|
28 ELR 10067 | Environmental Law Reporter | copyright © 1998 | All rights reserved Applying Cost Causation Principles in Superfund Allocation CasesRichard Lane White and John C. Butler IIIEditors' Summary: The question of how to fairly apportion cleanup costs at Superfund sites is a highly debated topic in the law of hazardous substances. This Article highlights the deficiencies found in common allocation methods, and offers cost causation as a rational approach to apportioning cleanup costs. After providing a background on the CERCLA liability scheme, the authors address the various equitable factors used to apportion cleanup costs and discuss cost causation's relationship with those factors. The authors then introduce cost causation in an examination of how cleanup costs are created at a site and who is responsible for the specific costs. Next, they use figures and hypothetical scenarios to explain cost causation analysis and to note the deficiencies of volumetric- and toxicity-based allocations of costs. They conclude that although cost causation analysis may not address every equitable issue, it addresses many of them, and is flexible enough to incorporate others. Richard Lane White is a director in the Cambridge, Massachusetts, office of Putnam, Hayes & Bartlett, Inc., an international economic and management consulting firm. John C. Butler III is a managing director of Putnam, Hayes & Bartlett, Inc., in its Los Angeles, California, office, and he heads the firms environmental practice. Both Mr. White and Mr. Butler serve as allocation experts in Superfund contribution cases. The views expressed here are those of the authors, and not necessarily those of Putnam, Hayes & Bartlett, Inc. The allocation of response costs at Superfund sites remains a contentious and unsettled issue. Statutory direction on allocation is minimal. What courts and responsible parties need is a set of clear allocation principles, and an approach that can provide a rational method for allocating these costs. This Article focuses on some of the key reasons that volume, weight, and toxicity analyses often fail to allocate cleanup costs appropriately and that employing cost causation principles to allocate costs is a rational approach. A set of benchmarks is developed to allow the reader to evaluate specific sites and determine whether the application of more rigorous causation analysis is merited. Basic Principles of Liability and Allocation Under the Comprehensive Environmental Response, Compensation, and Liability Act Allocating Superfund cleanup costs is a contentious issue because it is a zero-sum game.1 A party wins only when another loses. Allocation is complicated by the fact that there is little guidance available on how to distribute these costs, so potentially responsible parties (PRPs) must examine a range of allocation methodologies. The amount of material contributed by each PRP is usually the measure for comparing parties on a "waste-in" list, typically the starting point in developing an allocation. These contributions usually are measured as either volume or weight.2 When the wastes are liquids or homogeneous mixtures, volume is more frequently [28 ELR 10068] the measure. Weight is often used when waste types vary and it is necessary to calibrate them (for different densities) to place them on a common basis. But, as this Article highlights, some fundamental limitations to using a volumetric or weight measure as the basis for cost allocation limit its applicability. In many cases, a gallon of waste from Party A is not the same as a gallon from Party B. And while a waste that contains a hazardous substance, no matter how little, creates Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) liability, it is obvious that some wastes contain more hazardous substances than others and that some wastes are more toxic than others. Therefore, they should not be treated the same when it comes to apportioning liability.3 Adjusting waste-in data to reflect crude measures of hazardousness—for example, trying to limit the waste-in list only to wastes deemed hazardous under some definition—is also of limited value, because such adjustments are arbitrary and depend on the definition of hazardous waste that is employed. More specific toxicity adjustments to wastes under the guise of scientific principles fall short for the same reason, and they all do not lead to the same level of response costs. Moreover, more precise measures for evaluating toxicity generally are misguided by the notion that Superfund allocation should reflect an allocation of liability or harm rather than an allocation of response costs. Many allocation cases focus primarily on generator PRPs, because other classes of parties, such as owner/operators and transporters, are often insolvent and those shares become "orphan" shares that are absorbed by the viable PRPs. In apportioning liability among these PRPs, courts and experts are increasingly turning away from volume or weight and turning toward methods for relating contribution to cleanup costs. The application of cost causation principles is not the entire solution to allocation, but it does address many of the factors that parties and courts consider when apportioning liability. Further, it often is a good starting point for crafting an allocation that can incorporate all relevant equitable factors. When CERCLA was passed by Congress in 1980, it was designed to address remediation at the nation's worst hazardous waste sites. Though amended and reauthorized in 1986, the fledgling Superfund program still left a number ofkey issues unaddressed. Probably the single most contentious issue facing those involved at Superfund sites was, and continues to be, the allocation of cleanup costs. The allocation issue is contentious for a variety of reasons. First, CERCLA broadly identifies four classes of PRPs, including past and current site owners, past and current site operators, those who arranged for disposal of wastes or those who generated the wastes, and transporters of waste (if they helped select the site of disposal).4 CERCLA was designed so that liability would extend to any party who had any relationship to waste disposal; almost everyone who has anything to do with hazardous substances is potentially liable under CERCLA.5 What gives rise to CERCLA liability is the disposal, transportation, or handling of "hazardous substances." In other words, hundreds of substances give rise to liability.6 There is no "threshold" requirement for hazardous substances, and particular waste streams need not be identified on the hazardous substances list; courts have concluded that any material containing a hazardous substance is, for purposes of establishing CERCLA liability, a hazardous substance.7 Even disposal of low-toxicity waste such as municipal solid waste (MSW), which has been shown in numerous studies to contain small quantities of hazardous substances, imposes liability under CERCLA.8 And finally, there is a presumption that wastes contain hazardous substances.9 Under CERCLA, there is no causation requirement to a showing of liability. Anyone who disposes of hazardous substances, in whatever quantity, is potentially liable. And one need not "fingerprint" a party's waste or show that a particular party's waste stream caused or led to the need for remedial action.10 What is required is a showing that there is a release, or a threat of a release, at a site, which has created the need for response.11 Thus, what we have is a statute that is so expansive that it triggers liability for almost anyone associated with waste disposal if that waste makes its way to a Superfund site.12 In addition, it triggers liability for almost any waste stream, regardless of the quantity of hazardous substances in that waste stream, and without the need for showing that the waste in question actually led to the need for remedial action. But the backbone of CERCLA lies in its three key liability provisions. Liability under CERCLA is strict, joint and several, and retroactive. Because liability is strict, it is enough to show involvement, not negligence.13 Joint and several liability allows the U.S. Environmental Protection Agency (EPA) to pursue parties who contribute to a site and hold them entirely liable for all costs.14 And the statute can be applied retroactively. Actions that were acceptable, legal, "industry standard," or "common practice" can now, after the fact, be found to be illegal.15 Superfund's liability provisions effectively create the liability web that traps almost everyone who comes in contact with it. Yet Superfund cost allocation is contentious not only because it has such draconian liability provisions, but because the costs associated with Superfund are so large. Were the costs trivial, this liability morass would simply be an annoyance; but with average site remediation costs in excess of $ 30 million,16 the stakes for PRPs are quite high, so PRPs take extreme measures to minimize their exposure at these sites. Currently, there are 1,053 general Superfund sites on the national priorities list (NPL).17 At $ 30 million per site—with some sites actually estimated to cost hundreds of [28 ELR 10070] millions of dollars to remediate—the total bill will be more than $ 30 billion and it continues to escalate. Statutory Direction on Allocating Costs For a statute that goes into such detail to define the bounds of liability, CERCLA is noticeably quiet when it comes to crafting an approach to allocating this liability among the various PRPs. Liability is strict, so issues such as volume of waste, toxicity of waste, relative fault of the parties, or causation have no place in determining liability. Allocation, however, is not strict; it is "equitable." All of the factors that are ignored at the liability phase are fair game as allocation factors. In many instances, PRPs work together to craft an allocation, often with a neutral allocator. In some instances, however, parties remain at odds and take their dispute to courts or to arbitration. Allocation of Superfund cleanup costs has been assigned, by statute, to the courts, where they "may allocate response costs among the liable parties using such equitable factors as the court determines are appropriate."18 But beyond using "equitable factors," the statute is silent as to the actual approach to take in allocating response costs. So courts and allocators have been left to determine what equitable means and what factors are relevant or appropriate in developing equitable allocations of response costs. Equitable Factors and Superfund Sifting through the myriad of equitable factors to determine which of them are appropriate to a particular allocation is an examination that is often fact-and site-specific. However, just as the waste-in list is the starting point for developing an allocation, the Gore factors provide a starting point for determining equitable factors. These are a set of six factors delineated in the unsuccessful amendment to CERCLA proposed by then-Rep. Albert Gore (D-Tenn.).19 They are: 1. the ability of the parties to demonstrate that their contribution to a discharge, release or disposal of a hazardous waste can be distinguished; 2. the amount of the hazardous waste involved; 3. the degree of toxicity of the hazardous waste involved; 4. the degree of