The Heritage Foundation

Issue Bulletin #204

October 3, 1994

October 3, 1994 | Issue Bulletin on

Superfund Status Quo: Why the Reauthorization Bills Won't Fix Superfund's Fatal Flaws

(Archived document, may contain errors)


October 3, 1994


INTRODUCTION Congress may act soon on legislation to reauthorize the 1980 Comprehensive, Environ- mental Response, Compensation, and Liability Act (CERCLA, pronounced sir-k1uh), also known as Superfund. As it is enforced, Superhind is perhaps the most complex existing law. But it is now synonymous with failure and has become a symbol of wasteful, ineffec- tive environmental legislation. The Clinton Administration agrees that the Superfund law has not worked, and last Feb- ruary unveiled its reform proposals. Thew proposals have been substantially incorporated into two bills before Congress: S. 1834, introduced by Senator Max Baucus (13-MI), who is Chairman of the Committee on Environment and Public Works; and H.R. 3800, intro- duced by Representative Al Swift (D-WA), the Chairman of the Transportation and Haz- ardous Materials Subcommittee of the Committee on Energy and Commerce. Although $20 billion has been spent already on the program, Superfund has failed in its mission to clean up hazardous waste sites. So far, only a small fraction of the "worsf'haz- ardous waste sites have been cleaned up. To clean up the remaining sites will cost hun- dreds of billions more. Unfortunately, the legislation fails to address the two fundamental problem that have created this situation. Specifically, the legislation: Retains the law's retroactive. strict, joint and several liability scheme. This delays site cleanup as responsible parties attempt to avoid liability by shifting it to others. More- over, it adds to total site cleanup costs by creating a climate for increased litigation and other transaction costs. Maintam overly stringent and inflexible cleanup standards. Unreasonable cleanup standards are the other major factor behind the slow and costly pace of site cleanup. The current system not only fails to prioritize those sites posing the most significant risk, it also imposes excessive standards of "cleanliness." instead of fixing these fundamental problems. the bills mate the Environniental Insur- ance Resolution Fund (EIRIF), finided by an unnecessary and inequitable $8.1 billion tax on the insurance and reinsurance industry. The purpose of this fund is to settle the problem

created by current law that virtually forces companies to sue the insurance industry be- cause the extent of insurer liability is not easily settled. This fund, built on retrospective and prospective taxes, helps some insurance companies at the expense of others and still would leave insurers open to future risk. The right solution would be to repeal retroactive, joint and several liability and fix the cleanup standards. By doing this, only the most needy sites would be cleaned up. Instead of the litigious scramble for the responsible party and the ensuing litigation, the polluter would pay. For those sites where no one is legally responsible under the new law, the fed- eral government should decide if they warrant cleanup, and if so, clean them up. In addition, the bills contain other provisions that do nothing to fix the Superfund pro- gram and will only add to the cost of the bill and the cost of cleaning up hazardous sites, such as new Davis-Bacon Act expansions in the House bill. These have no place in such legislation. The many problems with these bills underscore the need to move cautiously on Super- fund reform rather than to rush through a bill in the final days of this Congress. With ex- ploding costs on the horizon caused by Superfund's flaws, real reforms are necessary now. These bills, however, would virtually guarantee five years more of the status quo until the next reauthorization. If Superfund reform cannot be done right this year, it should not be done.


The Superfund program was created in 1980 after widespread media attention to the ex- posure to high levels of dangerous chemicals of residents of Love Canal, near Niagara Falls, New York. The national media coverage thus provided the political impetus for legis- lative action.

Superfund, known formally as the Comprehensive Environmental Response, Compensa- tion, and Liability Act (CERCLA), was intended to force companies to clean up their haz- ardous wastes. It also created a government fund to pay for cleanup of "orphan sites " '@' The which are contaminated properties where the polluter, say, has gone out of business. law was hurried through in the final days of the Carter Administration and many provi-

1 It is important to note that the company that buried the chemicals in the area, Hooker Electrochemical Company, took great cam to prevent public exposure to the chemicals. Hooker buried the chemicals sealed in a thick protective tube of impermeable clay with the intent of maintaining ownership of the land. Unfbrtunately, the Niagara Falk Board of Education through threat of eminent domain, effectively forced the company to sell the land (fbr $ 1), despite the company's warmings that the land was contaminated and not fit for public use. A public elementary school was built on the site and city construction crews bulldozed through the clay seal, thus allowing the chemicals into the surrounding soil. Additionally, some of the land was sold off for residential housing. Tbus, the problems at Love Canal were the direct result of government interference. 2 42 U.S.C. sec. 9601 et. seq.; PL 96-5 10, as amended by 97-216, July 18, 1982; Pl. 97-272, September 30, 1982; P.L. 98-45, July 12, 1983; PL 99-160, November 25, 1985; PL 99-499 (Superfund Amendments and Reauthorization Act of 1986 (SARA)), October 17,1986; PL IOD-202, December 22,1987; PL 101-144, November 9,1989; PL 101-508, November 5,1990; PL 101-584, (Superfiind Surety Bonding), November 15, 1990; PL 102-389, October 6, 1992; and PL 102-426, October 19, 1992. sions were ill-considered. Consequently, CERCLA has cost approximately $20 billion or more while utterly failing in its mission of cleaning up hazardous waste sites. As Carol Browner, Administrator of the Environmental Protection Agency (EPA), explains, "A lot of time and money is taken y with companies suing each other over how much they owe to clean up a particular site." To deal with orphan sites, Congress originally authorized a $1.6 billion trust fund, or "Superfund." This sum was believed adequate by policy makers to clean up all orphan sites known or likely to be dis- covered. For many reasons, Federal Superiund Outlays Still Rising however, the program proved much more expensive. In 1986, V= Mftm of NowW Dohn the Superfund Amendments 1AW Too] FedwW SWdng dm aq - - - - - - - - - - and Reauthorization Act 1994:$IOAWW= (SARA) increased the trust fund by another $8.5 billion, to ------------------- $ 10.1 billion. The total was ----------------- raised by another $5.1 billion, ------------- to $15.2 billion, during the 1990 budget deal. The increases ------------- were financed through excise 40 - - - - - - - taxes, primarily on crude oil and petrochemicals, as well as a corporate environmental tax. Of 1"1 110 Ila 1W My 1"1 1"] 1"40 this amount, $10.9 billion had Amwd 0 br Find V@ No 01"4dmiseW -' been spent by the end of fiscal year 1994.4 The $10.8 billion spent so far from the Superfund, moreover, comprises only a fraction of the total cost of CERCIA. The cost spent by companies and individuals in compliance with CERCLA has raised the overall price tag of the program by billions of dollars more.5 Much of this cost has been spent not on cleanup, but on litigation and other "transac- tion" costs. In fact, only 237 of the 1,292 "worsf' hazardous waste sites placed on the Na- tional Priorities List (NPL), or about 18 percent, have been cleaned up. 6Moreover, only 60 sites have been deleted from the NPL. To clean up the remaining sites already on the NPL, the EPA's costs alone-excluding costs to businesses and insurance companies- will run an estimated $40 billion. Also, many more sites are expected to be listed on the

