Land Use and Cleanups: Beyond the Rhetoric

George Wyeth

Mr. Wyeth is an attorney in the Office of General Counsel at the U.S. Environmental Protection Agency (EPA). The views expressed are the author's and do not necessarily represent those of EPA. Discussion of pending legislation is based on language in the bills as of May 24, 1996.

[26 ELR 10358]

There seems to be agreement across a wide spectrum of those involved in Superfund cleanups that such cleanups should take into consideration the kinds of activities that are expected to take place at the site after the remedial work is completed. While cleaning every site to levels suitable for all conceivable uses may be a laudable goal, doing so can impose costs that are out of proportion to the added amount of protection obtained. Not surprisingly, therefore, every significant proposal for reform of the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA)¹ contains a provision tying the selection of remedies to expected land uses in some way.

The role of land use in remedy selection, however, is more complicated than many appreciate. Discussions of land-use based remedies frequently reflect oversimplified notions of how land-use considerations may bear on the remedy selection process and how land-use decisions are really made. This Dialogue seeks to provide a clearer understanding ofhow land use should be considered in remedy selection and addresses the long-term implications of tying remedies to anticipated land uses. In addition to highlighting the issues that are often overlooked in discussions of land use as a cleanup consideration, it argues that we should consider an entirely different kind of cleanup regime similar to permitting, under which regulatory oversight would be maintained on a continuing basis and cleanups phased in with an emphasis on addressing known short-term risks.

Introduction to Land-Use Based Remedies

The fundamental insight underlying the general enthusiasm for land-use based remedies is fairly simple. In the past, cleanup programs such as Superfund have often sought to clean up sites to levels that will be safe for unrestricted human use. In addition to the obvious policy reasons for wanting to achieve this goal, congressional calls for the maximum use of permanent remedies (embodied in the 1986 amendments to CERCLA) put pressure on government agencies to impose highly stringent cleanup standards.

Achieving unrestricted use at every site, however, can be very costly. There is broad consensus that one way to reduce costs without significantly increasing risks is to design cleanups around the land uses anticipated at the site.² For example, some land uses, such as commercial or industrial activities, are compatible with relatively high levels of on-site contamination because they result in less exposure than residential use. Therefore, less stringent cleanups may be satisfactory at sites where those uses are expected.

The Superfund program has, in fact, never required cleanups to residential standards in every case. The preamble to the national contingency plan (NCP), which provides the regulatory guidelines for CERCLA cleanups, indicates that risk assessments used in designing cleanups need not always be based on the assumption that the site will be used for residential purposes. It states:

In general, the baseline risk assessment will look at a future land use that is both reasonable, from land use development patterns, and may be associated with the highest (most significant) risk, in order to be protective. These considerations will lead to the assumption of residential use as the future land use in many cases…. The assumption of residential land use is not a requirement of the program but rather is an assumption that may be made, based on conservative but realistic exposures, …. An assumption of future residential land use may not be justifiable if the probability that the site will support residential use in the future is small.³

However, although this policy does not specify that residential use must be assumed in every risk assessment, it does arguably create a presumption in favor of making such assumptions. The approach recommended in the NCP might, for example, treat an existing industrial area as potentially residential in the future based on observable shifts in use that have occurred in similar areas over the past 25 years.⁴

[26 ELR 10359]

In 1995, the U.S. Environmental Protection Agency (EPA) responded to those who contended that the Superfund program placed too much emphasis on residential level cleanups by issuing guidance that states: "remedial action objectives developed during the [remedial investigation/feasibility study] should reflect the reasonably anticipated future land use or uses."⁵ The guidance also states that "the baseline risk assessment generally needs only to consider the reasonably anticipated future land use."⁶ This guidance document has not ended the discussion, however; interest remains strong in codifying a commitment to greater consideration of land use. All the major CERCLA reauthorization bills contain provisions to this end.⁷

Issues in Selecting Land-Use Based Remedies

Despite the amount of attention given to land-use based remedies, however, most of the discussion has tended to oversimplify the issue, obscuring important policy choices and possibly leading to unduly simplistic solutions. The following sections clarify what the selection of land-use based remedies entails, with the aim of untangling the various issues and misconceptions with which decisionmakers must wrestle.

Anticipating Versus Controlling Land Use

Discussions of land-use based remedies tend to blur the distinction between allowing less stringent cleanup levels at sites where the only likely future uses are compatible with those levels, and controlling future use as a risk management device (usually, as an alternative to more permanent forms of cleanup such as containment or treatment). Few would disagree with the proposition that a cleanup of a site that is currently industrial and is expected to remain industrial indefinitely should aim only for levels that are protective of that industrial use. Requiring more stringent cleanup could result in substantial costs to address relatively speculative risks.

When future uses are less certain, however, a decision to protect only current or immediately foreseeable uses is implicitly a choice to use controls on future land use as a tool for managing risk. Such a choice involves significant trade offs between cost and long-term reliability that are not raised by the easy case presented by a site that is expected to be used for industrial purposes for the indefinite future. It is because of these trade offs that using land-use controls as remedies has been controversial throughout the history of the Superfund program.

Broad agreement on considering future land use in selecting remedies should not, therefore, be equated with endorsing greater use of land-use controls as an alternative to more permanent cleanups. Yet that is the implicit result if one moves from an approach of protecting all reasonably anticipated land use to one of protecting only current or very likely future land uses (as at least two Superfund reforms propose).⁸ In the latter case, plausible future uses will have to be foregone and controls imposed to ensure that result.

This Dialogue does not argue for or against using land-use controls more widely as remedial tools; that is a fundamental policy decision beyond the scope of this Dialogue. The critical point is, however, that the decision should not be made invisibly under the guise of a seemingly uncontroversial statutory amendment regarding "land use."

