE SUBCOMM. ON SUPERFUND, OCEAN AND WATER PROTECTION, 98TH CONG., 2D SESS., LAUTENBERG-DURENBERGER REPORT ON SUPERFUND IMPLEMENTATION: CLEANING UP THE NATION'S CLEANUP PROGRAM (May 1984) [hereinafter LAUTENBERG-DURENBERGER REPORT]; U.S. GENERAL ACCOUNTING OFFICE, SUPERFUND, A MORE VIGOROUS AND BETTER MANAGED ENFORCEMENT PROGRAM IS NEEDED (Dec. 1989) [hereinafter GAO REPORT]; U.S. CONGRESS, OFFICE OF TECHNOLOGY ASSESSMENT, COMING CLEAN, SUPERFUND PROBLEMS CAN BE SOLVED (Oct. 1989) [hereinafter OTA REPORT]; U.S. ENVIRONMENTAL PROTECTION AGENCY, A MANAGEMENT REVIEW OF THE SUPERFUND PROGRAM (1989) (William K. Reilly, Administrator) [hereinafter 90-DAY REPORT].

14. OTA REPORT, supra note 13, at 18.

15. 40 C.F.R. § 300 app. B (1990).

16. Note that the only part of Superfund that has consistently received praise is the "removal" program under which EPA moves, typically at 100 percent federal expense, to activate its cleanup contractors to eliminate an actual or imminent threat to public health at a hazardous waste site. CERCLA § 104, 42 U.S.C. § 9604, ELR STAT. CERCLA 012. See SASSER REPORT, supra note 13, at 4 ("In the removal stage, EPA generally moves quickly to stabilize and contain toxic waste sites when they are reported."). While this authority is limited by statute to $ 2 million and to not more than one year, it can be extended if conditions continue to exist or if the removal action is consistent with the long term remedial action. CERCLA § 104(c), 42 U.S.C. § 9604(c), ELR STAT. CERCLA 013.

17. RAND CORPORATION, INSTITUTE FOR PUBLIC JUSTICE, PUB. NO. R-3838-ICJ, UNDERSTANDING SUPERFUND, A PROGRESS REPORT 16 (1989) (Jon Paul Acton) [hereinafter RAND REPORT].

18. 56 Fed. Reg. 5598, 5634 (Feb. 11, 1991).

19. CERCLA § 113(a), 42 U.S.C. § 9613(a), ELR STAT. CERCLA 038.

20. Id. § 104(a), 42 U.S.C. § 9604(a), ELR STAT. CERCLA 012. See 55 Fed. Reg. 9689 (Mar. 14, 1990); 55 Fed. Reg. 35504 (Aug. 30, 1990).

21. See OFFICE OF SOLID WASTE MANAGEMENT AND EMERGENCY RESPONSE, U.S. EPA, NO. 9832.13, THE SUPERFUND COST RECOVERY STRATEGY at 13 (July 29, 1989). See also LAUTENBERG/DURENBERGER REPORT, supra note 13, at 164.

22. 42 U.S.C. § 9606, ELR STAT. CERCLA 024.

23. INFORMATION NETWORK FOR SUPERFUND SETTLEMENTS, EPA's ENFORCEMENT FIRST STRATEGY FOR SUPERFUND IMPLEMENTATION at SSS-112 (Dec. 1, 1990).

24. 5 U.S.C. § 552, ELR STAT. APA 011.

25. Id.; 40 C.F.R. § 2.112 (1990) (EPA must respond to FOIA requests within 10 working days of receipt).

26. See U.S. v. Acton Corp., 733 F. Supp. 869, 873, 20 ELR 21191, 21193 (D.N.J. 1990) (noting that the Lone Pine landfill case involved more than 300 PRPs).

27. Lone Pine Steering Committee v. U.S. EPA, 600 F. Supp. 1487, 16 ELR 20009 (3d Cir. Nov. 22, 1985).

28. 42 U.S.C. § 9604(e), ELR STAT. CERCLA 015.

29. EPA may alternatively send PRPs a "special notice" stating that EPA will begin work at the site unless the PRPs agree within 60 days to make a proposal for "undertaking or financing" the cleanup. CERCLA § 122(e), 42 U.S.C. § 9622(e), ELR STAT. CERCLA 056.

30. INFORMATION NETWORK FOR SUPERFUND SETTLEMENTS, PRP ORGANIZATION HANDBOOK — A GUIDE FOR POTENTIALLY RESPONSIBLE PARTIES AT SUPERFUND SITES (June 1989).