3 Remarks by EPA Administrator Carol Browner to the U.S. Chamber of Commerce Policy Insiders Breakfast, February 16,1994. 4 7be 1994 figure includes estimated outlays for Fiscal Year 1994, which funds through September 1994. The cost in real dollars is $12.2 billion. 5 Estimates vary greatly-frorn $7.9 billion to $18 billion-and typically we not based on solid data; nevertheless, there is uniform agreement that the costs have been substantial. 6 EPA Superfund Hotline run by Booz-Allen & Hamilton, as of September 23, 1994. Note that some of those cleaned up sites, technically known as "construction completed," actually we undergoing long-term remediation. Also note that by the end of 1992, from which time many of the cost figures for various studies were derived, only 149 of 1,275 on the NPL had been cleaned up. NPL-for an ultimate number between EPNs Superfund: Lots of Sites, Few Cleaned Up 2, 100 to 10,000- lA00 thus driving up the costs even higher. 7 IAW ----- ------------------- : ------ Over the next thirty 1=1 ---------------------- years, according to lAW ---------------------- the General Account- ing Office, total costs are estimated at $300 6W ---------------------- billion in 1990 dol- 4W ---------------------- lars.8 The Congres- 2W 237 ------------ sional Budget Office, 60 on the other hand, has P kdo IN im iq Lft Consumcdon Cmnoated Was& put the total costs of &Verfisnd Ska Cle-W CarnoM NW Ckanw Congk". cleaning contami- SmAcr. EPA &VURM1400 Yet CeMed as Clean Cuffed as Clean al nated sites not owned by the federal govern- ment at between $106 Cleaning Supeftnd Sites: Massive Spending billion and $463 bil- Increases Coming Soon lion in nominal dol- Billiws arCurent Dollan lars, or $42 billion $16 and $120 billion in 14 - - - Cale - - - - - - - - - - - Cale discounted dollars, 12 --- --- Case ----------- over the next 78 years. 9 The lion's 10 -- ---- --------------- share of those costs 8 - - ........ --------------- will be incurred 6 . --- -------- --------------- within the next two 4 - - - - -- - - - - - - - - - - - - decades, with annual 2 -------- ---- - -- ----------- costs potentially ...... .. . . .... reaching a peak of 2OW 2010 2M 2030 : 2W 2= ... ... .... 2M 2M $14.4 billion by the year 2003. 10 if the

7 Congressional Budget Office, "TheTotal Costs of Cleaning Up Nonfederal Superfund Sites," January 1994, citing Milton Russell, E. William Colglazier, and Mary R. English, Hazanious Waste Remediadon: The Task Ahead (Knoxville, Tenn.: University of Tennessee, Waste Management Research and Education Institute, 199 1); and the Office of Technology AssessmenL Coming Clean. Superfund Problems Can Be Solved... (October 1989). 8 General Accounting Office, "Superfund Program Managemem" Imligh Risk Series, December 1992. 9 CBO, "Ibe Total Costs of Cleaning Up Nonfederal Superfund Sites," January 1994. Tbc baseline projection in nominal dollars was estimated at $228.3 billion, or $73.9 billion in discounted dollars. To calculate the discounted costs, the study used a discount rate of 7 percent, which biased the study's discounted costs downward. While proper if the figure being calculated was return on investment (ROI)--or payback on a present cost--to ensure a conservative estimate, it should not be used to estimate future costs. 10 ibid. Note that this figure is the high case estimate in undiscounted dollars for cleanup of non-federal facilities. costs of cleaning up federal facilities are added into the equation, the total cost could ex- ceed $750 billion or about $7,800 per household in America.


The high cost and slow pace of Superfund cleanups are caused principally by two prob- lems: the program's unworkable liability scheme and its overly stringent cleanup stand- ards. The law's retroactive, strict, joint and several liability scheme results in endless litiga- tion as potentially liable parties search for the "deepest pockets" to pay for cleanup. In ad- dition, parties can be held liable for cleanup whether or not they are at fault, even if they dumped only minute amounts of waste and regardless of whether the dumping was legal at the time. The program's unreasonable cleanup standards require hazardous waste sites, no matter what the location, be excessively clean and fail to set up a priority system that en- sures the maximum bang for the cleanup buck. The Liability Scheme Is Unworkable The most important reason why the Superfund has failed to clean up more hazardous waste sites is its highly unusual liability scheme. This retroactive, strict, joint and several li- ability system delays cleanups and increases total costs by fueling constant litigation. This liability scheme has led to countless examples of absurd and grossly unfair results. For instance, Russ Zimmer was named as one of the "potentially responsible parties" (PRP) 12 for a Superfund site located at a battery cracking plant in Torrington, Wyoming. in? He sold a bag of dog food in 1977 and a bag of seed in 1984, and took a third-party check as payment for the items. Since the checks originally had been issued by the now-bankrupt company that had owned the battery cracking plant, Zimmer was sued as a PRP. Zimmer decided upon advice from counsel that he should set- tle or it would cost him even more in legal costs. So he agreed to pay $3,500. 13 Joint and several liability is much to blame for causing legal woes such as Zimmer's. Un- der joint and several liability, a person who contributed even a very small amount to a problem, such as by tossing a car battery into a landfill, can be held responsible for the en- tire cost of, say, a multi-million dollar cleanup. 14