involvement by the parties in the generation, transportation, treatment, storage, or disposal of the hazardous waste; 5. the degree of care exercised by the parties with respect to the hazardous waste concerned, taking into account the characteristics of such hazardous waste; and 6. the degree of cooperation by the parties with federal, state, or local officials to prevent any harm to the public health or the environment.20 As courts have previously noted, the Gore factors are not exhaustive,21 and at their discretion courts can use one, several, all, or none of these factors.22 Courts have also identified a number of other equitable factors that they have considered in the context of allocation, including: 1. the financial resources of the parties involved;23 2. the benefits received by the parties from contaminating activities;24 3. the knowledge and/or acquiescence of the parties in the contaminating activities;25 4. the existing contracts between the parties on the question of liability, such as indemnity agreements;26 5. the relative fault of the parties (including causation);27 6. the degree to which each party made efforts to prevent and/or contain any known release of hazardous wastes at the property, at the time the releases [28 ELR 10071] occurred;28 7. the degree of care exercised by a party in managing a Superfund site and their duty as landowners;29 8. the relative economic benefits of PRPs;30 9. the state of mind of the parties;31 10. the benefit received by the owner from the operator's activities;32 and 11. the benefit to the current owner if after cleanup the land is cleaner than at the inception of the operation that caused the problem.33 Under CERCLA, the authority to develop the allocation of response costs is assigned to the courts; EPA does not generally play a role in such allocations. EPA has, however, addressed issues of liability and allocation from time to time. For example, EPA has published guidance documents on issues such as developing waste-in lists, settling with de minimis and de micromis PRPs, and developing nonbinding allocations of responsibility (NBARs). And over the past several years, EPA has attempted to address the issues of liability and allocation for parties involved with MSW. EPA's NBAR guidance is probably the best starting point in any examination of EPA's "thinking" on allocation. While many examples are discussed in terms of volume, the NBAR guidance clearly identifies cost causation as the primary tool for allocating costs among responsible parties. As the NBAR guidance notes, "where it is possible to do so, waste types and volumes that necessitate particular remedial activities will be fully attributed to the appropriate contributors."34 And as other commentors have noted, "EPA has consistently—if obliquely—articulated the need to connect a party's allocation to costs which its wastes cause since EPA adopted its Interim Policy on Nonbinding Preliminary Allocations of Responsibility."35 Most recently, EPA has proposed controversial guidance on MSW generators and transporters.36 Although purportedly addressing liability, it actually focuses on addressing allocation of response costs for parties associated with MSW wastes. EPA proposes to cash out MSW generators and transporters at a cost of between $ 3.05 and $ 3.25 per ton for wastes taken to Superfund sites. EPA characterizes this proposal as setting forth "a fair and efficient method for calculating an equitable and reasonable settlement contribution" for MSW generators and transporters.37 The share that EPA believes should be assigned to MSW parties is specifically designed such that it "reflects a reasonable approximation of the cost of remediating MSW."38 While there are a number of problems with EPA's proposal, it clearly illustrates EPA's view on cost allocation.39 In EPA's view, the appropriate cost to assign to a party is the cost that reasonably approximates what damage that party caused. Cost Causation and Equitable Factors The Gore factors do not specifically identify "cost causation" as an equitable factor, but it is an amalgam of several Gore factors and other equitable factors. For example, volume, or amount of material contributed, is a Gore factor. Why is volume of material relevant to an allocation? Because we do not measure frequency of disposal—whether one party used the site more than another. Instead, we measure how much material each party sent to the site. Implicitly, at least, we assume that the amount of material is related to [28 ELR 10072] the remedial activities for which the PRPs are being asked to pay. Why do we measure toxicity, which is yet another Gore factor? Is it simply because those parties who sent more toxic or more hazardous materials should pay more, or is it that more toxic and more hazardous materials disproportionately affect the need for and the level of the remediation? What is it that allocators and courts are trying to address? Whether a pound of solvent (e.g., trichloroethylene (TCE)) is worse than a pound of arsenic, or whether an allocation should consider how different wastes give rise to different costs? And why is the ability of a particular party to distinguish its contributions from those of other parties a Gore factor? Given volume and other factors, what is the value in showing the particular results, or impact, of a particular party, if not to relate that contribution to the cost that has arisen? Stepping back from the Gore factors themselves and examining other equitable factors used by courts, we repeatedly see that courts turn to the interrelated concepts of causation and relative fault. What exactly is causation? In the context of Superfund cost allocation, it is the remediation costs that have been caused by the disposals and actions of various PRPs. A party may dispose of a waste stream containing a hazardous substance, but the real question is: does that waste stream cause any problem or contribute to the need for a remedial action by itself or in combination with other waste streams? Relative fault, on the other hand, encompasses a range of issues, including such actions as lack of care. As the court in United States v. Cannons Engineering Corp. noted, for the court to approve a CERCLA settlement, "terms must be based upon, and roughly correlated with, some acceptable measure of comparative fault."40 Again, the question is fault for what? Fault for response costs. Under CERCLA, courts "allocate response costs among the liable parties."41 Cost causation—an examination of how costs are created and who or what parties are responsible for specific costs—addresses each of these equitable factors. The amount of material (volume) and its characteristics (toxicity or hazardousness) are evaluated in tandem, not in isolation. The allocation exercise is not about allocating pounds or toxic pounds; it is about allocating costs. A causation analysis is designed to distinguish the contributions to cost of one party from another. The interplay of causation and relative fault becomes a central focus of cost causation. So while cost causation is not specifically delineated in the Gore factors, it is central to an examination of equitable factors in the context of CERCLA cost allocation.42 Cost Causation and Stand-Alone Cost Analysis Before delving into an examination of various allocation scenarios, it is worthwhile to briefly discuss just what cost causation is and how it works, beginning with an examination of a volumetric allocation. Consider Figure 1. In this example, the cost we are trying to allocate is the record of decision (ROD) remedy estimate. The ROD cost estimate is $ 1.5 million to remediate the site, and the quantity of waste at the site has been estimated at 15,000 cubic yards of soil. The average cost per cubic yard for the remediation is $ 100 ($ 1.5 million/15,000). In the graph, this is demonstrated as the line labeled A "Volumetric" Allocation. In a volumetric allocation, each of the parties is assigned the average cost per unit of waste. In this example, we have identified two responsible parties. Both Party A and Party B are generators and the only identified parties at the site. Party A generated 5,000 cubic yards of waste; Party B generated 10,000 cubic yards of waste. The sum of their contributions equals the waste at the site—15,000 cubic yards. [SEE ILLUSTRATION IN ORIGINAL] In this example, Party A is assigned $ 500,000 ($ 100 per cubic yard x 5,000 cubic yards), which represents one-third of the $ 1.5 million remedy since it is responsible for one-third of the waste volume. Party B is assigned $ 1 million, representing the other two-thirds of the site's remediation cost. This allocation assumes a direct relationship between the quantity of waste and the cost of the remedy. It also assumes that all wastes are—at least from a remediation cost standpoint—identical. From an economics viewpoint, these costs are assumed to have constant unit costs and no economies of scale; that is, regardless of how many cubic yards are remediated, the cost of remediation will be $ 100 per cubic yard. Examining Basic Cost Relationships Costs, however, seldom exhibit constant unit costs because the scope of a project varies and some costs vary with quantity (i.e., scale), while others are "fixed." In our remediation example, it may be the case that the hauling charge and tipping fee for disposal of the wastes is constant, at $ 80 per cubic yard. If this is the case, that accounts for $ 1.2 million ($ 80 x 15,000 cubic yards) of the $ 1.5 million cost. The remaining $ 300,000 may be a fixed cost, such as the costs for set up and initial site work in preparation for the excavation and disposal work. In the case of fixed costs, a volumetric allocation may not adequately reflect cost relationships. In our example, if the cost of the remedy is not $ 100 per cubic yard, but is instead $ 300,000 plus $ 80 per cubic yard, then a 5,000 cubic yard excavation does not cost $ 500,000, but instead costs $ 700,000—an average of $ 140 per cubic yard. Likewise, a 10,000 cubic yard excavation does not cost $ 1 million, but instead costs $ 1.1 million—an average of $ 110 per cubic yard. Many costs do have a fixed cost component. They exhibit a declining average cost as the quantity of waste rises (within the relevant range). As another example, engineers use a "cost-capacity" factor to estimate how costs vary as the scope of a treatment plant or other process changes. This engineering economy of scale also shows the characteristic declining average cost as scope grows.43 Figure 2 shows several alternative cost functions that result in a ROD remedy estimate. One cost equation demonstrates the effect of a $ 300,000 fixed cost and an incremental or marginal cost of $ 80 per cubic yard. The other cost equation demonstrates an economies-of-scale effect akin to that in an engineering cost-capacity relationship. For Party A, either of these cost equations results in an allocation much different than the volumetric estimate shown in Figure 1. [SEE ILLUSTRATION IN ORIGINAL] The point where each cost function crosses the 5,000 cubic yard mark shows Party A's "stand-alone cost"—the cost that Party A would pay if it were the only party at the site. The same cost functions have been extended to show the impact for Party B, in Figure 3. The point where each cost function