31. Id. at 19.

32. Id.

33. See, e.g., United States v. Cannons Eng'g Corp., 899 F.2d 79, 20 ELR 20845 (1st Cir. 1990) (upheld consent decree apportioning generators' liability on the basis of volumetric share).

34. U.S. ENVIRONMENTAL PROTECTION AGENCY, INTERIM POLICY ON CERCLA SETTLEMENTS INVOLVING MUNICIPALITIES OR MUNICIPAL WASTE, 54 Fed. Reg. 51010 (Dec. 12, 1989), ELR ADMIN. MATERIALS 35225 [hereinafter EPA INTERIM POLICY].

35. 42 U.S.C. §§ 9607, 9613(f), ELR STAT. CERCLA 024, 039.

36. INSIDE EPA'S SUPERFUND REPORT, Nov. 21, 1990, at 6.

37. Transportation Leasing Co. v. California, 21 ELR 20826 (C.D. Cal. Dec. 10, 1990); B.F. Goodrich v. Murtha, 754 F. Supp. 960, 21 ELR 20777 (D. Conn. Jan. 8, 1991), permission to appeal from interlocutory order granted, No. 91-7450 (2d Cir. May 8, 1991). See also U.S. v. Kramer, No. 89-4340 (JFG) (D.N.J. Feb. 8, 1991) (EPA's decision not to name local governments as PRPs in an action under CERCLA § 107 does not insulate them from allegations of third-party liability).

38. CERCLA requires that EPA also establish an administrative record for future litigation. CERCLA § 113(k)(1), 42 U.S.C. § 9613(k)(1), ELR STAT. CERCLA 040. EPA has promulgated rules requiring establishment of such a record. 40 C.F.R. § 300.800 (1990) (administrative record for selection of response action).

39. OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE, U.S. EPA, NO. 9355.4-03, CONSIDERATIONS IN GROUND WATER REMEDIATION AT SUPERFUND SITES (Oct. 18, 1989).

40. 55 Fed. Reg. 35511 (Aug. 30, 1990).

41. Hearings Before the Subcomm. on Superfund, Ocean, and Water Protection of the Senate Comm. on Environment and Public Works, 101st Cong., 2d Sess. at 19 (June 21, 1990) (testimony of Donald R. Clay, Assistant Administrator for the EPA Office of Solid Waste and Emergency Response).

42. 42 U.S.C. § 9604(a)(1), ELR STAT. CERCLA 012.

43. Mr. Clay took this position in response to questions from members of Congress in his June 21, 1990, testimony, supra note 41. Several industry trade associations have challenged the Agency's position because, they argue, it is a de facto rule that should be subject to notice and comment rulemaking. See Chemical Mfrs. Ass'n v. EPA, No. 90-1460, (D.C. Cir. petitioners' motion for summary remand filed Nov. 2, 1990), digested at ELR PEND. LIT. 66124, 66129.

44. CERCLA § 122(e), 42 U.S.C. § 9622(e), ELR STAT. CERCLA 056.

45. CERCLA § 104(c)(3), 42 U.S.C. § 9604(c)(3), ELR STAT. CERCLA 013-014.

46. See RAND REPORT, supra note 17, at 16 (average time spent on RDs was 18 months, and RAs was 25 months, although many RDs and RAs were not complete at the time of the report). See also Reilly Testimony, supra note 9, at 4 (EPA expects it will take 13 years to begin construction on all sites currently on the NPL).

47. 42 U.S.C. §§ 9607, 9613(f), ELR STAT. CERCLA 024, 039.

48. New York v. Ludlow's Sanitary Landfill, No. 86-CV-853 (N.D.N.Y. fourth-party complaint filed Oct. 10, 1990) (digested at ELR PEND. LIT. 66110).

49. Id.

50. See United States v. Cannons Eng'g Corp., 720 F. Supp. 1027, 20 ELR 20159 (D. Mass. 1989), aff'd 899 F.2d 79, 20 ELR 20845 (1st Cir. 1990).

51. Prudential Ins. Co. v. U.S. Gypsum, 711 F. Supp. 1244, 19 ELR 21220 (D.N.J. 1989) (manufacturer of products containing asbestos did not arrange for disposal of those products); Florida Power & Light Co. v. Allis Chalmers Corp., 893 F.2d 1313, 20 ELR 20523 (11th Cir. 1990) (sale by manufacturer of transformers containing PCBs was not arrangement for disposal of PCBs); Edward Hines Lumber Co. v. Vulcan Materials Co., 685 F. Supp. 651, 18 ELR 21223 (N.D. Ill.), aff'd 861 F.2d 155, 19 ELR 20187 (7th Cir. 1988) (sale of creosote for the treatment of lumber was not an arrangement for disposal of creosote).