The base caw and low caw estimates are $9.1 billion and $4.4 billion, respectively. Inclusion of federal facilities in cost estimates, of course, would result in significantly higher estimates. 11 Kent Jeffreys, "Science, Economics, and Environmental Policy: A Critical Examination," Alexis de Tocqueville Institution, Arlington. Virginia, August 11, 1994, citing Milton Russell at al., Hazardous Waste Rernedlation. The author calculated the potential average household cost by dividing the total by the number of households in the United States, or 96.391 million, in 1992, according to U.S. Department of Commerce, Bureau of the Census, Current Population Reports, "Money Income of Households, Families, and Persons in the United States: 1992." 12 "Potentially Responsible Party," usually referred to by its acronym PRP, is the legal term referring to a person or institution that is identified as a liable party under CERCLA's liability scheme. 71bis term also is applicable before any court determination of liability, as the word "potentially" implies. 13 Jack Anderson, Michael Binstein, "How Superfund Hurt the Innocent," The Washington Post, March 7, 1994, P. CIL 14 Special rules apply now to very small (de minimis) contributors, but the problem discussed nevertheless still The concept is designed to find somebody-anybody-to pay for cleanups of hazardous waste. Under the doctrine, the "responsible" person presented with the tab often has no part or only a small part in creating the waste problem. In practice, this means that a firm with "deep pockete' ends up paying for the cleanup of wastes created by someone else. Naturally, any person or institution identified as a PRP has a strong incentive to prove that others are liable as well, in order to dilute the costs among more PRPs. Hence, litigation is virtually assured. Although EPA Administrator Carol Browner has admitted the need for reform, she justi- fies Superfund's liability scheme because "a lot of good has come from the program. The polluter-pays concept that was first adopted in the Superfund law has changed the way businesses in this country deal with their waste." Unfortunately, Superfund does not adopt the "polluter-pays" system, but rejects it. By imposing joint and several liability, Super- fund guarantees people like Zimmer pay for cleaning up the pollution of others. In short, joint and several liability works against the idea of making polluters pay the costs that their actions impose on others-no more, no less. 15 Petlpps the most controversial part of this liability scheme is the retroactive compo- nent.16 Retroactive liability mews that owners, operators, and other individuals and com- panies who otherwise would be liable under the strict, joint mid several standards are held responsible for pollution that took place before the Superfund law was passed in 1980. For instance, if a company made jeep parts during World War H, and during production some chemicals seeped into the ground, the company would be legally responsible for the cost of cleaning up those chemicals now. Thus, because of a law passed more than 35 years later, the company may be faced with tens of millions of dollars in cleanup costs-even though the co4ipany's actions not only were completely legal at the time, but were consid- . . IT cred patriotic. There am essentially three problems with retroactive liability. Retroactive liability is of questionable constitutionality. The standard is seemingly unconstitutional on its face. In the same pan of the Constitution that authorizes the exist- ence of Congress, it clearly and unequivocally states that "No ... ex post facto Law shall be passed." 18 An ex post facto law is defined as "the infliction of punishment upon a person for an act done which, when it was committed, was innocent... [or) every law which, in re- lation to the offense or its consequences, alters the situation of a person to his disadvan- tage."19

exists for many extremely small. as well as moderate, contributors. 7be bills before Congress offer a partial solution to the current system by exempting truly tiny (de micromis) contributors, but the bills would not clumnate the problem because they would have the burden of proving a negative--that they did not pollute more-to qualify for the exemption. Moreover, moderate polluters would remain subject to enormous liability. 15 For a clear elucidation of the principle as properly defined, see John Shanahan, "Ibe Conservative As Environmentalist," Heritage Lecture No. 358, November, 19, 1991. 16 Retroactivity is not expressly provided for in the text of CERCLA, but that it is clear from the usage of the past tense that Congress intended for it to be retroactive. Moreover, the courts have correctly interpreted that the intent of the law was to impose retroactive liability. 17 For many defense contractors, liability has amounted to hundreds of millions of dollars. 18 U.S. Constitution, Article 1. Section 9, Clause 3. Superfund very clearly is an ex post facto law because individuals and companies are surely punished and their situations "altered to disadvantage" when the government re- quires them to spend perhaps tens of millions of dollars to atone for actions that were legal when done. The intent of the Founding Fathers is clear, yet that intent has not been rigor- ously upheld by the Supreme Court. Relatively early in the history of the nation, in fact, the Court held that theprohibition against ex post facto laws applied only to either crimi- nal or penal measures.@lu Thus, an important question is whether the imposition on a single person of liability for millions of dollars is penal by nature. When individuals and compa- nies may be bankrupted based on their own past behavior-as opposed to, say, compliance with general regulations-it is difficult to conceive of a law that could be more penal with- out being criminal. Even in cases where a statute was not considered penal, the Court traditionally ruled against retroactive laws. 21 In the last twenty years, however, there has been a shift in the interpretation. A relatively modem case that squarely addresses ex post facto "non-penal" laws suggests that, if ruled on, Superfund could be held constitutional if it is found to be ra- tionally related to a legitimate government purpose. 22 This is considered a low legal hurdle to overcome. In addition, lower court decisions have found that the prohibition against ex post facto laws poses no constitutional barrier to Superfund's retroactivity. 23 To date, how- ever, the Supreme Court has not squarely addressed the issue of Superfund retroactivity. Retroactive liability is unfair. Fairness is another problem with retroactive liability. It is unfair to hold people accountable today for actions taken decades ago, perhaps by a pre- vious owner of the land with several intervening owners. Yet the law, as enforced, does just that. Thus, a farmer may find that his grandfather-or indeed, someone whom he has never heard of-fouled the sub-soil 50 years earlier using state-of-the-art equipment, and the farmer must fix the entire problem. no matter how much it costs. Even if the farmer caused the problem himself back in the 1960s, is it really fair to assume that a law-abiding citizen should know that his actions-reasonable at the time-could destroy his children's inheritance of the family farm? Not only envuonniental sensitivity, but environmental knowledge has increased exponentially since that tune. Retroactive liability has triggered an avalanche of litigation. The third problem with this liability scheme is a very practical one: it leads to constant litigation. Litigation prob- ably is the biggest contributor to Superfund's failure to achieve its primary objective of cleaning up hazardous waste sites.