crosses the 10,000 cubic yard mark shows the cost that Party B would pay were it the only party, and this is its stand-alone cost. [SEE ILLUSTRATION IN ORIGINAL] In the case of a volumetric allocation, the two allocations are additive: the cost associated with Party A, plus the cost associated with Party B, adds up to the cost associated with the ROD remedy. When cost functions are used to estimate what each party would pay, and those cost functions are not directly proportional to volume, the sum of the individual stand-alone costs will not sum up to the actual remedy. How can these estimates be used to "allocate" the ROD remedy? Stand-Alone Cost Analysis—An Example Some costs are caused uniquely or directly by a single party; those costs should be borne by that party. That is not standalone cost analysis; it is simply a basic premise of cost causation. Other costs that cannot be assigned or distributed uniquely to specific parties are common costs. There are a variety of methods for allocating common costs, some being [28 ELR 10074] better than others. For instance, consider stand-alone cost (SAC) analysis. Suppose two parties need a treatment plant to purify their water supplies. Party A could build its own treatment plant, at a cost of $ 15, as is shown in Figure 4. This represents the SAC for Party A—the cost to Party A if it works in isolation. Likewise, Party B can build a water treatment plant to suit its own needs. It needs a somewhat smaller treatment plant, and the cost for that plant is only $ 10. Thus, Party B's SAC is $ 10—the cost to Party B if it works in isolation. [SEE ILLUSTRATION IN ORIGINAL] Another alternative would be for the two parties to work together to build one large treatment plant capable of servicing the needs of both parties. This larger treatment plant would cost only $ 20, because of the economies of scale in building water treatment plants, and would result in a savings of $ 5 as compared with the cost of both parties building their own plants. The larger combined treatment plant is a "common" cost that is shared by Party A and Party B. This common cost is "caused" by the two parties, and it can be allocated by principles of cost causation. Certainly, some contend that common costs are not caused, because they are not uniquely attributable to a single party. But, in the example, the common cost arises because two parties jointly undertake the activity—the cost does not exist unless both act. These parties cause the common cost and, therefore, should collectively be assigned that cost. Common costs also can be allocated using principles of cost causation or cost engineering principles. SAC analysis examines what the alternative is for each party working in isolation—if that party causes the alternative cost to exist. These estimates then are used to allocate the common cost. There are rational and principled approaches for equitably allocating common costs by looking at the costs each party would have created on its own. Is SAC the only method that can be used? No. There is no unique way to allocate the cost between the two parties. For example, Party A could argue that Party B was willing to spend $ 10, so Party A should only have to pay the remaining $ 10. Or, it could be argued that since both parties are using the plant they should share the cost equally. And Party B could argue that Party A was willing to spend $ 15 on its own, so Party B should only have to come up with the remaining $ 5. Here, the SAC estimates for each party serve as reservation prices. If either party is allocated more than its SAC it would simply choose to build its own plant. There would be no incentive to save the $ 5 in that case. As one can see, there are many ways that the $ 5 savings associated with joint action of Parties A and B could be allocated. By using SAC analysis, we can equitably allocate the cost for the larger treatment plant. In this example, Party A was willing to pay $ 15 and Party B was willing to pay $ 10, for a total of $ 25. Thus, Party A's SAC is 60 percent of the total. That share can be used to allocate the actual cost. This results in Party A being assigned $ 12, and, through similar rationale, Party B is assigned $ 8. This allocation has some interesting consequences. First, each party is paying the same share of its SAC toward the common cost. Party A's SAC is $ 15, so $ 12 is 80 percent of its SAC. Likewise, Party B's SAC is $ 10, so $ 8 is also 80 percent of its SAC. And while in this example the common action provides a benefit, the same principle of allocation would hold if the actions of the two parties resulted in commingled wastes, creating a burden—a cost in excess of the sum that each would incur on its own. The second consequence is that each party shares in the "benefit"—the $ 5 saved by the joint action—in the same proportion as actual costs are shared. Party A receives a benefit of $ 3 (i.e., its SAC is $ 15 and its allocated cost is $ 12), representing 60 percent of the benefit. Similarly, Party B receives a benefit of $ 2, representing 40 percent of the benefit. SAC analysis is a reproducible approach that can be applied consistently from case to case. It also has a number of features, in addition to those noted above, that make it an equitable approach to allocating common costs. First, it stops the "free-rider" effect that can plague common cost allocations. Rather than any party "piggy-backing" on another, each party is evaluated relative to what it would do on its own. Second, the SAC approach can be used to allocate positive benefits (such as the $ 5 above) or negative benefits (the burden created by commingling the wastes of several parties that result in higher costs than the sum of the individual SAC estimates), and its principles remain constant whether the benefits are positive or negative. Third, the SAC approach explicitly accounts for the fact that parties can undertake the same activity, but on a different scale. This is reflected in the allocation. It accounts for the economies of scale often inherent in Superfund cost problems. Fourth, the SAC approach explicitly accounts for the fact that different parties can undertake different activities, and it uses this fact to allocate the common activity undertaken. So, here again it addresses one of the central issues in cost allocation, recognizing that different wastes at Superfund sites may lead to different remedies. A Cost Function That Shows a Change in Remedy The typical "economies-of-scale" cost functions were discussed previously, but an alternative cost function that can be quite common at Superfund sites occurs when contaminant "thresholds" dictate specific remedies and, within a particular remedy, volume has little, if any, relationship to cost. Figure 5 shows two new cost functions. Each approach shows that while a small amount of contaminant results in negligible cost, once the contamination reaches a threshold level the cost rises (because the remedy changes), and the cost from that point forward is relatively "fixed" regardless [28 ELR 10075] of the contaminant saturation. For example, if a small quantity of volatile organic compounds enter the soil at a site the costs may be trivial, but a larger quantity would require a pump-and-treat operation in addition to soil treatment. Although the plume becomes more concentrated (which may affect operation and maintenance costs), the basic cost of the pump-and-treat operation is an all-or-nothing proposition. [SEE ILLUSTRATION IN ORIGINAL] Under one formula, the "threshold" is met with a contribution less than that contributed by Party A. Therefore, Party A should shift to point A-1. As you can see, a volumetric allocation seriously understates the contribution of Party A. Under the alternative approach, Party A has not met the threshold yet, and the volumetric allocation seriously overstates its contribution. Party A should be at point A-2 on the diagram. In neither case will a volumetric allocation result in an allocation that is consistent with the cost contributions for either Party A or Party B. Different Cost Functions for Each Party The discussion thus far has highlighted the problems of using volumetric allocations, even for a single waste stream, when the cost relationships are not linear. Now, ignore the nonlinearity of the cost function and turn to what is a more common and fundamental problem—different wastes from different parties. At Superfund sites, different parties routinely contribute different wastes that subsequently become commingled. Figure 6 shows a situation where Party A and Party B have contributed different wastes that have different cost functions to a site. Assume that Party A is sending a small volume of high-cost waste, while Party B is sending a large volume of low-toxicity, low-remediation-cost waste. The difference between Party A's allocation under the volumetric function versus the cost function is substantial—Party A gets a deal in the volumetric allocation because that approach fails to account for costs associated with its waste. For Party B, the difference is also substantial. A volumetric allocation seriously overstates the contribution of Party B. Under the volumetric allocation, it pays substantially more than Party A. Failure to account for differences in waste types and remedies creates a substantial allocation bias and is one reason why allocation disputes are not resolved quickly as parties search for allocation methods that more fully account for each party's contribution to cleanup costs. Allocations based on cost causation can address the economies of scale in remediation, as well as the different remedies for different waste issues. [SEE ILLUSTRATION IN ORIGINAL] Analyzing Cost Allocation Scenarios Consider an examination of specific allocation scenarios that allow us to compare volumetric, toxicity, and cost allocation analyses. A very simple example will demonstrate, in subsequent examples, how these allocation approaches differ when modifying site-specific circumstances. Scenario 1—The Divisible Site, Part I Party A dumps 5 drums of waste solvent onto a parcel of property; Party B also dumps 5 drums of waste solvent onto an adjacent parcel of property. Later, the entire area must be remediated as an NPL site with two distinct areas of contamination. Each party's action, on its own, results in 100 cubic yards of soil contaminated to a level at which it must be excavated and taken to a Resource Conservation and Recovery Act (RCRA) landfill. A total of 200 cubic yards of contaminated soils require remediation. There are no other costs to allocate between the two parties. In this scenario, a volumetric allocation, a toxicity-based allocation, and a cost causation allocation provide identical answers: Parties A and B should each be assigned 50 percent of the cost of remediation. Ten 55 gallon drums of waste solvent have been sent to the site; each drum weighs approximately 450 pounds. Each party contributed 5 drums of waste solvent. Rather than using drums as the measure of material, if we used gallons (assuming each drum had 55 gallons of material) or weight (assuming approximately 8.3 pounds per gallon), the volumetric [28 ELR 10076] allocation results would be the same. On a waste-in list each party would be assigned 50 percent responsibility. This result is shown in Table 1.