52. New York v. General Elec. Co., 592 F. Supp. 291, 14 ELR 20719 (N.D.N.Y. 1984) (sale of spent transformer oil to racetrack for dust control constituted an arrangement for disposal); United States v. A&F Materials Co., 582 F. Supp. 842, 14 ELR 20432 (S.D. Ill. 1984) (manufacturer of jet aircraft that sold spent caustic to oil reclamation facility arranged for disposal of the caustic).

53. Compare United States v. Fleet Factors Corp., 901 F.2d 1550, 20 ELR 20832 (11th Cir.), cert. denied 59 U.S.L.W. 3481 (U.S. Jan. 14, 1991) (No. 90-504) (secured creditors may be liable as owners or operators by participating in financial management of facility to a degree indicating ability to influence facility's waste handling) with In re Bergsoe Metal Corp., 910 F.2d 668, 20 ELR 21229 (9th Cir. 1990) (creditors not liable as owners or operators).

54. EPA INTERIM POLICY, supra note 34; Anderson v. City of Minnetonka, No. CV 3-90-312 (formerly No. 4-90-371) (D. Minn. City of Deephaven's memorandum in support of summary judgment filed Apr. 24, 1991), digested at ELR PEND. LIT. 66141; U.S. v. Kramer, No. 89-4340 (JFG) (D.N.J. Feb. 8, 1991); B.F. Goodrich v. Murtha, 754 F. Supp. 960, 21 ELR 20777 (D. Conn. Jan. 8, 1991), permission to appeal from interlocutory order granted, No. 91-7450 (2d Cir. May 8, 1991); Transportation Leasing Co. v. California, 21 ELR 20826 (C.D. Ca. Dec. 10, 1990); New York v. Ludlow's Sanitary Landfill, No. 86-CV-853 (N.D.N.Y. fourth-party complaint filed Oct. 10, 1990), digested at ELR PEND. LIT. 66110; New Jersey v. Almo Antipollution Sources Corp. No. 89-4380G (D.N.J. filed Oct. 16, 1989).

55. See Joslyn Corp. v. T. L. James & Co., 696 F. Supp. 222, 19 ELR 20518 (W.D. La. 1988) (corporate subsidiary could not pierce the corporate veil to impose liability for response costs on parent corporation), aff'd 893 F.2d 80, 20 ELR 20382 (5th Cir. 1990), petition for cert. denied 59 U.S.L.W. 3564 (U.S. Feb. 19, 1991) (No. 89-1973).

56. See United States v. Pacific Hide & Fur Depot, 716 F. Supp. 1341, 19 ELR 20897 (D. Idaho 1989) (innocent landowner defense applied in CERCLA cost recovery action to private parties who inherited stock of corporation that owned property upon which a landfill was operated).

57. In recent testimony before Congress, an insurance industry spokesperson noted, for example, that the number of sites at issue in Allied Signal, Inc. v. Abeille-Paix Reassurances, No. MRS-L-226-88 (N.J. Super. Ct. Law Div., Morris County, filed Jan. 26, 1988), is 370; in United Technologies Corp. v. Liberty Mutual Ins. Co., No. 87-7172 (Mass. Super. Ct., Suffolk County, filed Dec. 24, 1987), the number is at least 145; in Monsanto Co. v. Aetna Casualty & Surety Co., No. 88C-JA-118 (Del. Super. Ct., New Castle County, filed Jan. 20, 1988), the number is 80; and in Champion Int'l Corp. v. Liberty Mutual Ins. Co., Nos. 87 Civ. 1634 and 88 Civ. 5041 (S.D.N.Y. filed Mar. 11, 1987 and July 19, 1989) and Champion Int'l Corp. v. Aetna Casualty & Surety Co., No. 90-2-09616-5 (Wash. Super. Ct., King County, filed May 11, 1990), the number is up to 73. Hearings Before the Subcomm. on Policy Research and Insurance of the House Comm. on Banking, Finance and Urban Affairs, 101st Cong., 2d Sess. at 3 (Oct. 11, 1990) (testimony of Thomas W. Brunner, Counsel for the Insurance Environmental Litig. Ass'n) [hereinafter Brunner Testimony].

