Testimony of The Environmental Defense Fund and The Natural Resources Defense
Council on S. 8, the "Superfund Cleanup Acceleration Act of 1997"
(revised Draft Chairman's Mark of 8/28/97)
before the Senate Committee on Environment and Public Works
Karen Florini, Senior Attorney, EDF
accompanied by Jacqueline Hamilton, Senior Project Attorney, NRDC
September 4, 1997

Introduction

On behalf of the Environmental Defense Fund (EDF) and the Natural Resources Defense Council (NRDC), I appreciate this opportunity to discuss the revised version of S. 8, the "Superfund Cleanup Acceleration Act of 1997," amending Superfund. EDF and NRDC have been actively involved in the Superfund reauthorization process, serving on EPA's NACEPT Committee on Superfund and on the National Commission on Superfund, and testifying repeatedly on Superfund during the last two Congresses. Most recently, I testified before this Committee's Subcommittee on Superfund, Waste Control, and Risk Management regarding the initial version of S. 8 on March 5,1997.

While some of the most problematic features in S. 8 as introduced have been moderated,1 we believe that there are still numerous fundamental flaws in the bill as revised, compelling us to continue to oppose the bill in its current form. These include features that will make cleanups less protective, disempower communities (particularly where authorities are transferred to States), and let large industrial polluters escape liability without policy justification and with unacceptable consequences for the future of the cleanup program.

In addition, numerous provisions, taken together, will dramatically slow the pace of cleanups. This is a particularly inopportune time for doing so, given that the Superfund program has finally begun to make substantial progress in recent years.

We urge Congress not to turn back the clock to an earlier era in which Superfund cleanups were abysmally slow. In particular, Congress must not divert funds needed for cleanups to sweeping liability carve-outs for polluters who can well afford to pay to clean up the messes they have made, nor divert EPA's resources by creating unnecessary, time-consuming new tasks. Rather, Congress should:

-- adopt a targeted set of broadly supportable provisions to enhance program effectiveness and public participation,

-- increase funding to allow cleanups to proceed as promptly as is consistent with good decision-making and full public participation, and

-- reinstate the now-defunct taxes that help finance the program.

We would welcome an opportunity to work with the Committee in developing a bill meeting these objectives.

The remainder of this testimony focuses on our principle criticisms of the current version of S. 8, the August 28 "Draft Chairman's Mark." Because the draft bill was made available to us only 4 business days before today's hearing, please note that we may subsequently identify additional concerns.\1\

\1\ In particular, we are glad to see that the "polluters in charge' provisions of S. 8 as introduced - under which polluter-written cleanup plans could have been approved by default- have been deleted. We also applaud the fact that the revised bill no longer allows Potentially Responsible Parties to serve as voting members on the Community Advisory Group [SCAA 303, adding CERCLA 117(h), p. 72]. And we support the provision dropping the requirement in existing law for matching contributions for Technical Assistance Grants (SCAA 303, adding CERCLA 117(g)(2), p. 80], and expressly allowing up-front payments for TAGs.

I. Community Disempowerment: How S. 8 Makes Communities Irrelevant.

Almost everyone agrees that early, robust public participation pays handsome dividends in avoiding controversy - and thus cleanup delays - down the line. Accordingly, it is surprising as well as disappointing that a bill denominated the Superfund Cleanup Acceleration Act would contain numerous provisions that systematically curtail public participation in key contexts, most notably those involving state roles.\2\ While EDF and NRDC do not oppose a greater role in cleanup for states that have adequate resources, authorities, and commitment, this expansion must not occur at the expense of curtailing the public's role in Superfund cleanups.

\2\ Although the community Participation provisions, Title Ill of S.8 are an improvement over existing law, they will be largely irrelevant for all sites transferred to states given the weaknesses of Title II's state roles.

Yet just such curtailment could well result from the state role provisions in Title II of S.8 as revised. Problems with S. 8's state role provisions include the process through which delegation or authorization occurs, as well as the consequences of delegation or authorization. In effect, both communities and EPA are forced to trust that state programs will contain adequate community involvement provisions without any way of verifying that such provisions will indeed be included - and without meaningful recourse if they are not. Similar problems exist in Title I, the Brownfields title. Both are discussed below.

A. Curtailing Community Participation Through Inadequate State Role Criteria and Procedures

Public participation is conspicuous by its absence from the list of criteria for EPA to evaluate in making delegation determinations (SCAA 201, adding CERCLA 130(e)(3)(C), p. 44]. To make matters worse, the bill expressly precludes EPA from including any conditions regarding public participation (or anything else) in approving a delegation request [130(e)(4)(D), p. 46]. Similarly, for authorization, states merely are required to have "procedures to ensure public notice and as an appropriate opportunity for comment" on cleanup plans (130(c)(1)(C), p. 37] - a loophole potentially big enough for a proverbial Mack truck. Similar language exists for Brownfields programs (SCAA 102(b), adding CERCLA 128(b)(2), p. 15].\3\

\3\This weakness in the Brownfield title is especially objectionable because NPL sites are (inappropriately, in our view) eligible to be included under Brownfield programs [SCAA 102(b), adding CERCLA 128(c), p. 16].