Each party is sending the same material in this example. Therefore, the toxicity factor is the same for both parties on a per-unit basis (e.g., per gallon). Further, the amount of material that each party is sending is identical, so whatever effect a toxicity factor might have on "weighting" waste shipments has no impact here. Each party's actions (the dumping of the 5 drums of solvents) result in the same remedial action. On its own, each would need to remediate 100 cubic yards of soil at a cost of $ 200 per cubic yard. And in each case, the treatment would be the same—taking the contaminated soil to a RCRA landfill. In this example, the cost of treatment is $ 20,000 for each party. Each party has created a unique and identifiable cost that can be directly assigned to that particular party. This is a direct cost. There is no area where both parties have jointly created a cleanup cost, so there is no joint action or interaction between the two parties. There are no common costs. Each party can be tied uniquely to an area of remediation that it alone created, and each should, therefore, be assigned the costs related to that area. In this case, that results in each party being assigned 50 percent of the total cost of the remediation. Under each of the above approaches, the result is the same: each party should be assigned 50 percent of the response costs. The results match not simply because each party sent 5 drums of solvent, but because each party sent 5 drums of the same solvent and disposed of that waste in such a way that it created the same problem, and the remedy necessary to address that problem was the same both in nature and in scope. Rarely, in the context of Superfund sites, do all these factors line up so precisely. Furthermore, while each approach yields the same result, the cost causation approach is correct; the others only coincidentally match. Because costs are being allocated, examining costs on a cost causation basis is the most appropriate method. When volume is a good proxy for cost, it works well. When toxicity approximates cost, it too can work. This is a view not simply held by allocators, but by EPA itself; the scenario examined is nearly a textbook NBAR example: In a limited number of cases, it is possible to link particular remedial activities with specific waste types and volumes. For example, in the easy but rare case of divisible waste, the cost of removing barrels from a warehouse on a larger site can be separately attributed to the contributors of the barrels. Or, the cost of incinerating soil contaminated solely by [polychlorinated biphenyls] (PCBs) can be attributed to the PCB contributors. Where it is possible to do so, waste types and volumes that necessitate particular remedial activities will be fully attributed to the appropriate contributors.44 In the previous example, Party A sent drums of solvent; it could just as easily have sent the barrels that were in the warehouse. Party B also sent drums of solvent, but it could just as easily have sent the PCBs that were in the soil. Each contribution can be segregated, so the costs associated with those wastes and volumes that were disposed of should be fully attributed to those particular parties. The site defined in this scenario should be a good candidate for divisibility. As the NBAR guidance notes, when contributions from the parties can be segregated, it is that "easy but rare" case in which divisibility may apply45 Note that use of terminology; EPA couches its discussion of allocation as an easy but rare occurrence. After all, where costs can be clearly allocated, or where waste contributions can be clearly segregated, those contributions can be divided. And when contributions are divisible, joint and several liability does not apply.46 Joint and several liability is a key enforcement tool used by EPA. It is not surprising that they would characterize clear allocation scenarios as rare.47 Scenario 2—The Divisible Site, Part II Party A dumps 5 drums of waste solvent onto a parcel of property; Party B also dumps 5 drums of waste solvent onto an adjacent parcel of property, then later returns and dumps 5 more drums onto the same parcel, for a total of 10 drums. Later, the entire area must be remediated as an NPL site with two distinct areas of contamination. Each party's action, on its own, results in 100 cubic yards of soil contaminated to a level where it must be excavated and taken to a RCRA landfill. A total of 200 cubic yards of contaminated soils require remediation. There are no other costs to allocate between the two parties. Analysis The only difference between this and Scenario 1 is that Party B has dumped twice as much material at this site. How that action affects the allocation depends on whether you believe volume, toxicity, or cost causation should be the guiding principle for allocation. In this scenario, these allocations do not provide the same answer. Both the volumetric and toxicity analyses will give Party B a greater share of the allocation, while a cost causation analysis recognizes that the costs created by each party are the same in this example. Each party initially contributed 5 drums of waste solvent, but Party B returned and contributed an additional 5 drums, for a total of 10 drums of waste solvent. On the waste-in list Party A is assigned a 33.3 percent share, Party B is assigned a 66.7 percent share. Whether costs are allocated on a volume or weight basis, the results are the same. This result is shown in Table 2.
Although Party B sent twice as much material to the site, the type of material that each party sent to the site is identical. Thus, a toxicity weighted volumetric allocation would produce identical results: 33.3 percent to Party A and 66.7 percent to Party B. It is worth noting, however, that while Party A has contaminated 100 cubic yards of soil, which, for purposes of discussion, we can say is contaminated at 50 parts per million (ppm), Party B has not contaminated twice as much soil at the same level; it has contaminated 100 cubic yards of soil, but that soil is twice as contaminated—at 100 ppm in this example. Just as in the previous example, each party's actions result in the same remedial action both in nature and in scope. The only difference is that Party B's soil is more heavily contaminated. In many instances, this additional "loading" on the soil could result in a greater amount of contamination or a different and more expensive remedy. If that were to occur, cost causation would account for that fact. But in this example, although the soil is more heavily contaminated, it does not affect the remedy. Because the remedy is the same, it does not affect the cleanup cost. Under a cost causation approach, therefore, each party is assigned 50 percent of the site cleanup cost. Both the volumetric and toxicity allocations recognize that one party has sent more material to the site than another. That the cost causation allocation does not adjust in this example might cause one to think that the cost causation allocation is incorrect. Would it be any more correct if the cost of the remedy had increased for Party B, but it did not result in the same allocation as a volumetric- or toxicity-based approach would conclude? Of course not. It is easy to see why a volumetric- or toxicity-based approach is incorrect here. Assume that the costs were allocated on the basis of volume. Party A is assigned 33.3 percent, and Party B is assigned the remaining 66.7 percent. And remember, each party deposited its waste in distinct areas—so distinct, in fact, that this site is divisible. If the site is divisible, how much waste Party B sent is irrelevant. Party B should be assigned the cost associated with remediating the area it contaminated. Since joint and several liability does not apply to Party B, it is not liable for any costs other than its own. No assessment for costs to other areas of the site. No orphan shares. Party B only pays for its portion of the site. Furthermore, because the site is divisible, the volumetric allocation not only does not apply, it provides the wrong answer. So does the toxicity-based allocation. In each case, the allocation approach attempts to allocate two-thirds of the site cost to Party B, when, in fact, Party B should be assigned only one-half of the site cost. Only the cost causation allocation provides the correct allocation result. These first two scenarios are premised on the assumption that each party dumps its waste in a uniquely identifiable area and there is no commingling of wastes. Unfortunately, Superfund sites rarely break out into uniquely identifiable parcels that can be assigned back to a specific party. As the next several scenarios will show, however, the basic premise of cost causation is still valid even in cases where wastes are commingled and sites are not generally considered divisible. Scenario 3—Commingling Creates a Benefit Party A dumps 5 drums of waste solvent onto a parcel of property; Party B simultaneously dumps 10 drums of waste solvent onto the same parcel of property. A third party, C, dumps an additional 5 drums of waste solvent onto the same parcel of property. Later, the site must be remediated as an NPL site with a single distinct area of contamination. Had either Party A or Party C been the only party at the site, their individual contributions would have caused the need to remediate 100 cubic yards of soil, which would have been sent to a RCRA landfill. Had Party B been the only party at the site, its individual contribution would have caused the need to remediate 200 cubic yards of soil, which again, would have been sent to a RCRA landfill. The actual remedy at the site calls for 300 cubic yards of soil to be excavated and taken to a RCRA landfill. According to the remediation contractor, the cost associated with this remediation is estimated at $ 80 per cubic yard plus a $ 12,000 setup fee for any volumes less than 50,000 cubic yards. There are no other costs to allocate between the three parties. Analysis This scenario adds a third party, commingles the wastes, and provides a formula for a remediation contractor to use to estimate the cost for remediating the site. The waste streams remain identical across the parties. The volumetric-based, toxicity-based, and cost causation allocations do not provide the same answer. Party A contributed 5 drums of waste solvent; Party B contributed 10 drums of waste solvent; and Party C contributed 5 drums of waste solvent. The total quantity of waste solvent at the site is 20 drums of waste. In a volumetric allocation, Party A would be assigned 25 percent, Party B would be assigned 50 percent, and Party C would be assigned the remaining 25 percent. This result is shown in Table 3.
Although the parties sent different quantities of material to the site, the type of material that each sent to the site is identical. Again, a toxicity-weighted volumetric allocation would produce identical results: 25 percent to Party A, 50 percent to Party B, and 25 percent to Party C. In this example, the entire cleanup cost is a common cost, because the costs are caused by more than one party. No direct or uniquely identifiable costs can be assigned to a particular party. Although each party is contributing the same type of material, the costs associated with remediating the waste are not directly volume variable, so the first step is to estimate the costs associated with addressing each party's waste, as well as the cost for remediating the entire site itself. Both Party A and Party C contributed 5 drums of waste solvent, and each disposal, on its own, would have required remediation of 100 cubic yards of contaminated soil. The cost to remediate that quantity of soil is $ 20,000. Party B, which sent twice as much waste, and contaminated twice as much soil, would need to spend $ 28,000 if it were the only party at the site. The site itself, which has 300 cubic yards of contaminated soil, will cost $ 36,000 to remediate, and this is the cost that is to be allocated among the three parties.48 Table 4 shows how to allocate this $ 36,000 site cost using the SAC estimates for each of the three parties. In this example, the total of the three SACs is $ 68,000—nearly twice the cost of the actual remediation. There are "economies of joint action" from the joint disposal and joint cleanup of this site because the cost of remediation, on an average cost per cubic yard basis, falls as the amount of material to be remediated rises.
In this example, economies of joint action result in a savings of $ 32,000. At the same time, each party is actually allocated a cost that is proportionately less than its SAC estimate. Because the premise of Superfund cost allocation is to equitably distribute the response costs, it is worth noting two results that flow from using SAC estimates to allocate these common costs. First, each party in this case is paying only a fraction of its SAC, but each party is paying the same fraction of its SAC. Second, there is a benefit from joint action. That benefit is distributed among the parties on the same basis as their relative liability. Both of these results can be seen in Table 5.