19 Henry Campbell Black, M.A., Black's Law Dictionary, Sixth Edition (St. Paul, Minn.: West Publishing Co., 1990). 20 Calder v Bull, 3 U.S. (3 Dall.) 386, 1 L.Ed. 648 (1798). 21 See Railroad Retirement Board v. Alton Railroad Co., 295 U.S. 330,55 S.Ct. 758, 79 L.Ed. 1468 (1935).7be Court bias against retroactive laws has not solely rested on the ex post facto law prohibition, but also has rested on several other constitutional grounds such as the due process. 22 Usery v. Turner Elkhorn Mining Co., 428 U.S. 1, 96 S.Ct. 2882,49 L.Ed.2d 752 (1976). 23 See U.S. v. Northeastern Pharmaceutical & Chemical Co., Inc., 810 F.2d 726 (8th Cir. 1986), cert. den., October 5, 1987. The fact that the Supreme Court declined to hear this ca lends some weight to the belief that the Court supports the 8th Circuit court's decision, but does not set Supreme Court precedent.


Perhaps the most important practical problem is that retroactivity increases the raw num- ber of PRPs for any given site. Numerous individuals and companies may have owned the land over the decades. Moreover, each of the previous owners typically had dealings with others who might be liable. When the fact that large numbers of owners and other poten- tially liable parties often have poor records, it is little wonder that those already targeted by the EPA as deep pockets turn to the courts to defray the costs. Another problem is that records for actions taken decades ago can be difficult to obtain. If an entrepreneur conducted regular business in the 1950s through the 1970s with a chemi- cal company that created a hazardous waste site, he usually would not have kept records dating back that far. He would be W-suited to defend himself in court against canceled checks and other such records from the chemical company. His own culpability may be small or nonexistent, but without records of his own, the evidence may point to his liabil- ity. Moreover, since the law also is strict, joint and several, the entrepreneur is potentially liable for the entire cleanup if he has the deepest pockets. Thus, liability potentially can oc- cur due to a failure to maintain decades-old records.

Another practical problem of retroactivity Average Cost per Superfund Ske: is that companies recog- C@P=W5vA*A=wW3-m @kasthm$100MIm nize the government could very well decide Lepi to increase standards 39.0% beyond those required today. Aftr all, retroac- Nan4& F ffiqpr6dw wfth Fbtat ly Reqmwbk tivity has proved no Pwdm EPA and Insuram bander with this law. conpWan Odwr 19A% 9A% Thus, companies am vulnerable to uncertain Gmrmwm and unknowable liabil- poffmm ity costs in the future. OWUP 7.6% Naturally, this further 152% 1 5j=UWdS.DbV%OabomhS.D_w uWWmKAW... 1%tt&SWwCWww-WE I aw -s in- i. motivates companie Go= at 13 kp&UW SW RAND Gorpwadom 1"3. volved in litigation to reduce their share of the liability by suing as many other companies as possible.24 In all, transaction cost accounts for approximately 32 percent of total private sector CER- CLA costs, although the average varies from 17 percent to 60 percent according to firm size. 25 Litigation comprises 65 percent of this amount, or 21 percent of the total average site cost. While this figure is significantly lower than the widely publicized, yet inaccurate, estimate of 80 percent of site cost, it nevertheless represents a figure of about $6.7 million per site.

24 This list of problems is not exhaustive, but is representative only. 25 Jan Paul Acton. Lloyd S. Dixon, with Deborah Drezner, Laural Hill, and Steven McKenney, Superfiod and Transaction Costs. The Experience of Insurers and Very lArge Industrial Firms (Santa Monica, CA: Rand institute for Civil Justice, 1992). Cleanup Standards Are Overly Stringent, Inflexible The extremely high costs of actually cleaning up Superfund sites also contribute to delay of cleanups and litigation. Although EPA estimates that hazardous waste sites are a me- dium to low health risk, the. average site cleanup is now estimated at $24 million to $46 million.26 Yet this figure is unnecessarily high because of overil stringent and inflexible cleanup standards. This leads to numerous and costly problems. 7 The most egregious problem was highlighted by the Office of Technology Assessment when it found that cleanup decisions at the majority of Superfund sites are based on hypo- thetical future risks rather than on actual or likely future risks.28 This means that remote or inaccessible contaminated sites that pose no current or likely future danger to people's health are given a priority just as high as sites that expose people to a real present danger. In fact, the Agency for Toxic Substances and Disease Registry announced in 1990 that un- der 12 percent of all Superfund sites pose an actual or current risk to human health or the environment.29

To place sites on the National Priorities List, EPA uses a system known as the "hazard- ous ranking systenf' (HRS) to score sites on a 100 point scale. If a site ranks 28.5 or above. it qualifies for the NPL. Yet this system has been widely criticized. As Peter Guer- rero, Associate Director of Environmental Protection Issues at the General Accounting Of- fice (GAO) stated, the FIRS "utilizes an arbitrary criterion... not based on the risks posed by the sites."30 Another problem is CERCLA's preference for permanent cleanup, as opposed to on-go- ing treatment or containment. This preference drives up the costs of remediation because containment and isolation are discouraged. Even if blanketing layers of plastic and clay on the soil, surrounding the property with a fence, and posting guards would completely elimi- nate exposure to the public for a small fraction of the cost of remediation, this option is not available. Moreover, Superfund is structured so that states pay only 10 percent of remedial construction costs, while it requires them to pay 100 percent of the Operation and Mainte- nance (O&M) costs of Superfund sites in the state. Although ongoing treatment may cost much less over the long run and may reduce public exposure equally well-and in some situations, more-states have a built-in incentive to demand permanent and extremely