Not only is public participation omitted as a criterion for transfer of authorities to States, but the public is excluded from decisions about whether to transfer such authorities. There is no allowance for public notice and comment in proceedings either for delegation or authorization. The public is even more emphatically excluded from participating in the context of "expedited authorization" applications, given that states meeting any 3 of 5 specific criteria are "conclusively presumed" to warrant authorization [130(d)(1), p. 38]. And expedited authorizations cannot be challenged in court (SCAA 130(d)(4)(D), p. 42]. So, if there are grave deficiencies in a state's program, the public will have no opportunity to call these to EPA's attention. Similarly, the public will have no opportunity to voice concerns about state enforcement authorities despite the fact that even delegated states are to use state enforcement authorities (130(e)(1)(C), p. 43].

Moreover, the bill's liberal use of default approvals mean that delegation or authorization can occur without any actual review by EPA of the adequacy of the state program. Decisions about toxic waste dump cleanup programs are too important to be relegated to the flipping of pages on a calendar. This is true regardless of the cause of any bureaucratic delays in making decisions - whether they be due to a personal tragedy that befalls an EPA reviewer, or a change of personnel, or a government-wide shutdown, or even simple inertia. Deadlines play a legitimate role, but default approvals do not.

B. The Draconian Consequences of State Delegation and Authorization

Exclusion of the public from authorization and delegation decisions is particularly troubling because those decisions have profound consequences under 5.8 as revised. For example:

-- State roles override CERCLA's citizen-enforcement provisions, because the bill provides that neither EPA nor any other person can take judicial enforcement action against any person regarding a transferred site (SCAA 201, adding CERCLA 1 30(h)(4), p. 52]. Thus, citizens will be unable to use existing 310 of CERCLA even to enforce cleanup agreements. The Brownfields title contains similar strictures for facilities "subject to" state cleanup plans, apparently regardless of whether those plans meet any criteria whatsoever [SCAA 103, adding CERCLA 129(b)(1), p. 18]. These limitations are radical and unwarranted departures from prior law not only under Superfund, but indeed virtually all federal environmental programs. There is no justification for barring citizen enforcement of Superfund requirements.

-- State delegation/authorization eliminates virtually all EPA authority. The public needs and deserves an effective federal fallback where states fail to carry out their environmental responsibilities appropriately for toxic site cleanups, just as occurs for air and water pollution programs. Yet, extraordinarily, the bill provides that EPA cannot act at a site covered by a delegation agreement unless the agency goes to court and obtains a declaratory judgment that the state has failed to make reasonable cleanup progress (SCAA 201, adding CERCLA 1 30(h)(4)(B)(ii)(II), p. 53]. For a bill that supposedly seeks to accelerate cleanups and reduce litigation, forcing EPA to wait helpless pending completion of a lawsuit against a state is as curious as it is counterproductive. Provisions almost as onerous apply in the Brownfields title, with regard to any facility that is "subject to" a State remedial action plan - despite the fact that EPA apparently has no role in reviewing state remedial programs at all (SCAA 103, adding CERCLA 129(b)(4), p. 18].

All the preceding problems are compounded by the fact that the bill offers EPA no option of partial de-delegation or de-authorization. Instead the only option is the "nuclear" one of total program withdrawal - a seldom-used tactic.

C. Other Features That Undercut Effective Public Participation.

In addition to the state role and Brownfields provisions discussed above, several other features of the bill undercut public participation as well. These are briefly discussed below.\4\

\4\ In addition, the presumptive remedies section as now written further disempowers communities. While the concept of presumptive remedies can be beneficial, S.8 as revised seems to make the presumption an irrefutable one - regardless of community concerns. Specifically, the bill provides that the Administrator may select a presumptive remedial action without consideration of (other) technologies, approaches, or methodologies' (SCAA 403, adding CERCLA 1 32(c)(2), p. 120]. This could be read to allow the Administrator to adopt a presumptive remedy regardless of community views at a particular site. At the same time, the bill specifies that identification of presumptive remedies does not constitute rulemaking and need not go through public notice and comment procedures (1 32(b)(3), p. 120]. The text of the bill should make clear that nothing in the presumptive remedy section authorizes EPA to disregard comments and alternative remedies suggested by interested parties-at sites for which presumptive remedies exist.

1. Shifting the Public Out through Silent Vetoes.

Yet another way the public is shut out of meaningful participation arises from provisions under which new sites can be added to the Superfund list "only with the concurrence of the Governor of the State" in which the sites are located [SCAA 802, adding CERCLA 105(i)(3), p. 246]. Similarly, a state can block any administrative cleanup order under 106 by failing to concur within 90 days (SCAA 103, adding CERCLA 129(c), p. 20].