Scenario 4—Commingling Creates a Burden Party A dumps 5 drums of waste solvent onto a parcel of property; Party B simultaneously dumps 1 drum of PCB-laden oils onto the same parcel of property. Later, the area must be remediated as an NPL site with one distinct area of contamination. Party A's waste solvents would, on their own, result in 100 cubic yards of soil contaminated to a level at which it must be excavated and taken to a RCRA landfill at a cost of $ 20,000. Party B would have to incinerate its contaminated soil, but that volume of soil would be only 25 cubic yards if Party B were the only contributor. The cost of off-site incineration would be $ 25,000. Unfortunately, due to the interaction of the solvents with the PCBs, the PCBs are migrating throughout the 100 cubic yards of solvent-contaminated soil at a level sufficient to require incineration of the entire 100 cubic yards. The cost of the actual remedy is $ 200,000. There are no other costs to allocate between the two parties. In this scenario, the volumetric estimates, the toxicity relationship, and the cost causation allocation do not provide the same answer. A volumetric analysis, which implicitly assumes that the wastes have the same cost relationship, leads to an outcome where the small contributor of high-cost waste pays far less than its share when its cost contribution is taken into account. Neither party, on its own, is responsible for the entire problem at the site. The site cost is exacerbated by the interaction of the two parties' wastes. Party A contributed 5 drums of waste solvent; Party B contributed 1 drum of the PCBs and oils. On the waste-in list, Party A is assigned 83.3 percent of the volume since it contributed 5 of the 6 drums of waste contributed in total; Party B is assigned the remaining 16.7 percent of the remediation cost. If, alternatively, one uses weight as the basis (to account for the difference in density), the analysis varies slightly to adjust for the fact that the PCBs are heavier than most waste solvents. Whether the allocation is based on volume (gallons) or on weight, such an approach implicitly assumes that waste solvents and PCB-laden oils should be treated as equivalent in the allocation. This result is shown in Table 6.
Each party sends a different waste stream to the site. Arguably, the PCBs are "more hazardous" than the solvents (although this depends on the solvents), which might merit an increase in the share for the PCB contributor. In this example, we assume a toxicity factor of 8 is applied to the waste solvents, and a factor of 24 is applied to the PCBs. Implicitly, this higher factor means that we think the PCBs are three times as "bad" as waste solvents. While the volumetric analysis resulted in a 16.7 percent share being assigned to Party B, the toxicity-based allocation results in a 37.5 percent share being assigned to Party B. While this difference is substantial, the resulting share still does not reflect the relative cost caused by each of these parties.49 Toxicity adjustments, as shown above, attempt to compare different contaminants and to rank them, but they do not account for how these contaminants are remediated or how they contribute to response costs. Party A's dumping would prompt a response action—off-site landfilling—and would cost, in the absence of any other party, $ 20,000. Party B's dumping would prompt a response action—incineration—and would cost, in the absence of any other party, $ 25,000, even though its volumetric contribution was only a fraction of that contributed by Party A. The actual remedy, however, is $ 200,000, which is much more than the sum of the costs each party would pay on its own. Some have questioned the application of SAC analysis in cases where the interaction of parties leads to a burden or increased cost, rather than a savings or benefit to be shared.50 It is true that SAC analysis was historically used where the costs being allocated created a benefit and the SAC estimates for a given party served as a reservation price—a ceiling price. If the cost for a common project was above the ceiling price, the party would undertake the activity on its own. In the case where the joint action leads to a burden, not a benefit, the SAC continues to serve as a reservation price—as a floor to the allocation. Each party should be allocated at least what it would cause on its own. But obviously, no party would voluntarily join in a common action if that result led to a cost greater than what it would incur on its own. It would never voluntarily sign on for an activity with a burden. But in the Superfund context, parties are retroactively held responsible for their activities. No one weighed the cost of joint activity against the cost of other unique disposal options. And that is not really the problem that is sought to be addressed here. The question to answer is whether there is an equitable method for distributing a common burden among the parties. The principles of SAC provide an equitable method for such an allocation. SAC would allocate this cost as follows: Party A would pay [$ 20,000/($ 20,000 + $ 25,000)] 44.4 percent, and Party B would pay [$ 25,000/($ 20,000 + $ 25,000)] 55.6 percent. Those shares would apply to the actual remedy, whatever it ultimately costs. Here, Party A would pay approximately $ 88,800, while Party B would pay the remaining $ 111,200. In this example, a $ 155,000 "interaction" cost results from the commingling of these two waste streams. There is no unique way to distribute this cost between the two parties, but clearly these parties are somehow responsible for the incidence of this cost and it must be apportioned between them. The PCB generator might argue that if not for the solvents, the PCBs would not have migrated below the surface, thus the entire burden of this interaction should be borne by the solvent generator. The solvent generator can argue, however, that if not for the PCBs, its solvents would have contaminated the same soil, but that the remedy for such contamination would be a mere fraction of the actual remedy, thus the burden should be borne exclusively by the PCB generator. Here again, we see the "gaming" deficiency of using incremental cost. Scenario 5—Allocating Actions of Parties Party A transports 10 drums of solvents to a warehouse and abandons them, then sells the warehouse to Party B who does not know that the drums are in the warehouse. Party B, upon discovering the drums, does not contact Party A, but instead chooses to dump the contents of the drums onto the ground where they then contaminate 150 cubic yards of soil, that soil in turn requires remediation. Had Party A remediated those drums initially, it would have overpacked the drums and had them hauled away, at a cost of $ 5,000. Although Party B did not physically generate any waste, its actions resulted in the release of Party A's waste into the ground. The actual remedy, removal of 150 cubic yards of contaminated soil to a RCRA landfill, will cost $ 24,000. There are no other costs to allocate between the two parties. In this scenario, there is no way to perform the typical volumetric calculation. Volume is not appropriate here because Party B does not contribute any volume—it contributes action. To address this kind of situation, we must recognize that causation includes not only an examination of each party's wastes, but actions it takes on the wastes of others as well. This situation is frequently encountered at a site involving successive owners. One approach to a cost causation allocation would be to develop SAC estimates for each party. For Party A, the SAC would be based on the cost of overpacking and removing drums of solvent (assuming the drums would not have leaked if they were not emptied). In this example, that cost would be $ 5,000. For Party B, the SAC relies on the conclusion that Party B should be held responsible for its actions. Party B could have (a) ignored the drums and left them alone; or (b) contacted Party A and have it address the drums directly. By taking a course of action that creates the remediation problem, Party B creates its own cost. Here, the appropriate SAC for Party B is the cost of remediating the 150 cubic yards of soil, because it was the dumping of that waste that created that cost. This does not result in an incremental cost allocated entirely to Party B. Although Party B's SAC is identical to the actual remedy, Party A also has a SAC estimate, and this highlights one of the realities of codisposal at Superfund sites. The actions of individual parties can lead to costs that, when added together, greatly exceed the actual cost of remediation. Because multiple parties take similar actions there are economies of joint action. Once a specific area is completely saturated with solvents, pouring additional solvents onto the parcel has no additional impact (unless it changes the remedy or increases the duration or cost of treatment). Here, the allocation result is that Party A bears [$ 5,000/($ 5,000 + $ 20,000)] 20 percent of the allocated cost, while Party B bears 80 percent of the allocated cost. Scenario 6—The Industrial Facility Groundwater Issue Party A operates an industrial facility on a parcel of property for a period of 40 years. During the period of operation there are ongoing releases from the plant, which seep into the soil and into the groundwater. The type of contaminant is unimportant, but it is uniquely different from that used by Party B, who purchases the property and continues to operate the facility for another 10 years, during which there are ongoing releases that contaminate the same area of soil and groundwater (under the footprint of the plant). Its releases, however, are of different and, again, uniquely identifiable contaminant. The quantity of waste released by Party A is not quantified, but it is substantially larger than the quantity of contaminant released by Party B (also unquantified). The groundwater is remedied by the standard pump-and-treat technology; the level of each contaminant is sufficient to merit the same technology on its own. In this scenario, the volumetric estimates, the toxicity relationship, and the cost causation allocation do not provide the same answer. The focus is on groundwater for this example. Each party contributed to the common groundwater plume of contamination. The volume of releases from each party has not been specified in this scenario, and as is often the case at these types of sites, determining the "contribution" as measured in the standard