26 Ibid. The $32 million figure is based on 108 companies at 18 sites with annual revenues less than $20 billion between 1981-1991. 27 The problems discussed are not exhaustive. 28 "Coming Clean: Superfund Problems Can Be Solved...," Office of Technology Assessment, October 1989. Others also have identified this problem. 29 Kathryn A. Kelly, Dr. P.H., Nanette C. Cardon, M.S., M.U.S., "The Myth of 10(-6) as a Definition of Acceptable Risk," EPA Watch. Vol. 3, No. 17 (September 15, 1994), p. 8, citing reference in C.B. Doty and C.C. Travis, "Is EPA's National Priorities List Correct?" Environmental Science and Technology Vol. 24, No. 12, pp. 1778-1782. 30 Richard L. Stroup, Sandra L. Goodman, "Rights vs. Regulation: How to Reform Superfund" (Political Economy Research Center, Bozeman. MT, September 1994), endnote #17. p.37, citing Peter F. Guerrero, testifying before the Subcommittee on Oversight of the Committee on Ways and Means, June 11, 1992. Serial 102-122, p.37. strict cleanup standards in an effort to avoid fu- Federal Superhmd Obligations: FY 199Z ture O&M costs. Thus, Direct Response: Includes hasiMlity states reduce their costs, SoAe and Ranedineon but companies, through 51.8% increased taxation, need- lessly pay more. ..... ..... Another problem is Mansgernentand that EPA uses a range of Enm gamy Response Admhdrtradon wedih EPA 'tisl&'-froni one in ten Support 8.8% thousand to one in a mil- 17.7% lion-to &fine when the EPA and Nadonal I of Ei A = w HesM probability of contracting ra IMCM.M Riesearch and Dewdopment cancer crosses the line 15.1% 6.6% from being an acceptable s..=-ftTwcara..%upiw&w &*.w9waio.*.yiqm. risk to an unacceptable 11 M@ risk. Thus, one official may consider a site to be "hazardous," and thus subject to CER- CLA, if the risk of cancer from the materials at the site is one in ten thousand over a life- time. But another EPA official may use a risk level of one in a million. There is much dis- agreement over whether using a range of risks is proper or whether EPA should set a sin- gle specific risk level, or risk-point, that is acceptable. The more important controversy, however, centers not on the range argument, but on whether using a one in a million risk 31 of contracting cancer over a lifetime of exposure is appropriate at all. Conceptually, it helps to put this risk level in context of overall cancer rates. The chance of developing cancer from all causes is I in 3, or 33 percent. Thus, if a site increases the cancer risk to an exposed person by one in a million, it increases the person's overall can- cer risk by only .0003 percent. This is an extremely small increase in the overall chances of someone contracting cancer. Realistically, moreover, it is unlikely that many people would be exposed over an entire lifetime. Thus, the real cancer risk increase is smaller yet. Oddly, there is no scientific justification for choosing a one in a million risk level. It has "never received widespread debate or even thorough regulatory or scientific review. It is an arbitrary level proposed 30 years ago for completely different ref@ations .... the circum- 3 , stances of which do not apply to hazardous waste site regulation." Moreover, no one seems to know how this risk level came to be the standard of a one in a minion risk. In fact, when representatives of Environmental Technologies International, Inc., asked one federal agency where the standard came from, they were told, "You really shouldn't be ask- ing these questions.,,33

31 Any specific numerical risk level should be at the lower end of the current range, and should act as a threshold floor for prioritizing cleanups. 32 Kelly and Cardon, "Ibe Myth of 10(-6) as a Definition of Acceptable Risk," p. 4. 33 Ibid. PROBLEMS WITH THE REFORM BILLS Given that the Superfund program is facing hundreds of billions in cleanup costs in the next few years it is critical that the five-year reauthorization bill address these two funda- mental problems. Unfortunately, both the House and Senate bills make only half-hearted attempts to address them. Instead of repealing the law's liability scheme, it sets up instead an unwise and inequitable fund that will settle only part of Superfund's litigation problems. With regard to cleanup standards, the bills fail to adopt acceptable risk standards for cleanup and site prioritization. Finally, they include unnecessary giveaways for organized labor and environmental activists that will only increase the cost of the Superfund pro- gram. Failing to Establish Real Reform in CERCLA's Liability Perhaps the most controversial liability reform in the bills is the proposed Environmental Insurance Resolution Fund (EIRF). The EIRF ostensibly is designed to settle the problem created under current law that virtually forces companies to sue the insurance industry be- cause the extent of insurer liability is not easily settled. Not only does this drive up litiga- tion costs, but it hurts the insurance industry's ability to make sound business decisions. In- surance is based on spreading risk, which in turn is based on actuarial calculations-in short, the insurance industry is built on the predictability of risk. Yet Superfund defeats this purpose by inserting huge uncertainties into the process. This has hurt both hisurance companies offering commercial multiperil coverage and those insured. Some insurance companies face bankruptcy, and those remaining are unable to gauge their liability exposure for policies written decades ago. Hence, CERCLA creates uncertainties in premium rate setting today because insurers must somehow recoup these losses through cost shifting- 34 The EIRF has been proposed as a way to enable insurers to "insure againsf' this uncer- tainty by creating a fund financed by an $8.1 billion tax over ten years on the entire insur- ance industry offering commercial multiperil coverage. The tax, which would be levied on both direct insurers and reinsurers, would fund voluntary settlements by PRPs. Essentially, PRPs found liable for cleanup could agree to accept ursement from the fund in ex- change for agreeing to settle any claim they had instituted against the insurers prior to this . . 35 reauthorization. This often would be advantageous to insured companies since their li- ability would be certain and litigation costs would be reduced. The choice to accept the set- tlement would be voluntary for the insured company, so even though an insurer pays the tax, it still could be liable for all existing claim depending on the PRPs' decision-mak- ing.36 Moreover, since future claims would not be covered by the fund, insurance compa- nies would continue to be exposed to uncertain risk.