While it may be appropriate to give states "first dibs" on cleanups at sites that will be appropriately addressed through state action, these provisions go much too far. A state could, through simple inaction, bar an NPL listing or a 106 order even though the site will not otherwise be cleaned up. The State need not even give any reasons for failure to concur, inviting potential abuses (if, for example, a major PRP at the site also happened to be a campaign contributor to a high-ranking State official). Moreover, these provisions invite creation of "pollution havens" by Governors seeking to lure business from other states by declaring an indefinite moratorium on NPL listings. EPA should defer to a state program only upon affirmatively determining that the State will conduct an adequate, timely cleanup absent the listing or 106 order.

2. Shifting the Public Out of Cleanup Decision Revisions.

As written, the bill's provisions for reopening existing cleanup decisions essentially eliminate opportunities for effective public participation. Given that review boards are to complete their review within 180 days (SCAA 406, adding CERCLA 136(d), p. 144], communities often will have inadequate time to receive notice and respond. This is particularly true at sites where no Technical Assistance Grant is currently in place. (Even where TAGs already exist, the limited number of community-oriented technical experts would be unable to provide effective support if large numbers of reopener petitions are submitted - a possible outcome under the bill as now drafted, see section Ill below.) To assure that the public is meaningfully involved, the Administrator should be able to extend the deadline for the Board to complete its review.

II. S. 8's Inadequate Cleanup Provisions

Although the revised version of 5.8 has dropped the egregious provisions that let polluters run the cleanup decision-making process, the remedy title still has several major deficiencies that make it highly objectionable. These include a preference for treatment of "hot spots" that is worse than useless; critical omissions from the range of cleanup objectives; and important weaknesses in the cleanup standards themselves. These are discussed in turn below.

A. The Hot Spots "Preference"

Given current EPA practice of cleanup up to unrestricted use at only one-third of sites even with the existing preference for treatment, we have increasing reservations about whether there is any rationale for changing this portion of the law. However, if the preference for treatment contained in current law is to be narrowed, it is essential to provide a preference for treatment of "hot spots." While S. 8 as revised now includes such a preference, as currently drafted it applies only when contaminants "cannot reliably be contained" and "present substantial risk... because of high toxicity ... and high mobility" and there is "a reasonable probability of actual exposure based on site- specific factors" (SCAA 402, amending CERCLA 121(c)(3), p. 108].

Such an approach is highly objectionable because it implies that treatment will occur only when these onerous and unworkable requirements are met. More generally, this approach entirely misses the point of having a preference for treating hot spots: to avoid intrinsically uncertain guesstimates about whether material cannot reliably be contained, and whether and how future exposures will occur. Because it is impossible to see into the future with the level of confidence these phrases suggest, a preference for treatment is vital.

Another approach to this issue may also warrant consideration. The current statute's preference for treatment and mandate for permanent remedies have caused problems primarily at sites with high volumes of low-toxicity wastes. That problem could be dealt with explicitly, by maintaining the preference for treatment while creating an exception for high-volume, low-toxicity sites. Rather than making containment the rule and hot spots the exception, Superfund would maintain treatment as the rule and make the problematic type of sites (high-volume, low toxicity) the exception.

B. Weaknesses in Institutional Controls Provisions

Even with Superfund's existing mandate for permanence and broad preference for treatment, many sites have been cleaned up only part way, to a degree that allows for some but not all types of use of land or water (e.g., industrial use only, or no excavation). To assure that restricted-use sites are in fact only used in a manner consistent with their restrictions, legal mechanisms known as "institutional controls" may be employed. Unfortunately, while S. 8 would do much to increase the prevalence of restricted-use sites,\5\ it provides no real assurance that any institutional controls adopted as part of such cleanups will actually work. As experience at Love Canal itself amply illustrates, institutional controls that fail can be a disaster on many fronts.\6\

\5\ For example, the bill expressly provides that use of institutional controls is `to be on an equal basis with all other remedial action alternatives" (SCAA 402, amending CERCLA 121 (c)(4)(E), p. 112] - despite the fact that institutional controls are inherently more uncertain than treatment-based remedies. The bill also requires facility-specific risk evaluations to "consider the use of institutional controls" (SCAA 403, adding CERCLA 131(b)(1)(D), p. 116].

\6\ Love Canal is a classic illustration of the failure of institutional controls. The deed contained a notice of the presence of chemical wastes, but a subsequent owner eventually disturbed the waste when doing construction on the site. See U.S. v. Hooker Chemicals & Plastics Corp. , 722 F. Supp. 960,962 (W.D.N.Y 1989).

Among other problems in this critical section, the definition of "institutional controls" is itself overly broad (SCAA 402, amending CERCLA 121(c)(4)(A), p. 1 10]. While zoning, land use plans, and notification systems may be extremely valuable as supplements to institutional controls, these devices are too ephemeral and/or too weak to serve as institutional controls in this context: protecting human health and the environment from the effects of toxic contaminants left on land or in water after cleanup activities are "complete."