volumetric sense is difficult. While this example has two unique contaminants—both of which contaminate the same area—would (or should) the allocation be different if the contaminants were the same across the parties? In a case where the amount of contribution is unknown, it is basically impossible to calculate a volumetric analysis. Toxicity Analysis Each party is sending a unique contaminant to the groundwater. Assume that Party A has sent a benzene toluene ethylbenzene xylene (or BTEX)-type contaminant, while the contaminant from Party B is TCE. A toxicity analysis would recognize that these are not the same contaminants and would rank BTEX versus TCE, leaving one party with a larger allocated share of the cleanup costs. On this measure alone, however, no account is given of the different volumes of material released by each party since volume is unknown. Each party's actions (their contributions over time) result in the same remedial action and, for purposes of this example, the same cost. One contaminant may disperse more than another, and there may be more of one in the plume relative to the other. But the remedy is the same for either, and the costs of cleanup are the same as well. Under a cost causation analysis both should share the cost of cleanup equally. There are several issues to consider. First, these are owner/operators, not generators, although they are acting as generators, and one owner/operator was at the site for a much longer period than the other. Both ran basically the same operation and their cost impacts can be determined easily. Why not allocate this on a "years-of-use" basis?51 Contrast this with a landfill-type site, where generators create the waste and owner/operators "handle" the waste created by others. Liability attaches to the owner/operators precisely because they handle the waste. If the site is a facility, however, there are often no generators delivering waste, only successive owners of the property whose operations lead to the need for remedial action. In that case, the owner/operators can be considered to be generators of waste. The imposition of an allocation share based simply on a years-of-use basis makes little sense. It is, at best, a crude attempt at a volumetric allocation (where you assume "all else being equal"). In the present example, where one party is at the site for 40 years and the other for only 10 years, the difference in years may have no relationship to the costof cleanup. In fact, as the example is specified, the cost of cleanup has no relationship to years of use. Likewise, volumetric allocations provide a distorted allocation. First, the example did not provide enough data on which to estimate volumes for each party—a common problem at facility-type operations and groundwater sites. Second, the volume of different contaminants does not allow them to be calibrated, unless they are adjusted for toxicity. Even so, toxicity factors often have no relationship to cost. For example, BTEX and TCE would probably be seen as different in a toxicity analysis. But from a cost of remediation viewpoint, both might lead to approximately the same level of cleanup costs. Is cost causation the "right" approach to use in this example? Although the owner/operator who was at the site for only 10 years might disagree, on a cost basis both parties share the liability for cleanup costs equally—and this is not just a viewpoint of allocators, it is the view of courts as well. This scenario is basically Ellman v. Woo, where the court applied principles of cost causation and allocated 50 percent of the cleanup costs to the defendant.52 The defendant in that case was responsible for one type of contaminant and assigned 50 percent. The plaintiff was either responsible for the other contaminant, or had the recourse to seek contribution from the responsible party for the other contaminant, and was assigned the other 50 percent. Although the volume of the contaminants was not determined, the court noted that each contaminant led to the same remedy and that the cost associated with cleaning up each contaminant was the same. Thus, the equitable allocation was to assign the defendant 50 percent of the cleanup costs.53 After more than a generation of Superfund cost allocation experience, there is still no clear approach to dividing cleanup costs among various PRPs. Because allocation is, by definition, a zero-sum game, parties rightly focus on site-specific and party-specific issues that should affect allocation. But, a consensus is still lacking on what the underlying methodology should be when allocations are developed. The allocation process is complicated by the fact that there is no statutory guidance on how to allocate these costs, and only a few cases that have developed allocations through the courts and arbitration proceedings that can then be used as a template to focus the allocation process. Although many out-of-court settlements do occur, their solutions are frequently made to avoid the substantial transaction costs associated with taking a case to court and seeking the court's view on allocation. What PRPs need is guidance on an approach to allocating costs and clarity on key issues that affect cost allocation. It has taken nearly a generation to move from volume as the default, to volume as an option. It has taken just as long to move from Gore factors as the equitable factors, to Gore factors as examples of equitable factors. Parties still seek clarity on whether they are allocating "harm" or whether, as the statute says, they are allocating "response costs." PRPs need to recognize that cost allocation is just that—cost allocation and not allocation of pounds or gallons of waste. Cost causation is a fundamentally sound method for apportioning these Superfund costs. While cost causation-based allocation may not answer every equitable issue, it often does address most of them. Cost causation is a good starting point for developing equitable allocations from which other adjustments, if appropriate, can be made; adjustments to account for site-specific equitable factors not adequately addressed in the causation-based allocation. [28 ELR 10082] These adjustments tend to be minor because cost causation allocation addresses most of the major equity issues.54 While PRPs and EPA itself are turning toward cost causation, few costs are so uniquely caused that they can be tied to the actions of a single party. Many of these costs, however, can be tied to the activities of a class or subset of parties. And as EPA itself indicates, those parties should be responsible for those categories of cost. Assigning cost elements to groups of PRPs who collectively are responsible for them is one step in the process. The more contentious step is allocating those costs among the parties within that group or class. Here again, cost causation has a role to play in the allocation of these common or joint costs. Although there is no unique way to allocate common costs, some methods are more equitable than others. This Article highlights the use of cost causation concepts and SAC analysis as methods for apportioning common costs. The focus of this Article has not simply been to advocate cost causation, but to highlight the deficiencies in other allocation methods which fail to account for how remediation costs are caused in dividing costs among responsible parties. 1. CERCLA or Superfund refers to the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), 42 U.S.C. §§ 9601-9662, ELR STAT. CERCLA §§ 101-312. In 1986, Congress reauthorized and amended CERCLA by the Superfund Amendments and Reauthorization Act of 1986 (SARA), Pub. L. No. 99-499, 100 Stat. 1613. 2. The U.S. Environmental Protection Agency (EPA) generally discusses the broad range of allocation issues in terms of volume. And although "gallons" is a volumetric measure and "pounds" is not, EPA acknowledges that weight is an appropriate measure to use when the waste types contributed by the various parties are heterogeneous. Weight is often a measure at landfill sites, which, due to their codisposal nature, receive a variety of waste types of differing densities. Many settlements (e.g., de minimis settlements) indicate that weight, not volume, is the measure. In short, when allocators talk about "volume" as the common measure, they usually mean "amount of waste," which, depending on waste type, may be more appropriately compared as volume or as weight. It seems equally clear that when EPA discusses allocation issues in terms of "volume," they also mean "amount of waste" as either volume or as weight. See GUIDANCE ON PREPARING WASTE-IN LISTS AND VOLUMETRIC RANKINGS FOR RELEASE TO POTENTIALLY RESPONSIBLE PARTIES (PRPs) UNDER CERCLA, OSWER Directive No. 9835.16 (Feb. 20, 1991). A waste-in list is a listing of the volume and nature of substances contributed by each PRP at a facility (OSWER 9835.16 at 3). [In developing a waste-in list] waste-in information should be converted to a common unit of measurement. In general, most sites will be receiving hazardous substances in drums or tankers, making gallons the preferable unit in which to express volume. However, some sites, such as landfills, may have large amounts of solid waste, trash, and other hazardous substances coming in by weight, in which case pounds or tons may be more appropriate. (OSWER 9835.16 at 5-6). For consistency, "volume" will be discussed, while also recognizing that in many instances it is actually weight, but in this case both encompass the "amount of waste." 3. The courts generally have recognized that using a pure volumetric allocation does not meet the fundamental "equity" criteria set up by CERCLA because different wastes have different toxicities and different response costs. Consequently, "toxicity" is often considered a relevant factor in addition to volume. See Ridgeway W. Hall et al., Superfund Response Cost Allocations: The Law, The Science and The Practice, 49 BUS. LAW. 1512-13 (1994). Even CERCLA has recognized this. See Superfund Program: De Minimis Contributor Settlements, 52 Fed. Reg. 24338 (June 30, 1987). 4. See 42 U.S.C. § 9607(a), ELR STAT. CERCLA § 107(a). 5. "One of CERCLA's primary goals is to extend liability to all those involved in creating harmful environmental conditions." Schiavone v. Pearce, 79 F.3d 248, 253, 26 ELR 20824, 20827 (2d Cir. 1996). 