34 Moreover, the inevitable cost shifting hurts modem companies with newer policies because the insurance industry's liability costs for old policies are shifted to them indirectly. 35 1he bills include specific provisions as to eligibility. timing, and necessary documentation. The term "insurers" refers to both direct insurers and reinsurers. 36 The reimbursement offer made to the PRPs would be based on a fixed percentage of the cleanup costs. These fixed reimbursement rates, broken into categories, vary according to the state. 7he amounts were based roughly on the reMve average awards by courts against insurers in each state. 7bus, each state is assigned to

If fewer than 85 percent of the PRPs potentially eligible for awards did not accept the fund's settlement offer within 150 days of the bill's enactment, however, the fund would not be created. 37 If this occurs, and there really is no way to assess adequately the prob- abflity of reaching the 85 percent threshold, one of the cornerstones of the "reform' 'would disintegrate. 38 Even if the fund became operative, and thus brought some predictability to the situation, the EIRF is ill-conceived. Under H.R. 3800, 70 percent of the tax raised in the first four years is to be fmanced through retrospective taxes on net premiums charged between 1968 and 1985. 39 But retrospective taxes simply are unfair. While they may not be penal in na- ture, they nevertheless exact taxes based on financial activity between 26 and 9 years into the past. While many insurance companies may not be adversely affected by the tax, since they will recoup the loss through reduced future liability, others will be net losers. Many of these companies might have been more prudent when writing 4rlicies, or for other reasons may have smaller liability risks in relation to their premium . Yet they would be taxed just as much. A prospective tax is even more unfair, and some 30 percent of the proposed tax for the first four years-and 75 percent for years five through ten-is prospective. The entire pur- pose of the EIRF is to settle liability uncertainties from old policies, yet a prospective tax is based on future policies. This means insurance companies offering multiperil insurance coverage that cut back market share or were less cautious in decades past reap a windfall at the expense of the cautious or expanding companies. For example, Lloyd's of London has cut back its market share considerably, but still faces enormous liability for past policies.

one of three percentage categories for reimbursement settlenients: 20 percent , 40 percent , or 60 percent. Most states am classified in the 40 percent reimbursement category, with a few states in each of the 20 percent and 40 percent reimbursement categories. The categories, however, we unnecessarily limited and broad. Also, it is claimed that at least two stun have been miscategorized. 37 Note that there would not be a five month accumulation of money in the EMF because, although taxes am being accrued through the first 150 days, payments we due only after die fund becomes operative, if at all. This period could be up to 150 days (known as die contingency period). This is a sound approach. 38 The reauthorization attempts to increase the likelihood of achieving at least 85 percent by penalizing companies that decide to take advantage of their right tD settle the matter in court, and either lose or we awarded less than die settlement offer. In such cases, die insurance companies' attorneys' fees must be paid by the company. While an effective strategy, it is unfair and only is necessary because the reform is so timid in its overall approach to real reform. 39 As this paper goes to press, the taxing structure in S. 1834 is indeterminate. This will change, however, during mark-up of the bill by die Senate lFinance Committee scheduled for September 28. Most notably, the Administration has proposed that the tax on reinsurers be different from that on direct insurers by making the tax on reinsurers 100 percent retrospective. This would address some of the problem discussed regarding H.R. 39M. but would not eliminate the problems from the Senate bill. Moreover, the financing mechanism that would emerge fi-om a conference of both chambers cannot be accurately predicted. Note that die 70 percent figure in die text accompanying this footnote applies to years one to four only. During years five to ten, retroactive taxes will account for approximately 25 percent of the annual total. 40 Them can be many reasons for this situation. Insurance companies could have collected higher insurance premiums per policy for their level of Superfund liability exposure because their policies subjected them to what then was perceived to be greater risks. Conversely, they could have insured companies, that on average, pose lower Superfund liability risk to the insurance company. Under a prospective -tax, the firm's total exposure is reduced considerably at the expense of American insurers, who would pay disproportionately. VAfile the Administration proposes to amend this structure in the Senate bill to eliminate prospective taxes on reinsurers, the problems of the EIRF's inequitable collection of taxes cannot be completely cured. Despite the Clinton Administration's assurances that the EIRF is a "consensus" position, only a small percentage of insurance companies support this tax. The over 2,000 compa- mes opposing the EIRF typically are much smaller than the companies supporting the EIRF. But oven so, those opponents represent approximately two-thirds of total premiums written. This hardly represents consensus in favor of the new tax. As Jack Ramirez, Execu- tive Vice President for the National Association of Independent Insurers, stated, "We are strongly opposed to the insurance fund. The product they have come up with is not the in- dustry's position, although it is being represented as such."41 Rather than impose an inequi- table $8.1 billion tax that has limited support at best, Congress would be better served by dealing with one of the fundamental problems of Superfund: a flawed liability standard. Another important provision also is ill-advised. The bills attempt to reduce legal costs by allowing companies to use an arbitration process in which the costs would be allocated among the liable PRPs. On its surface, this proposal sounds like a step in the right direc- tion because it would alleviate much of the litigious burden of joint and several liability. However, the proposal has a very serious built-in problem. Many companies may be un- willing to use this system because it is an informal process that denies them the protection of the normal rules of evidence and procedure in courL To discourage companies from rejecting a settlement offer based on the arbift-ator's deci- sion, any company choosing to use the courts still would face joint and several liability. The Administration has stated candidly that this is intended as a stick to influence compa- nies to choose arbitrated settlements. In fact, if a company rejects what it considers an un- fair settlement offer, the bills expressly make the company subject to liability for the por@- tion of the cleanup that the arbitrator allocated to the orphan share. For example, suppose a company sought a court ruling on a $30 million cleanup because the company was allo- cated 10 percent of the liability when it caused only one percent, or $300,000, of the prob- lem. By rejecting the government's settlement offer to pay 10 percent, the company poten- tially would face liability for the entire $30 million. The proposed structure thus would be severely punish companies for seeking the legitimate protections of the courL Limited Steps in the Right Direction While these provisions, and the lack of reform of the underlying liability scheme would ensure that costs continue to explode for five more years, not all provisions in the bills move in the wrong direction. Although the additional provisions are only limited steps that largely would be unnecessary if real reforms were enacted, they nevertheless move in the right direction. One positive provision in the bills is lender liability reform, which would provide needed stability to the lending industry. This is significant because an unintended conse-