Similarly, the bill's current "requirements" for institutional controls - that they are "adequate to protect human health and the environment," "ensure... long-term reliability," and "will be appropriately implemented, monitored, and enforced" - are far too vague to be meaningful (SCAA 402, amending CERCLA 121(c)(4)(c), p. 112]. Rather, the bill must explicitly require that specific criteria be met for any institutional control that is adopted as part of a remedy. These include, at a minimum:

-- permanence (i.e., the control will remain in effect until removed following an affirmative, site-specific determination that it is no longer needed because the contamination is gone);

-- universality (i.e., applies to all current and future interest-holders of the land or water);

-- enforceability (i.e., by all interested parties, including citizens); and

-- permanent notice (i.e., in land records unless inappropriate given the specific nature of the control).

Given the Byzantine complexity of much of American property law, some jurisdictions may lack mechanisms that meet these criteria. Congress should create an array of federal institutional controls to assure that qualifying mechanisms are available in all jurisdictions. The only other alternatives are either unlikely (disallowing institutional controls in jurisdictions that lack qualifying controls and requiring that all sites be remediated to unrestricted use) or intolerable (allowing use of inadequate institutional controls).

C. Weaknesses in Cleanup Standards

The cleanup standards in S. 8 continue to commit critical sins of omission. In particular, there is still no explicit requirement for protecting the health of children and other highly susceptible or exposed groups.\7\ Likewise, "protection of health" is still defined as a cancer risk in the range of 10-4 to 10-6 [SCAA 402, amending CERCLA 121(a)(1)(B)(i)(l), p. 85], but without the National Contingency Plan's provision specifying that 10-6 is the "point of departure." As a result, cost considerations are likely to tilt remedies toward the less-protective outcome, since cleaning up to a less- protective level is almost always cheaper.

\7\ This problem is exacerbated by the fact that the bill calls for use of central estimates in risk communication principles (SCAA 403, adding CERCLA 131 (d)(2), p. 118]. This is a specific statistical technique that is only appropriate under particular circumstances- ones generally not found in the Superfund context- and are otherwise affirmatively misleading. This language should be stricken.

Similarly, S. 8 continues to lack explicit objectives of protecting clean groundwater, and making contaminated land and groundwater available for beneficial use. These important objectives have, until now, been inherent in the program, given the existing mandate for permanence and preference for treatment. If those provisions are to be narrowed, the list of objectives must grow (with the recognition that not every remedy may be able to attain these additional objectives).

While the revised bill has taken some steps in this direction, it does not go far enough. Beneficial use of land is now included, but only as one element in developing future land use assumptions (SCAA 402, amending CERCLA 121(b)(1)(B)(ii)(IV), p. 97]. Protection of groundwater shows up only in an amorphous way - the bill merely provides that remedial action "shall seek to protect uncontaminated groundwater], and "shall seek to restore ground water to a condition suitable for beneficial use" (SCAA 402, amending CERCLA 121(c)(?)(B) & (C),\8\ p. 100]. It is not clear how these aspirational statements relate to the bill's express objectives and balancing factors. Moreover, even they are "not required to be attained in an area in which any hazardous substance, pollutant, or contaminant is managed in place" [SCAA 402, amending CERCLA 121(b)(1)(B)(v), p. 102] - potentially an immense loophole if interpreted to mean areas other than those directly underlying landfills or other clearly and narrowly delineated areas.

\8\ The numbering of this subparagraph appears to be erroneous.

More generally, S. 8 continues to provide only limited protection for water resources. In particular, protection of groundwater is dependent on its anticipated use -- with all the inherent uncertainties of predicting both who will need the water when, and where the water will be at that time - rather than its status as a valuable and limited resource. Moreover, the bill provides that assumptions about future water use are to take into consideration state water use plans (SCAA 402, amending CERCLA 121(c)(2), p. 98]. Unfortunately, in many cases, these plans were originally developed with no meaningful public input, often many years prior to the cleanup decisions and in a generalized statewide rulemaking or policy making context in which it was not clear to any member of the public or affected community that the decision would have any effect upon a particular site's cleanup. \9\

\9\ Moreover, it is not clear what will happen when a State has designated an area as a low priority for protection under one program--such as a classification of groundwater protection program--but as an underground source of drinking water (USDW) in a rulemaking under the underground injection control program. Clearly, the most protective of the state actions should control In addition, it is important to clarify two additional points: first, that point-of-use devices may only be used on a temporary basis (i.e., while more permanent arrangements are being made), or where no other approach is technically feasible; and second, that technical feasibility" means what can be accomplished from an engineering and technical perspective.

Finally, one other point bears mention with regard to cleanup standards. The bill exempts on-site activities from otherwise-applicable provisions of hazardous waste regulations under the Resource Conservation and Recovery Act (RCRA) [SCAA 402, amending CERCLA 121(a)(1)(C)(i)(II), p. 89]. The bizarre result will be that Superfund sites will be the only locations in the United States where untreated hazardous-waste soils can lawfully be placed in substandard landfills. This provision is an artifact of a problem that EPA has already taken formal steps to alleviate, through the proposal of rules tailoring hazardous waste standards to cleanup situations.\10\ Rather than eviscerating RCRA's applicability to on-site cleanups, the tailored rules should themselves become the applicable standards.