6. See, e.g., B.F. Goodrich v. Betkoski, 99 F.3d 505, 515, 27 ELR 20329, 20331 (2d Cir. 1996). The court in Betkoski notes that in CERCLA § 101(14), "Hazardous substance" is expansively defined to include: (A) any substance designated pursuant to section 1321(b)(2)(A) of Title 33, (B) any element, compound, mixture, solution, or substance designated pursuant to section 9602 of this title, (C) any hazardous waste having the characteristics identified under or listed pursuant to section 3001 of the Solid Waste Disposal Act …, (D) any toxic pollutant listed under section 1317(a) of Title 33, (E) any hazardous air pollutant listed under section 112 of the Clean Air Act …, and (F) any imminently hazardous chemical substance or mixture with respect to which the [EPA] Administrator has taken action pursuant to section 2606 of Title 15. Id. See also Table 302.4, 40 C.F.R. § 302.4 (1995) for a lengthy listing of hazardous substances. 7. Hazardous substances are defined in CERCLA § 101(14). 42 U.S.C. § 9601(14), ELR STAT. CERCLA § 101(14). There is no threshold requirement for hazardous substances to establish liability. "Twice we have said that quantity 'is not a factor' when determining CERCLA liability because had Congress wanted to distinguish liability on the basis of quantity, it would have so provided." Betkoski, 99 F.3d at 517, 27 ELR at 20329. See also United States v. Alcan Aluminum Corp., 990 F.2d 711, 720, 23 ELR 20706, 20710 (2d Cir. 1993) ("The statute on its face applies to 'any' hazardous substance, and it does not impose quantitative requirements."). The absence of threshold quantity requirements in CERCLA leads logically to the conclusion that the Act's "hazardous substance" definition includes even minimal amounts. "A disposal of any detectable hazardous substance into the environment, regardless of amount, constitutes a release." Dartron Corp. v. Uniroyal Chem. Co., 917 F. Supp. 1173, 1183, 26 ELR 21056, 21060 (N.D. Ohio 1996) (citing HRW Sys., Inc. v. Washington Gas Light Co., 823 F. Supp. 318, 23 ELR 21586 (D. Md. 1993)). 8. See, e.g., B.F. Goodrich Co. v. Murtha, 754 F. Supp. 960, 965-66, 21 ELR 20777, 20779 (D. Conn. 1991), aff'd, 958 F.2d 1192, 22 ELR 20683 (2d Cir. 1992) ("If the MSW, including household waste, contained a 'hazardous substance' as defined in § 9601(14) and there is a release or threatened release of the substance at the sites, the municipalities are liable under CERCLA regardless of the origination of the substance."). 9. See Dartron Corp., 917 F. Supp. at 1184, 26 ELR at 21060-61, for its discussion regarding a party who allegedly disposed of waste oil. EPA presumes to be hazardous wastes from the interior of a tank that held petroleum product. See United States v. Western Processing Co., 761 F. Supp. 713, 720-22, 21 ELR 20976, 20979 (W.D. Wash. 1991); 40 C.F.R. § 279.10 (1996) (detailing a part of this presumption). The only way to rebut this presumption is to test the waste oil and show that it does not contain contaminants. This presumption makes perfect sense, as it furthers the goals of CERCLA to achieve prompt cleanup of hazardous waste sites and to impose the cost of cleanup on those responsible for the contamination. If a person dumped waste oil onto a site that is later found to be contaminated with hazardous substances, that person cannot escape liability merely by arguing that there is no evidence the dumped waste oil actually contained a hazardous substance. Allowing such an argument would protect not only a simply negligent dumper of waste, but also a person who willfully cast a blind eye to the possibility that the waste he was dumping contained dangerous contaminants. 10. See, e.g., United States v. Colorado & Eastern R.R., 50 F.3d 1530, 1534, 25 ELR 20309, 20311 (10th Cir. 1995). "Causation" in the context of this case means whether CERC caused the increase in the expense of remediation. If it did, that would be an equitable factor for the court to consider under § 113(f)(1). All concerned acknowledge that liability under § 9607 is strict and causation would be irrelevant in that context. Id; see also Farmland Indus., Inc. v. Colorado & Eastern R.R., 922 F. Supp. 437, 441 (D. Colo. 1996). Causation in the context of a § 9607(a) action is irrelevant to liability for response costs because the statute sets forth the categories of persons that are liable for response costs and imposes joint and several liability. A liable party under § 9607(a) is limited to the statutory defenses in § 9607(b). If a liable party is unable to establish a § 9607(b) defense, he becomes potentially liable for all response costs regardless of his proportionate fault. Thus, the statute imposes harsh results on parties who may have minimum or de minimis responsibility for the contamination but who the government proceeds against to facilitate clean-up in the most expedient manner. Id. The court also noted that "at this [the allocation] juncture, causation may, if appropriate, become a relevant factor in the contribution equation. One who caused the contamination is likely to be accountable for some of the response costs but, as discussed, one who did not cause the contamination is not necessarily without culpability." Id. 11. See FMC Corp. v. Aero Indus., Inc., 998 F.2d 842, 23 ELR 21312 (10th Cir. 1993). The Tenth Circuit held that for a plaintiff to establish a prima facie case of liability in a CERCLA cost recovery claim, it must show that "(1) the site is a facility as defined in 42 U.S.C. § 9601(9), (2) defendant is a responsible person defined in 42 U.S.C. § 9607(a), (3) the release or threatened release has occurred, and (4) the release or threatened release has caused the plaintiff to incur response costs." Id. at 845, 23 ELR at 21313. As the Farmland court noted after reviewing the FMC finding, "regardless of whether the claim is brought under § 9607 or § 9613 the liability determination is based on the same elements. Causation is not part of the liability inquiry." 922 F. Supp. at 441. 12. "Congress enacted CERCLA with the expansive, remedial purpose of ensuring 'that those responsible for any damage, environmental harm, or injury from chemical poisons bear the costs of their actions.'" Schiavone v. Pearce, 79 F.3d 248, 253, 26 ELR 20824, 20827 (2d Cir. 1996) (quoting S. REP. NO. 96-848, at 13 (1980)). 13. Liability under CERCLA is strict. See New York v. Shore Realty Corp., 759 F.2d 1032, 15 ELR 20358 (2d Cir. 1985); see also St. Paul Fire & Marine Ins.Co. v. Warwick Dyeing Corp., 26 F.3d 1195 (1st Cir. 1994). Under CERCLA, a person that generates hazardous substances and arranges for their disposal is strictly liable, regardless of whether the person was at fault or whether the substance actually caused or contributed to any damage, for all costs of remediating environmental damages at the site where the substances ultimately are disposed. Id. at 1197-98. (Emphasis in original.) 14. Liability under CERCLA is joint and several if the environmental harm is indivisible. See B.F. Goodrich Co. v. Murtha, 958 F.2d 1192, 1198, 22 ELR 20683, 20684 (2d Cir. 1992); Roberts v. Picillo, 883 F.2d 176, 178-79, 20 ELR 20115, 20116 (1st Cir. 1989), cert. denied sub nom. American Cyanamid Co. v. O'Neil, 493 U.S. 1071 (1990); United States v. R.W. Meyer, Inc., 889 F.2d 1497, 1506-08, 20 ELR 20319, 20324-25 (6th Cir. 1989), cert. denied, 494 U.S. 1057 (1990); see also Farmland Indus., 922 F. Supp. at 441. In deciding whether liability is joint and several the Court is guided by the test set forth in the Restatement (Second) of Torts § 433A (1965) which provides: (1) Damages for harm are to be apportioned among two or more causes where (a) there are distinct harms, or (b) there is a reasonable basis for determining the contribution of each cause to a single harm. (2) Damages for any other harm cannot be apportioned among two or more causes. United States v. Monsanto Co., 858 F.2d 160, 172, 19 ELR 20085, 20089 (4th Cir. 1988). 15. See, e.g., United States v. Olin Corp., 107 F.3d 1506, 27 ELR 20778 (11th Cir. 1997). 16. See National Priorities List for Uncontrolled Hazardous Waste Sites, 60 Fed. Reg. 20334 (Apr. 25, 1995). This appears to be the most current EPA estimate of the average NPL site cost. It is estimated in 1994 dollars and totals $ 30.74 million. It includes $ 22.5 million for the remedial action, $ 1.35 million for the remedial investigation/feasibility study, $ 1.26 million for remedial design, and $ 5.63 million in discounted operation and maintenance costs ($ 400,000 per year for 30 years, discounted at 5.8 percent). 17. See National Priorities List for Uncontrolled Hazardous Waste Sites, 62 Fed. Reg. 50445 (Sept. 25, 1997). With the new sites added in today's rule, the NPL now contains 1,204 sites, 1,053 in the General Superfund Section and 151 in the Federal Facilities Section. With a proposed NPL rule published elsewhere in today's Federal Register, there are now 52 sites proposed and awaiting final agency action, 46 in the General Superfund Section and 6 in the Federal Facilities Section. Final and proposed sites now total 1,256. 18. 42 U.S.C. § 9613(f)(1), ELR STAT. CERCLA § 113(f)(1). 19. See Amoco Oil Co. v. Borden, Inc., 889 F.2d 664, 20 ELR 20281 (5th Cir. 1989). A number of cases have applied these factors, including United States v. R.W. Meyer, Inc., 932 F.2d 568, 575-76, 21 ELR 21062, 21063-64 (6th Cir. 1991); United States v. Tyson, No. 84-2663, 1989 WL 159256, 17 ELR 20527 (E.D. Pa. 1986); Amoco Oil Co. v. Dingwell, 690 F. Supp. 78, 86 (D. Me. 1988), aff'd sub nom. Travelers Indem. Co. v. Dingwell, 884 F.2d 629 (1st Cir. 1989); Colorado v. ASARCO, Inc., 608 F. Supp. 1484, 1487, 15 ELR 20523, 20525 (D. Colo. 1985); United States v. A & F Materials Co., 578 F. Supp. 1249, 1256, 14 ELR 20105, 20108 (S.D. Ill. 1984). For a discussion of these factors, see John C. Butler III et al., Allocating Superfund Costs: Cleaning Up the Controversy, 23 ELR 10133 (Mar. 1993). See, e.g., Control Data Corp. v. S.C.S.C. Corp., 53 F.3d 930, 935, 25 ELR 21378, 21380 (8th Cir. 1995) (citing John Copeland Nagle, CERCLA, Causation, and Responsibility, 78 MINN. L. REV. 1493, 1522-23 n.133 (1994)). A primary focus on these [Gore] factors is the harm that each party causes the environment. Those parties who can show that their contribution to the harm is relatively small in terms of amount of waste, toxicity of the waste, involvement with the waste, and care, stand better position to be allocated a smaller portion of the response costs. 20. Some courts have expanded this to include "the degree to which each party made efforts to remediate the release of hazardous substances at the Property after the time the releases occurred, including their cooperation with governmental authorities." Dartron Corp. v. Uniroyal Chem. Co., 917 F. Supp. 1173, 1185, 26 ELR 21056, 21061 (N.D. Ohio 1996). 