41 "House Subcommittee Approves Reform Bill; State Role, Voluntary Cleanup Titles Added," National Environment Daily, Bureau of National Affairs, May 13. 19%. quence of CERCLA, as now written, is that it dries up real estate loans. As interpreted by the courts, lenders who foreclose on properties held as collateral or exercise financial man- 42 agement, perhaps to rehabilitate a loan, are subject to Superfund's liability scheme. Thus, lenders have become hesitant to make loans based on collateral that might be subject to later litigation under the Act. Lenders are at risk of not only losing their original loan in- vestment, but of becoming liable for cleanup costs of perhaps millions of dollars. More- over, because banks typically are "deep pockets," they are likely to end up paying more of the cleanup costs. So, lenders now typically require that studies be undertaken to assure the property is free of any contamination. If it is not, an original or refinance loan on the prop- erty is virtually impossible to obtain. The benefit to the government of this situation is that banks investigate or require poten- tial applicants to investigate properties around the nation for contamination, thus helping it to identify contaminated sites more swiftly and cheaply. In effect, CERCLA has imposed an unfunded mandate on the lending industry. Under the pending legislation, lenders' liability essentially would be limited to their level of investment in the mortgaged property. This is an important step in the right direc- tion, it stops short of taking this concept to its logical conclusion. Lenders should not be h- able for the costs of cleaning land that may have been polluted without their knowledge or effective control. While the extent of liability may now be quantified, which provides needed stability, even limited liability distorts the economic markets and unfairly shifts the burdens of Superfund onto innocent investors. Another worthwhile provision would reduce the state's contribution to Operation and Maintenance costs of cleaned sites from 100 percent to a comparably small rate consistent with construction remediation costs. This change will reduce the incentive on states to de- mand unreasonably stringent cleanup standards to 'Yront-load" the costs so that the federal government will pay the vast majority of the project cost. 43 Yet another useful provision would create an exemption for so-called micromis, or truly tiny, contributors. While this is welcome relief, it would be unnecessary were it not for the current liability scheme. Moreover, the provision is not without problems. If Superfund's liability scheme is kept in place, truly tiny contributors should be relieved of the burden of proving a negative-that they did not contribute significantly to the pollution. Failing to Establish Reasonable Standards The House and Senate bills, unfortunately, fail to correct many of the problems created by inflexible and overly stringent cleanup standards. Moreover, while some issues substan- tively would be resolved, other standards and requirements would create new problems.

42 For a full discussion of the legal issues, see in Re Bergsoe Metal Corp., U.S. Ct. App., 9th Cir., 910 F.2d 668 (1990). 43 Under current law, although the states must Pay 100 percent of future O&M costs, they are only liable for 10 percent of the front-end cleanup costs. The bills set a 15 percent uniform share for state governments. Ilis aspect of the provision has bow criticized because it will increase the burden to the states for front-end construction costs. One of the most controversial provisions would deal with the goal of protecting human health through cleanup of Superfund sites. This new national goal "shall be expressed as a single numerical health risk level that ensures a reasonable certainty of no harm from expo- 44 sure to carcinogens and... noncarcinogens. The risk level would be decided through ne- gotiated rule-making which would allow EPA to set, and likely result in, an -excessively strict and expensive risk level of one in a million. This would cause the cost of Superfund cleanups to skyrocket because it would mandate the cleanup of extremely small risks. Another major problem with a national goal is that it would allow an inflexible "bright- line" for deciding when cleanup should or should not occur. A more prudent policy would base the cleanup decision on the risk reduced per dollar of cleanup cost. 45 Thus, for every dollar spent on Superfund cleanup, the most immediate risks to the public would be ad- dressed first. Essentially, the law should be structured to get the "biggest bang for every en- vironmental. buck." But as the Washington, D.C.-based non-profit National Research Coun- cil found, none of the agencies with responsibility over waste sites, including EPA, "have developed [their] over-all prionty-setting process in a manner that is explicit, adequately documented, and sufficiently open to scientific and public scrutiny-'A6 But these prioritiza- tion changes, plus similar changes in other environmental laws, would allow the greatest risks to the public to be reduced first, regardless of the type of environmental problem. In this way, the nation could more efficiently spend fewer doll= while pursuing a chosen acceptable, risk level. The bills grant exemptions to PRPs from achieving the new numerical risk standard if the necessary cleanup would be either technically infeasible or unreasonably costly. In either case, the EPA Administrator would be allowed to set a lesser cleanup level at a cost that, while possibly quite large, is not unreasonable. At first glance, this seems like a rea- sonable step in the right direction. But a closer look reveals that the bills further codify the use of unnecessarily strict cleanup standards. The exceptions granted in the bill are qualified by the language that the lesser cleanup standard chosen to avoid a unreasonable cost cannot "result in a unaccep risk to hu- @M=p man health ...... 7bus, any relief from unreasonable costs must result in an acceptable risk. Two questions about this exception must be asked: 1) Why should PRPs not granted an ex- ception pay huge sums if lesser standards would result in an acceptable risk? 2) What does an "unreasonable' cost mean? Either the bills are in most cases mandating expensive clean ups to meet excessive cleanup standards, or the exception is meaningless since both reason- able and unreasonable costs must be set at the same risk level-acceptable. Not all provisions in the bills, however, are so poorly crafted. To its credit, the Admini- stration effectively exempts industrial sites and other properties from Superfund's unneces- sarily stringent requirements. Currently, CERCLA is indifferent to the traditional or ex- pected future use of the contaminated property. CERCLA's standards effectively require

S. 1934, T"ide V, Sec. 501. 45 See John Shanahan, "How to Help the Environment without Destroying Jobs," Heritage Foundation Memo To: President-Elect Clinton No. 14, January 19, 1993. 46 Bureau of National Affairs. National Environmental Daily, September 12,1994, discussing National Research Council, Commission on Geosciences, Environment and Resources, "Ranking Hazardous Waste Sites for Remedial Action," September 8, 1994. that the dirt be clean enough to eat. 47 MeWle this arguably may make sense in a residential community, it is certainly unnecessary in an industrial park where there is little likelihood of exposure to the soil. As a result of these high standards, industrial operators invariably decide to locate their plants on pristine sites in order to avoid future liability for cleaning up another company's expensive mess. Often, the companies that contaminated plots have gone out business in the intervening years, so the properties have become "orphan sites," effectively increasing the cost to the public. 48 Even if these sites are cleaned up, companies have little incentive to locate on them for fear of future litigation. Since many of these older polluted industrial sites are located in urban areas, Superfund effectively acts as barrier to urban revitalization. Thus, the poor and minorities who disproportionately live in urban areas, are the unin- tended victim of this policy.49 The bills properly relax the unnecessary cleanup standards by allowing the EPA Administrator to take into account the intended and historical uses of in setting the standards. Thus, albeit by a circuitous path, the bills do take into ac- count real expected exposures for some sites. Financing Environmental Pork The bills also contain plenty of pork barrel spending for organized labor and environ- mentalists. The House bill now under consideration contains a grovision applying the Davis-Bacon prevailing wage law to Superfund cleanup sites.5 By mandating that PRPs pay union wages, Congress is moving in the diametrically opposite direction of containing Superfund costs, reducing litigation, and accelerating cleanup. There can be no justification for mandating that cleanups be more expensive than necessary. Perhaps the most egregious provision, however, effectively creates an environmental welfare program diverting Superfund's money to environmental activists.51 Grants totaling up to 4 percent of the Superfund pay-out in any given year, or y $63A mil- lion, would be given to private citizens to investigate hazardous waste sites, including those not on the National Priorities List. 52 The grants may be Used to hire "experts" for in- formational reasons, to hire a community liaison to PRPs, "to hire experts to file comments