\10\ 61 Fed. Reg. 18780 (Apr. 29, 1996); The proposal is due to be finalized in the next several months.

III. Superfund Slowdown: How S. 8 is the Superfund Cleanup Deceleration Act

More than a dozen provisions of the revised bill impose major new or expanded obligations on EPA. But far from assuring that additional resources will be available so that EPA can accelerate the rate of cleanup completions while meeting these new and largely unnecessary demands, the bill does precisely the opposite: it allows dollars now available for cleanups to be diverted to polluter-pays liability rollbacks, with costs shifted from polluters to the Fund, and with no "firewall" between cleanup costs and these pay- the-polluter funds (see discussion below in section IV).

Even beyond the pernicious effect of the changes in liability on the speed and thoroughness of cleanups, 5.8 as revised has numerous features that will slow down cleanups. These include potentially creating expansive new rights to re-open existing cleanup decisions as well as bottlenecks in the Remedy Review Board process, and requiring EPA to issue a slew of complex new rules implementing changes imposed under the Act - most of which are unnecessary and counterproductive.

The ROD re-opener provisions warrant particular scrutiny. Although the bill provides that a re-opener petition "may" be accepted if certain criteria are met (SCAA 406, adding CERCLA 136(b)(3)(A), p. 141], under existing case law that language could well be construed to require that all petitions meeting those criteria must be accepted.\11\ Such a result would likely lead to an unmanageable explosion in EPA's workload, forcing the agency to divert additional resources from making progress in cleanups to rehashing existing decisions.

\11\ Although we believe that such interpretations are fundamentally inconsistent with the use of the discretionary term "may," the D.C. Circuit recently adopted just such a reading of section 211(f) of the Clean Air Act. There, the statute provided that EPA "may" grant a petition allowing use of a gasoline additive known as MMT upon finding that the additive would not foul automotive pollution-control systems. EPA made such a finding, but rejected the petition on the basis of concerns about the additive's potential health effects. The court ruled that EPA lacked discretion to consider any factors other than the one expressly stated in the statute, e.g., effect on automotive systems, regardless of the fact that the Clean Air Act's objectives expressly include protection of human health. Ethyl Corn. v. Environmental Protection Agency. 51 F.3d 1053,1058-59 (D.C. Cir. 1995).

Similarly, because the bill creates a new role for the Remedy Review Board in assessing reopener petitions [SCAA 406, adding CERCLA 136(b)(3)(A), p. 141], while also requiring Board involvement in reviewing a third of new cleanup decisions [SCAA 404, adding CERCLA 134(e)(2)(B)(ii), p. 131], the Board may well become a major bottleneck. To avoid that result, EPA may have to establish multiple Boards, in which case more and more EPA personnel will have to be involved in Boards instead of actual cleanup activities.

With regard to regulations, the bill requires EPA to issue a slew of new rules, most within six months of the bill's enactment. Even aside from forcing EPA to divert considerable resources to re-writing rules, the pendancy of this array of rulemaking will very likely stall future cleanups, and ongoing ones, while everyone waits to see how the new rules will come out.

Specific rulemaking obligations coming due within 180 days of the bill's enactment include:

-- revise the National Priorities List to delete over-lying parcels from NPL [SCAA 407(b), p. 148];

-- revise the National Contingency Plan within 180 days of enactment to reflect changes made by the Act (SCAA 404, adding CERCLA 133(a), p. 122];

-- issue regulations for providing polluter paybacks (SCAA 502, adding CERCLA 1 12(g)(4), p. 161];

-- issue regulations establishing procedures for the remedy review board [SCAA 404, adding CERCLA 134(e)(2)(A), p. 130]

-- issue regulations for selection of allocators (due within 90 days of enactment) [SCAA 504, adding CERCLA 136(d)(3)(A), p. 172];

-- issue regulations incorporating Results Oriented Cleanup requirements into the National Hazardous Substances Response Plan (SCAA 801(b), p. 244];

-- issue regulations implementing risk assessment and risk communication provisions (due within 18 months of enactment) [SCAA 403, adding CERCLA 131(f), p. 119].

Furthermore, the bill establishes a broad mandatory allocation process that the Administrator must conduct (SCAA 504, adding CERCLA 136(b)(1)(A), p. 165]. Allocations are mandatory even for sites at which consent decrees and settlements have long since been established, if any additional costs will be incurred. In addition, the Administrator (or the Attorney General with EPA staff participation) will need to participate in such allocations in order to assure that the Fund is not drained by unduly enthusiastic attribution of expenses as "orphan" shares that will be paid for by the Fund. Given that multi-party sites with 1 or more viable parties currently lacking a final settlement will use the allocation process - potentially covering several hundred sites - this resource drain is likely to prove substantial.