58. For example, 110 attorneys have entered appearances in Monsanto; 196 in American Centennial Ins. Co. v. Union Carbide Corp., Nos. 89-3244, 89-4367, 90-1014 (N.Y. Sup. Ct., Suffolk County, filed July 19, 1989); 96 in United Technologies; 65 in Allied-Signal; and at least 60 in Shell Oil v. Accident & Casualty Ins. Co. of Winterthur, No. 278953 (Cal. Super. Ct., San Mateo County, filed Oct. 5, 1983). Moreover, these are just the lead attorneys. Hundreds of other attorneys are also involved in these cases supporting the lead attorneys. There are literally hundreds of depositions recording the testimony of witnesses taken in these cases. For example, in Shell, 232 witnesses were deposed in a one-year period. More than 100 depositions have been taken, or are expected to be taken, in each of the Union Carbide, United Technologies, Allied-Signal, and Monsanto cases. Brunner Testimony, supra note 57, at 3.

59. The amount in dispute in Shell is $ 1.8 billion; up to $ 2 billion in Allied Signal; more than $ 1 billion in E.I. DuPont de Nemours & Co. v. Admiral Ins. Co., No. 89-C-AV-99 (Del. Super. Ct., New Castle County); and more than $ 400 million in Monsanto. Brunner Testimony, supra note 57, at 3.

60. AIU Ins. Co. v. Superior Court of Santa Clara County, 799 P.2d 1253, 21 ELR 20315 (Cal. Nov. 15, 1990) (response costs are covered damages). Compare Continental Ins. Co. v. Northeastern Pharmaceutical and Chem. Co., 811 F.2d 1180, 17 ELR 20616 (8th Cir. 1987) (costs of cleanup are recoverable damages), reh'g en banc, 842 F.2d 977, 18 ELR 20819 (8th Cir. 1988) with Maryland Casualty Co. v. Armco, Inc., 822 F.2d 1348, 17 ELR 21277 (4th Cir. 1987) (response costs are not covered damages).

61. See American Motorists Ins. Co. v. General Host Corp., 667 F. Supp. 1423 (D. Kan. 1987) (releases were not unexpected and therefore coverage was denied).

62. See, e.g., National Grange Mut. Ins. Co. v. Continental Casualty Ins. Co., 650 F. Supp. 1405 (S.D.N.Y. 1986) (pollution exclusion clause does not relieve insurer from obligation to defend).

63. See Technician Elecs. Corp. v. American Home Assurance Co., 544 N.Y.S.2d 531 (Ct. App. 1989) (court held long-term discharge of toxic chemicals by insured not covered).

64. Id.

65. Compare State Farm Mut. Auto. Ins. Co. v. Coleman, 441 F.2d 329 (5th Cir. 1971) (notice one year after occurrence was not timely and insurer was released as a matter of law) with Colonial Gas Energy Sys. v. Unigard Mut. Ins. Co., 441 F. Supp. 765 (N.D. Cal. 1977) (insurer not released as a matter of law when notice was given almost two years after occurrence).

66. See, e.g., Westinghouse Elec. Corp. v. Aetna Casualty-Surety Co., 227 N.J. Super. 504, 547 A.2d 1167 (1988) (in two consolidated cases, Westinghouse claims against 140 insurance companies for declaratory judgment of coverage related to 81 environmental sites dismissed on grounds of forum non conveniens, except as to nine sites located in New Jersey; Westinghouse claims against more than 100 insurance carriers for declaratory judgment of coverage related to 2,675 asbestos actions dismissed on grounds of forum non conveniens, except as to 128 actions filed in New Jersey).

67. Hirschhorn testimony, supra note 2, at 3.

68. Reilly Testimony, supra note 9, at 4.

69. Hearings Before the Subcomm. on Policy Research and Insurance of the House Comm. on Banking, Finance and Urban Affairs, 101st Cong., 2d Sess. at 5 (Sept. 27, 1990) (statement of James M. Strock).