47 While there is no official standard that the soil be "clean enough to M" it serves as an effective criterion because of the exposure assumptions built into die standards, such as ingestion of a certain amount of soil per day. 48 While the costs we directly levied on businesses, much of this cost is passed on to the consumer through increased prices. Additionally, them at the losses to the public in the form of reduced dividends to shareholders and income to workers for costs that companies cannot pass on to the consumer 49 Unfortunately, the Administration through S. 1834 and H.R. 3800 unwisely attempts to address ..environmental justice" through the creation of a Superfund affirmative action plan. An in-depth analysis of the problems created by "environmental justice" provisions. however, and other provisions effwfively targeted to the same groups, is beyond the scope of this paper. So The Clinton proposal did not include Davis-Bacon provisions. 51 This provision is in both bills. 52 S. 1834, Title I, sec. 101. Those sites not listed on the NPL would be limited to I/Sth of total grant expenditures. The dollar figure estimate cited in the text is based on 4 parcent of estimated Superfund outlays for Fiscal Year 1995, or $1,585,979,000. Naturally, this dollar figure would climb as Superfund figures climbed. with the federal government and generate other documents,"' and "for training funds for in- terested affected citizens to enable them to more effectively participate in the remedy selec- tion process." None of the rules that govern agencies in selecting experts would apply since all decisions would be made by private citizens, although "guidelines" would be pub- lished. Thus, the grants would be open to large-scale abuse. The practical effect, if not the purpose, of this provision would be to make it easy for en- vironmental activists to get paid with federal tax dollars for what they now do with volun- tary donations. Environmental activists would be able to increase the scope of their activi- ties as well as their payrolls. Another provision in both bills pursues a completely different environmental objective- recycling. The bills exeLnpt from liability recyclers who inadvertently contaminate property with hazardous waste. 53 Essentially, this provision would reward recyclers, for their envi- ronmentally benign intent despite the actual damage done by their physical actions. It is ironic that polluters pursuing other societal objectives, such as national defense and health care, which arguably are as important to the nation as recycling, are denied this favorable status. Indeed, this provision would shift the recycler's costs onto those other sectors be- cause, as written, they would have to pay the recycler's shares. 54

A WORKABLE SOLUTION The best solution to the current problems with Superfimd would be to repeal retroactive, joint and several liability. With the repeal of retroactive liability, all sites contaminated be- fore 1980-when Superfund was enacted-would be "orphaned," meaning that no one is liable under the law. With the repeal of joint and several liability, moreover, whenever ac- tual polluters that partially contaminated sites after Superfund was enacted could not be found, their shares of the cleanup costs would become orphan shares. Cleaning up these or- phan shares would thus require expenditures from the Superfund created for just this pur- pose. If society wishes to clean up orphan sites for which no responsible party can be found, then society should pay the tab-Congress should not use a tax to shift the cost to another set of innocent "deep pockets." The fund, however, should not be given additional appropriations to account for this li- ability scheme change. Rather, substantive reforms should be-enacted to make cleanup standards more flexible and less stringent. These two changes, together with the lower EPA oversight costs that would follow, would reduce the average cost of cleanups consid- erably. Limiting Superfund finances in this way would force EPA to prioritize site clean- ups according to its assessment of which sites posed an actual present threat to public safety and would force the agency to use judgment in reducing each site's cleanup costs.

53 S.1834, sec. 410, adding new section, sec. 130(j). 54 Moreover, recyclers who have been unsuccessfully sued to pay for site cleanup would be allowed to recover attorney's fees. However, this treatment is not given to Other PRPs that have successfully sued a non-qualifying racycler. While the "English rule" of allowing the recovery of attorney's fees is preferable to the dominant "American rule" of no recovery of fees as a general matter, there is no justification for this unequal treatment. The section also has an arbitrary cut-off date of February 3, 1994; this date coincides with the bill's introduction. Congress should give EPA the flexibility to make these judgments. Adding more money to Superfund, or creating the Environmental Insurance Resolution Fund, would remove this pressure for substantive reform.

CONCLUSION The Superfund reauthorization now being considered by Congress has been touted as a major reform of the costly Superhind program, which is widely considered one of-if not the-most ineffective and unfair laws today. Unfortunately, the reforms being considered are as disappointing as the current law. If the proposed reauthorization becomes law, it will: create yet another fund-this one financed through a new $8.1 billion tax on the in- surance industry; create an "environmental porle' grant program that will benefit environ- mental activists, and inequitably shift recyclers' shares of costs of cleanups to others who often are already paying more than their fair share. Moreover, those few reforms that make positive strides, such as limiting lender liability, are only half-measures. The most striking aspect of the reform, however, is what it will not do. It will not reform the retroactive, strict, joint and several liability system that is causing much of the litiga- tion now plaguing Superfund. Nor will it do much more than tinker around the edges of the fundamental cost drivers of cleanups-overly strict and inflexible standards. Hence, the most important accomplishment of this reform will be to lock in the most unworkable provisions of the current law for another five years during a time of exploding costs, thus ensuring even greater waste into the future.

John Shanahan Policy Analyst

John Barry contributed to this study.

abolAptXmTra a @a4 dawn:'Mmd i" 0. AW, 0 H v 09,

About the Author