Last but by no means least, several provisions relating to risk assessment will slow down cleanups unnecessarily and will drain EPA resources. For example, the bill requires use of "chemical-specific and facility-specific data in preference to default assumptions whenever it is practicable to obtain such data" in facility-specific risk assessments (SCAA 403, adding CERCLA 131(b)(1), p. 116]. This language may force EPA to engage in massive data-gathering, to little purpose. Defaults are appropriately chosen for policy purposes, including protection of health where the science is uncertain. Unless someone makes chemical- or facility- specific data available to the Administrator (and the Administrator concludes those data are reliable), generic default values should be used.\12\

\12\ Similarly, the bill imposes an unworkable requirement to identify research needs emerging from each risk assessment, and peer-reviewed studies that are relevant to or fail to support estimates of public health effects and methods used to reconcile inconsistencies in scientific data (SCAA 403, adding CERCLA 131(d)(4), p. 1 18]. while such steps can be useful, they should be done generically, not in each risk assessment. Moreover, the requirement in paragraph (d)(5) to discuss individual studies that fail to support' any risk estimate is at odds with a weight of evidence approach. It would end up focusing as much attention on a single study that contradicts or fails to support the risk estimate or assumption as the multitude of studies which do support that position.

Similarly inappropriate is the requirement to use "the best" science in accordance with "objective" practices (SCAA 403, adding CERCLA 131(e), p. 1 19]. This is excessive, amorphous verbiage that invites endless wrangling to no useful purpose. If any such provision is to be included, it should simply direct the agency to use available, reliable data.

IV. Overly Broad Liability "Reforms": Still Corporate Welfare By Another Name\13\

\13\ Because the liability provisions of S. 8 as revised are substantially similar to those of S. 8 as introduced, this section of our testimony closely parallels the liability section of EDF's testimony of March 5, 1997.

There is no dispute that Superfund's existing liability system has often been abused by some Potentially Responsible Parties (PRPs) who have filed massive contribution actions against entities with minimal or no connection to the site. Curbing these abuses is necessary, but does not necessarily require legislation, since EPA routinely provides contribution protection to settling parties.

Even if legislation on this point were viewed as desirable, S. 8 as revised continues to go far beyond the boundaries of common sense. The bill inappropriately rolls back liability for vast numbers of companies that are well able to help pay for cleaning up their own messes, and who should remain responsible for doing so. In several instances, these overly broad carve-outs apply to future as well as past conduct, undercutting Superfund's vitally important incentives for safely managing today's wastes.

A. The Trade-off Between Liability Rollbacks vs. Cleanup Dollars

Although the bill provides that parties who have already received cleanup orders must carry out the cleanup, it also specifies that they will be repaid for all costs attributable to a party whose liability is limited (SCAA 502, adding CERCLA 112(g)(1) & (2), p. 160]. These paybacks apparently apply even for all future costs incurred under existing settlements. Payback payments "shall be made upon receipt" of an application [112(g)(3), p. 160], and must be made within a year (112(g)(6), p. 161]. In addition, parties that settle pursuant to an allocation have "an entitlement" to be promptly reimbursed for any costs they incur attributed to an orphan share (SCAA 504, adding CERCLA 136(m), p. 186].

This language creates a legal entitlement, as contrasted with discretion under current law to use the Fund for cleanups and other purposes, so paybacks will have first claim on the funds. Because there is no "firewall" between funds for paybacks and funds for cleanups, all of the moneys in the Superfund could be exhausted providing polluter paybacks, leaving none for actual cleanups, oversight, and enforcement by EPA, as well as vitiating programs for Technical Assistance Grants. If funds remaining in the Superfund are inadequate, one of three outcomes will occur: taxes will have to be raised, cleanup standards will have to be further weakened, or cleanups will again slow to a snail's pace. The latter two are unacceptable from an environmental perspective, while the first appears politically implausible.

B. Overly Broad Exemption for "Co-disposal" Sites

S. 8 repeals polluter-pays liability for generators and transporters of industrial wastes at hundreds of "co-disposal" sites at which those wastes were dumped along with municipal trash (SCAA 501(b), adding CERCLA 107((t)(1)(B), p. 153]. Even giant chemical companies will get entirely off the hook for wastes they sent to those sites. And even private dump-owners - those in business to make a profit - get their liability capped at 40% of cleanup costs (or the cost of closure) (501(b), adding CERCLA 107(t)(1)(D)(i), p. 156].

C. Overly Broad Exemption for "Small" Businesses

While we do not necessarily oppose curtailing liability for truly small businesses with a limited connection to a site who have limited ability-to-pay in any event, the current exemption is ill-crafted. First, the $3 million annual-revenue threshold is simply too high [SCAA 501(b), adding CERCLA 107(s), p. 152]. Moreover, the exemption applies to companies with either fewer than 30 employees, or less than $3 million gross revenues. This potentially exempts wealthy corporations that happen to have few employees.