21. Environmental Transp. Sys., Inc. v. ENSCO, Inc., 969 F.2d 503, 509, 22 ELR 21361, 21363 (7th Cir. 1992) ("the Gore Factors are neither an exhaustive nor exclusive list"). Certain Gore factors are tied more closely to specific classes of parties. For example, factors four through six frequently are cited as applying most closely to the owner/operator class. See, e.g., Farmland Indus., Inc. v. Colorado & Eastern R.R., 944 F. Supp. 1492, 1499-1500 (D. Colo. 1996). 22. "A court may consider several factors, a few factors, or only one determining factor … depending on the totality of the circumstances presented to the court." Environmental Transp. Sys., Inc., 969 F.2d 503, 509, 22 ELR 21361, 21363. Courts have broad discretion in the use of equitable factors. FMC Corp. v. Aero Indus., Inc., 998 F.2d 842, 846, 23 ELR 21312, 21314 (10th Cir. 1993). 23. See, e.g., B.F. Goodrich Co. v. Murtha, 958 F.2d 1192, 1206, 22 ELR 20683, 20689 (2d Cir. 1992). 24. See, e.g., Weyerhaeuser Co. v. Koppers Co., 771 F. Supp. 1420, 22 ELR 20168 (D. Md. 1991). 25. See id. 26. See, e.g., Kerr-McGee Chem. Corp. v. Lefton Iron & Metal Co., 14 F.3d 321, 327, 24 ELR 20369, 20370 (7th Cir. 1994). 27. See, e.g., United States v. Stringfellow, No. CV 83-2501 JMI, 1993 U.S. Dist. LEXIS 19113, at *302 (C.D. Cal. 1993). ("The primary equitable factors that the Special Master considered are the interrelated concepts of relative fault and causation."). See also B.F. Goodrich Co., 958 F.2d at 1206, 22 ELR at 20689 (noting that the relative cleanup costs incurred as a result of disposal of MSW may be an equitable factor). 28. See e.g., Dartron Corp. v. Uniroyal Chem. Co., 917 F. Supp. 1173, 1185, 26 ELR 21056, 21061 (N.D. Ohio 1996). See also Farmland Indus., Inc. v. Colorado & Eastern R.R., 944 F. Supp. 1492, 1499-1500 (D. Colo. 1996). The court noted CERC's "duty as a property owner" and "the degree of care exercised by a party in managing a Superfund site" as equitable factors. Id. 29. See Farmland Indus., 944 F. Supp. at 1499-1500. 30. See, e.g., Louisiana-Pacific Corp. v. Asarco, Inc., 21 Chem. Waste Litig. Rep. 1165, 1168 (W.D. Wash. 1991). "Another equitable factor I think applies here, and I think is very important, is an analysis of the economic benefit or loss caused by a pollution or contamination situation." Id. See the discussion by Dr. George R. Hall on the use of economic benefits as a basis for the allocation and the court's reliance of this approach in Gould, Inc. v. A & M Battery & Tire Service, No. 3:CV-91-1714, 1997 U.S. Dist. LEXIS 15728 (M.D. Pa. 1997). 31. See, e.g., United States v. R.W. Meyer, Inc., 932 F.2d 568, 572, 21 ELR 21062, 21063 (6th Cir. 1991). See also Farmland Indus., 944 F. Supp. at 1500. Perhaps the most stark contrast between Farmland and the CERC Parties is the difference in mindset apparent during remediation. Farmland seemed eminently interested in cleaning up the Site while the CERC Parties seemed entirely engrossed with how to avoid any responsibility for the cleanup. For example, rather than granting immediate access to the property to erect a fence, Flanders stalled, apparently in an attempt to make a profit off of Farmland's need to fence the property. In addition, there is strong evidence to suggest that Flanders did everything he could to divert all of CERC's assets to other entities in an effort to avoid paying any remedial costs. 32. See Weyerhaeuser Co. v. Koppers Co., 771 F. Supp. 1420, 22 ELR 20168 (D. Md. 1991). 33. This argument was also raised by EPA in United States v. Nalco Chem. Co., where EPA argued that its assigned percentage to a party was fair in light of the fact that "present owners of contaminated property reap the benefit of any increase in the value of their property due to cleanup by the U.S. EPA." United States v. Nalco Chem. Co., 1996 U.S. Dist. LEXIS 13089 at n.13 (N.D. Ill. 1996); see also Farmland Indus., 944 F. Supp. at 1500. The EPA published its notice of intent to delete the Site from the [NPL], and it was deleted. The parties stipulated that the value of the CERC Parcels increased by more than $ 600,000 as a result of the cleanup. Given that Farmland garnered no tangible benefit from the cleanup of land it no longer owns, it would be inequitable not to allocate significant additional remedial costs to the CERC Parties. 34. See 52 Fed. Reg. 19919-20 (May 28, 1987). 35. See B. DOUGLAS BERNHEIM ET AL., COMMENTS OF AIR PRODUCTS AND CHEMICALS, INC.; LUCENT TECHNOLOGIES; MACK TRUCKS, INC.; PENNSYLVANIA POWER & LIGHT COMPANY; RAYTHEON APPLIANCES, INC.; TARKETT, INC.; GAF CORPORATION; GENERAL ELECTRIC COMPANY; AND INGERSOLL-RAND CONCERNING THE UNITED STATES ENVIRONMENTAL PROTECTION AGENCY'S MUNICIPAL SOLID WASTE SETTLEMENT PROPOSAL 6 (Aug. 1997). 36. See EPA Proposal for Municipal and MSW Liability Relief at CERCLA Co-Disposal Sites, 62 Fed. Reg. 37321 (July 11, 1997). 37. Id. at 37233. 38. Id. 39. For a more thorough discussion of the MSW settlement proposal and why EPA does not have the authority to address issues of allocation, see Richard Lane White, Unequal Superfund Treatment: CERCLA Dupli-City, ENVTL. CORP. COUNS. REP., Oct. 1997, at 1 [hereinafter Unequal Superfund Treatment]; see also Richard Lane White, EPA's New Municipal Liability Proposal Sidesteps Equitable Allocation by Courts, 12 TOXICS L. REP. 312 (1997). 40. United States v. Cannons Eng'g Corp., 899 F.2d 79, 87, 20 ELR 20845, 20848 (1st Cir. 1990). 41. See 42 U.S.C. § 9613(f)(1), ELR STAT. CERCLA § 113(f)(1). See also Control Data Corp. v. S.C.S.C. Corp., 53 F.3d 930, 936, 25 ELR 21378, 21380 (8th Cir. 1995). Under CERCLA, if a responsible party … releases hazardous materials into the environment, and that release "causes the incurrence of response costs," then the party is liable. 42 U.S.C. § 9607(a). The question then becomes, liable for what? CERCLA's answer is that the party is liable for "any other necessary cost of response incurred by any other person consistent with the national contingency plan." 42 U.S.C. § 9607(a)(4)(B). Thus, a plain reading of the statute leads us to the conclusion that once a party is liable, it is liable for its share, as determined by Section 9613(f), of "any" and all response costs …. 42. See, e.g., Dartron Corp. v. Uniroyal Chem. Co., 893 F. Supp. 730, 737, 26 ELR 20115, 20119 (N.D. Ohio 1995). Having reached this conclusion, [denying summary judgment on particular grounds] the Court adds an important observation. Given the current state of the record, it appears that summary judgment in Dartron's favor on Uniroyal's CERCLA counterclaim may be appropriate on another ground. "Congress intended that those responsible for problems caused by the disposal of chemical poisons bear the costs and responsibility for remedying the harmful conditions they created." United States v. Akzo Coatings of America, Inc., 949 F.2d 1409, 1418, 22 ELR 20405, 20407 (6th Cir. 1991) (quoting United States v. Reilly Tar & Chem. Corp., 546 F. Supp. 1100, 1112, 12 ELR 20954, 20957 (D. Minn. 1982)). "The cost of cleaning up hazardous waste sites is to be borne by those responsible for the wastes." United States v. Bogas, 920 F.2d 363, 369, 21 ELR 20356, 20359 (6th Cir. 1990); see also 42 U.S.C. § 9613(f)(1), ELR STAT. CERCLA § 113(f)(1) (if more than one party is responsible for the release of hazardous substances, "the court may allocate response costs among liable parties using such equitable factors as the court determines are appropriate."). 43. A typical engineering "rule of thumb" is C[2] = C[1](Q[2]/Q[1]) 44. See 52 Fed. Reg. 19919, 19920 (May 28, 1987). 45. See, e.g., United States v. Colorado & Eastern R.R., 50 F.3d 1530, 1535, 25 ELR 20309, 20311 (10th Cir. 1995). Due to the impossibility of determining the amount of environmental harm caused by each party where wastes of varying and unknown degrees of toxicity and migratory potential have mixed, the courts have been reluctant to apportion costs between PRPs, and hence have adopted the rule that "damages should be apportioned only if the defendant can demonstrate that the harm is divisible." Id. (quoting Roberts v. Picillo, 883 F.2d 176, 178, 20 ELR 20115, 20116 (1st. Cir. 1989)). "Where defendants bear the burden of proving divisibility, responsible parties rarely escape joint and several liability." Id. 46. See supra note 14. 47. That is why it is so surprising that EPA has so clearly advocated an approach to addressing MSW allocation, which threatens to defeat joint and several liability. See Unequal Superfund Treatment, supra note 39, at 3. 48. For Party A and Party C, the cost is calculated as $ 12,000 + (100) x ($ 80) = $ 20,000. For Party B the cost is calculated as $ 12,000 + (200) x ($ 80) = $ 28,000. The cost to remediate the site as it now exists is calculated as $ 12,000 + (300) x ($ 80) = $ 36,000. 49. Also, while the tripling of the toxicity factor had a substantial impact on the share assignable to Party B (it raised it from 16.7 percent to 37.5 percent), the allocation becomes less sensitive as the PCB factor rises. For example, if the factor were raised to 32, the share assignable to Party B would rise to 44.4 percent. 50. See Kenneth T. Wise et al., Economists, Orphans, and Superfund Allocation: Clearing Up the Controversy, 34 Chemical Waste Litig. Rep. 763-66 (1997). 51. "The relative allocation among successive owners and/or operators may be determined, where all other circumstances are equal, by the relative length of time each owned and/or operated the site." Interim Guidelines for Preparing Nonbinding Preliminary Allocations of Responsibility, 52 Fed. Reg. 19919, 19929 (May 28, 1987). 52. See Ellman v. Woo, 34 ERC 1969, 22 ELR 20875 (E.D. Pa. 1991). 53. See id. at 1971-72. Here, the court recognized that there were two distinct contaminants and that each would require the same remedy on their own. The defendant was held responsible for the solvent contamination and was assigned 50 percent of the liability. The plaintiff was assigned 50 percent of the liability for the petroleum contamination, which was not caused by the defendant. 54. It is important to note that a cost causation allocation will not assure agreement of all PRPs since parties often advance self-interested allocation positions. However, we believe that a causation allocation is fair and logical and one most likely to be accepted by a trier of fact. 28 ELR 10067 | Environmental Law Reporter | copyright © 1998 | All rights reserved |