70. Id. at 5.

71. EXECUTIVE OFFICE OF THE PRESIDENT, BUDGET OF THE U.S. GOVERNMENT, FISCAL YEAR 1992, at 2-9b.

72. OTA REPORT, supra note 13, at 29.

73. Id.

74. See, e.g., RAND REPORT, supra note 17; LAUTENBERG-DURENBERGER REPORT, supra note 13; SASSER REPORT, supra note 13.

75. RAND REPORT, supra note 17, at 16-17.

76. While the Rand Report chronicles the actual time taken for each step in the process, EPA has prepared its own suggestion for the ideal time frames for various milestones. OFFICE OF SOLID WASTE MANAGEMENT AND EMERGENCY RESPONSE, U.S. EPA, NO. 9851.3, INTEGRATED TIMELINE FOR SUPERFUND SITE MANAGEMENT (June 11, 1990). EPA estimates that PRP search and notification will take between 18 and 39 months. RI/FS negotiation and settlement will continue for one year. The RI/FS will be completed, and the ROD will be signed at the end of 24 months. RD/RA settlement/referral and implementation will take at least another 18 months depending upon the nature of the settlement process. Because some of these milestones overlap, EPA's time frame provides for approximately 4-1/4 years for the process before design and construction of the cleanup can begin. Id.

77. SASSER REPORT, supra note 13, at 5-6.

78. Id. at 6. There are currently 1,189 sites on the NPL. 56 Fed. Reg. 5598, 5599 (Feb. 11, 1991).

79. Id. at 2, 6-7.

80. EPA has recognized that conflicting incentives are built into the Superfund process. See 90-DAY REPORT, supra note 13. For example, EPA noted that the desire for a prompt cleanup often conflicts with the need for full and responsive public participation in the cleanup process. Quick cleanup, EPA argues, also requires more regional autonomy, which might come at the expense of a consistent national program direction. Finally, EPA noted that it can "pursue either complete cleanup at some sites, or incremental cleanup at many sites, but cannot fully accommodate both goals simultaneously." Id. at 5.

81. CERCLA § 122(g), 42 U.S.C. § 9622(g), ELR STAT. CERCLA 058.

82. GAO REPORT, supra note 13, at 51.

83. 90-DAY REPORT, supra note 13.

84. Id. at 2-9 to 2-12.

85. See United States v. Cannons Eng'g Corp., 899 F.2d 79, 20 ELR 20845 (1st Cir. 1990) (EPA negotiated separate settlements with de minimis PRPs who settled early, de minimis PRPs who settled late, and major PRPs).

86. CERCLA § 122(b)(1), 42 U.S.C. § 9622(b)(1), ELR STAT. CERCLA 054-055. See OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE, U.S. EPA, No. 9225.1-01, PROCUREMENT UNDER PREAUTHORIZATION/MIXED FUNDING (Apr. 19, 1989) [hereinafter PROCUREMENT GUIDANCE].

87. GAO REPORT, supra note 13, at 51.

88. Id. at 53.

89. Id.

90. See PROCUREMENT GUIDANCE, supra note 86, at 3-6.

91. 90-DAY REPORT, supra note 13, at 2-9 to 2-12.

92. CERCLA § 122(e), 42 U.S.C. § 9622(e), ELR STAT. CERCLA 056-057. See Michigan ex rel. Kelley v. Thomas Solvent Co., 717 F. Supp. 507, 517, 20 ELR 20109, 20114 (June 5, 1989) (noting that EPA is authorized to develop NBARs and listing factors the government should consider in apportioning responsibility).

93. CERCLA § 122(e)(1), 42 U.S.C. § 9622(e)(1), ELR STAT. CERCLA 056.

94. OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE, U.S. EPA, INTERIM GUIDELINES FOR PREPARING NONBINDING PRELIMINARY ALLOCATIONS OF RESPONSIBILITY, 52 Fed. Reg. 19919 (May 28, 1987), ELR ADMIN. MATERIALS 35065.

95. GAO REPORT, supra note 13, at 92.

96. Id. at 51-57.

97. 90-DAY REPORT, supra note 13, at 2-9, 2-12. At a recent conference, Bruce Diamond, Director of EPA's Office of Waste Programs Enforcement, confirmed only one NBAR developed by the government. Address by Bruce M. Diamond, Environmental Law Institute Annual Advanced Seminar, Washington, DC (Oct. 4-5, 1990).

98. U.S. EPA, ADMINISTRATIVE ORDER ON CONSENT FOR REMEDIAL INVESTIGATION/FEASIBILITY STUDY (Feb. 5, 1990), ELR ADMIN. MATERIALS 35243.

99. See Richard H. Mays, Alternative Dispute Resolution and Environmental Enforcement, 18 ELR 10087 (Mar. 1988).

100. CERCLA § 122(f), 42 U.S.C. § 9622(f), ELR STAT. CERCLA 057.