In addition, the exemption applies to conduct in the future, thus eliminating incentives for small businesses to manage hazardous substances carefully in the future: an unjustifiable "pollute with impunity" clause for small businesses. In addition, any liability exemption for small businesses should be conditioned on cooperating with appropriate information-gathering and cleanup activities. Similarly, the exclusion should be inapplicable where the Administrator determines that the material has or may significantly contribute to the response costs at the site (cf. SCAA  501(b), adding CERCLA  107(r)(2), p. 152 (exception to exemption for de micromis contributors)).

V. The NPL Cap: Dumping Cleanups on Communities and States\14\

\14\ Because the NPL cap provisions of 5.8 as revised are identical to those as introduced, this section of our testimony is identical to that in EDF's March 5 testimony (other than with regard to bill citations).

Another highly objectionable feature of the bill is its inclusion of an arbitrary cap on the number of additional sites that can be added to the National Priorities List. Under S. 8, EPA can only add a total of 90 sites to the Superfund National Priorities List before 2000, and then 10 sites/year thereafter (SCAA 802, adding CERCLA 105(i)(1)(A), p. 245]. A cap has profound consequences because, unless a site is listed, EPA cannot undertake cleanup activities (other than a short-term, low-cost emergency removal). In effect, this provision dumps the problem of Superfund site cleanups into the laps of the States -- regardless of whether they have the resources or capacity to conduct those cleanups.

The General Accounting Office recently estimated that the cap could force States to accept responsibility for 1,400 to 2,300 sites (1,100 already identified by EPA, along with an estimated 300 - 1,200 yet-undiscovered sites). The estimated cleanup costs range from $8.4 to $19.9 billion.\15\

\15\ U.S. General Accounting Office, Impact on States of Capping Superfund Sites. GAO/RCED-106R. March 1996.

The GAO report makes painfully clear that the States are in no position to take on this added burden. Indeed, States are having difficulty securing resources for their current cleanup efforts. Of the states surveyed by GAO,

"three of the seven states with active programs said that taking on these additional cleanups would exacerbate an already difficult financial situation. Two other states said that they expect to face funding shortfalls beginning in fiscal year 1997 that will make it difficult to absorb the additional cleanup responsibilities, at least for a few years subsequent to that time. Another two states said that while they had sufficient funds to manage their own inventories, funding the additional cleanups would be difficult."\16\

\16\ Ibid., p. 2.

This provision also undercuts two of the valuable incentives created by Superfund: that which prompts voluntary cleanup of non-NPL sites in order to avoid a potential future NPL listing, and that which prompts careful management of wastes generated now.

An example of Superfund's effectiveness in the former arena emerges from a recent story in the Cleveland Plain Dealer about the Ashtabula River Partnership, a group that is working to avoid a potential Superfund listing by creating "a better-than-Superfund cleanup plan" for the river's heavy-metal and PCB contamination problems. The paper quoted Rep. Steve LaTourette (R-OH) as remarking that "(t]he prospect of a Superfund designation has proven to be a more effective tool than the Superfund itself. Without Superfund, however, most parties wouldn't even be at the table."\17\

\17\ "Toxic Cleanup: Ohioans Aim to Skirt Superfund Listing," Greenwire (electronic newsletter), June 14, 1995 (synopsis of story from June 11 Cleveland Plain Dealer).

Similarly, GAO noted that State program managers "pointed out that a major incentive for private parties to clean up sites is to avoid having their properties added to the list of the most contaminated sites in the country."\18\ In short, a cap on the number of Superfund sites may have the perverse effect of creating a greater need for more Superfund listings, by reducing incentives for non-Superfund voluntary cleanups.

\18\ GAO, p. 3.

The NPL cap will also undercut incentives for sound prospective waste management. Facilities will be able to gamble that states will lack, or forego use of, cleanup enforcement authorities for tackling sites created after the NPL list is effectively closed. The continuing nominal availability of litigation authorities under 107 is far from an adequate substitute, given that 107 suits can only be brought to recoup expenditures - thus requiring cash-strapped States to front all the cleanup money. Where states are unable to do so, today's polluters will evade cleanup responsibilities, and sites will remain unaddressed.

In short, the cap should be eliminated.

VI. Natural Resource Damage Provisions\19\

\19\ This portion of our testimony was prepared by Sarah Chasis of the Natural Resources Defense Council.

While there have been some improvements in the revised version of S. 8 relating to natural resource damages, there have also been some weakening changes. Most importantly, the draft retains the major deficiencies of S. 8 as introduced: it arbitrarily prevents trustees' from factoring heritage values - the values people place on passing on to their children and their grandchildren a pristine wilderness, a population of endangered whales or a national symbol such as the Grand Canyon - into their restoration decisions and from recovering damages for the impairment of these values. This approach has the effect of valuing least our most pristine and endangered resources.

The revised bill accomplishes this result by retaining the prohibition on the recovery for impairment of heritage values, referred to in the draft as "nonuse" damages [SCAA 701, amending CERCLA 107(f)(1), p. 231] and by its prohibition on trustees' recovery for the costs of conducting contingent valuation studies (SCAA 702, amending CERCLA 107(f)(2), p. 234], a methodology that Nobel laureate economists recognize as legitimate and that market researchers and businesses use regularly.

Other deficiencies of the revised bill include its limitations on the trustees' ability to recoup for the interim losses that may be suffered pending restoration of damaged natural resources. The bill has at least four significant limitations on interim losses:

-- It limits such recoveries to "temporary replacement of the services provided by injured . . . resource" (SCAA 701,amending CERCLA 107(f)(1), p. 230; see also, SCAA 703, amending CERCLA 107(f), p. 240]. This language artificially limits recoveries to measures that are temporary and replacement in nature (thus precluding acquisition, for example) and also potentially limits recoveries to prospective losses, those for which temporary replacement costs are incurred, omitting compensation for past losses. The term "services" also could be construed too narrowly to mean just human services, rather than ecological services as well.

-- It precludes recovery of any lost uses that occurred prior to December 11,1980 [SCAA 701, amending CERCLA 107(f)(1), p. 231];

-- It precludes recovery of interim losses, no matter how significant, if the resource has returned to baseline condition before trustees have had a chance either to file a claim or to incur assessment or restoration costs (SCAA 701, amending CERCLA 107(f)(1), p. 232]; and

-- It prohibits recovery of any lost heritage values (SCAA 701, amending CERCLA 107(f)(1), p. 231].

With respect to the selection of restoration options, we strongly support the revised version's deletion of the "reasonable cost" criterion that was in S. 8 as introduced. However, we remain concerned about the criteria that are included (SCAA 703, amending CERCLA 107(f), p. 239]. First, we believe they should be listed as considerations, rather than as absolute criteria, as is the case in Interior's regulations.

Second, arguably the most important criterion is not even mentioned, namely effectiveness in restoring the resource to baseline. This should be included. Cost- effectiveness is included as a requirement, but the term is not defined. To avoid confusion and to clearly distinguish this criterion from a reasonable cost criterion, a definition of the term should be included. Finally, we strongly object to the limitation placed on the last factor, "timely" to the extent consistent with cost and the other three factors. This factor should be included without limitation, just as the other factors are. The current language renders this factor potentially irrelevant. Natural recovery will tend in many instances to be more cost effective than active restoration. If timeliness is not considered as a separate factor but must always be consistent with what is most cost effective, natural recovery will tend to win out, even if it will take decades to occur.

We strongly object to the deletion from the revised bill of the provision contained in S. 8 as introduced for judicial review of the restoration plan on the administrative record. If the trustee goes through the process of compiling an administrative record, which we believe is highly desirable to ensure openness and fairness in decisionmaking, then the evaluation of the decision reached by the trustee should be based on that administrative record. The deletion of this provision from the revised bill defeats the whole purpose of providing for an administrative record with public participation. It means that a PRP could come into court with entirely new evidence that it kept out of the administrative process and use that evidence to discredit the trustee's restoration plan. There will be no incentive for the trustee to compile an administrative record since the PRP will be free to ignore the process and submit its evidence later in court. This change represents a serious step back from S. 8 as introduced.

The revised bill's language on the relationship between response actions and restoration (SCAA 703, amending CERCLA 107(f), pp. 239-240] is an improvement over S. 8 as introduced, as is the provision for a Federal "administrative" trustee (as opposed to a lead Federal "decision-making" trustee).

The provision [SCAA 705, adding to CERCLA 1 13(g)(1), p. 241] allowing for an extension of the current statute of limitations only where, in effect, the PRP agrees (by entering into a cooperative agreement) is ineffective in addressing trustees' concerns on this issue. We strongly recommend that the provision on mediation (SCAA 706, adding to CERCLA 136, p. 242] be made optional. To require trustees to go through a mediation process when there is no prospect of cooperation from the PRP only introduces delay and expense into the process. Mediation works only when there is a real interest on both sides; otherwise, it is a waste of time and money and further delays restoration of the resources.

Three final points. We oppose the provision on double recovery which would broaden current law in a number of ways (e.g., by extending to actions brought under state law, as well as federal law, and potentially limiting recoveries in such state actions for damages other than restoration costs, as well as extending to response actions -which are not designed to achieve restoration) [SCAA 701, amending CERCLA 107(f)(I), p. 231].

We also are concerned with the failure to call for the development, as part of the regulations, of simplified damage assessment methods. This combined with the call for "facility-specific" information (SCAA 702, amending CERCLA 107(f)(2), p. 233] could be used to call into question the ability of trustees to utilize simplified assessment techniques, which not only save time and money, but ensure that smaller spills and sites are assessed and restored.

Finally, we have serious objections to the grandfathering provision (SCAA 707, p. 242] that seeks to carve out a special exception for the Clark Fork case in Montana.

Conclusion.

Thank you for this opportunity to present our views.