The Role of Land Use in Remedy Selection

The relationship between expectations about land uses and the design of cleanups is more complicated than is generally recognized. Proposals to base remedies on land use often overlook the distinction between predictions about future land use, as part of the risk assessment process, deciding what land use the remedy should actually be designed to allow, which is a matter of risk management. They often treat land use as something that is "selected" before remedy selection based on a combination of predictions about likely use and preferences expressed by some body of community representatives, with the remedy then designed to provide for this selected use.

Risk assessment, an early and critical step in the cleanup process, is the largely technical matter of predicting the likelihood of harm occurring (e.g., cancer incidence) as a result of exposure to contaminants at a site. In this connection, predictions about land use are of interest because different land uses result in differing degrees of exposure. As a result, it is important to evaluate as best we can what the land use at the site is likely to be in order to know how risky it is.

On the other hand, the selection of a remedy, including the determination of what land uses the remedy will protect, is not a predictive matter but a policy choice requiring various considerations to be balanced against each other. Such discussions depend on a wide variety of factors including, but by no means limited to, the cost of achieving any particular level of cleanup, the degree of uncertainty about future uses, the feasibility of controlling future uses, and the economic and social value of making the site suitable for a wide range of potential uses. Certainly predictions about future land use are relevant to this decision — for example, in assessing the threat posed by a site and evaluating the likelihood of changes in uses, which affect the reliability of remedies leaving high levels of contamination onsite. Community preferences also play a role, although in a different way — they shed light on the potential value of alternative cleanup levels.

In some cases the predicted or preferred use may also be the one forwhich the cleanup should be designed, but in others the balancing of all these factors will lead to a different conclusion. It may be decided that the cleanup should permit either a wider or narrower range of potential uses. For example, the cost or incidental risks of achieving alternative levels of residual contamination, or the presence of ecological impacts, might make it appropriate to select [26 ELR 10360] cleanup levels that are either more or less stringent than would be suitable for the current or anticipated use. In areas where land uses are likely to change, the burden of maintaining controls may outweigh the advantages of a less stringent cleanup. In such cases, the added cost of achieving a residential level cleanup may be low enough, and the risk that a land use may change enough, to warrant the more extensive cleanup, even if the current or immediately foreseeable use only demands less stringent cleanup. In other cases, even protecting reasonably foreseeable uses may not be cost-effective, and the only reasonable approach from a cost perspective is simply to dedicate some of the land to very restricted uses. This might be true, for example, when part of a site is devoted to a waste management unit. Restricted use of, and possibly even access to, a site will also likely be part of the remedy at sites where it could be more hazardous or unreasonably costly to move large quantities of contaminated soil elsewhere for treatment and disposal than to leave them on site.

Therefore, predicting land uses or even assessing the community's preference as to future land use does not answer the question of what kinds of what land uses the remedy should be designed to allow. Such predictions are necessary to the remedy selection decision, but they are not, as some proposals seem to assume, sufficient to determine what the cleanup levels for the site should be. Instead, because remedial decisions involve such a complicated weighing of factors, any statutory directives on land use should allow decisionmakers flexibility while ensuring that land use is appropriately considered. For example, mandating that all cleanups protect current and likely future uses — no more and no less — will lead to cleanups that are too stringent in some cases and not as stringent as they should be in others.

Some of the currently pending Superfund reform legislation risks would impose such a uniform standard. For example, H.R. 2500, a bill introduced by Rep. Michael G. Oxley (R-Ohio), states: "Appropriate remedies shall be based on the current and reasonably anticipated future uses of land, water and other resources."⁹ The bill also states that "reasonably anticipated uses may include current uses or potential future uses that have a substantial probability of occurring."¹⁰ If stating that remedies "shall be based on" current uses or those with a "substantial probability of occurring" means that the uses will be determined first, and remedies shall be designed to protect those uses and only those uses in every case, erroneous remedy selection decisions will result in at least some cases for the reasons just discussed.¹¹

Similarly, S. 1285, the Superfund reform legislation introduced by Sen. Robert C. Smith (R-N.H.), states that the protectiveness of a remedy shall be determined "considering the expected exposure associated with the actual or planned or reasonably anticipated future use of the land or water resources."¹² Elsewhere, the bill defines "actual or planned or reasonably anticipated" uses to mean the actual use, use that is authorized by local zoning or landuse decisions, and any other reasonably anticipated use that has a "substantial probability of occurring."¹³ Thus, S. 1285 shares H.R. 2500's focus on current or substantially likely uses.

The requirement that protectiveness be assessed in terms of current and substantially likely uses could be read to mean that a cleanup had to be considered sufficient so long as it reduced residual risks to acceptable levels assuming those uses of the site. EPA might be prohibited from conducting a more stringent cleanup even when the added cost of doing so was limited, the cost of maintaining long-term oversight was substantial, and the risk that institutional controls might fail was significant. Conversely, and perhaps ironically, such language might prevent EPA from restricting future use more narrowly, in the unusual case in which cleanup even to that level was very costly; such a remedy would not, by definition, be protective.

In contrast, H.R. 228, introduced by Rep. John D. Dingell (D. Mich.) and based on Superfund reform legislation from the 103d Congress, specifies that "In selecting a remedy, the President shall take into account the reasonably anticipated future uses of land at a facility as required by this Act."¹⁴ While it lays out in some detail the kinds of information to be considered, it leaves considerable discretion to determine how such information will be used. This approach allows the Agency to select remedies that are most appropriate, taking into account all the relevant factors, while conveying a clear direction to focus most heavily on actual and plausible land uses rather than remote and hypothetical ones.

If Congress wants EPA to focus more on protecting current or likely land uses, and less on uses that may conceivably occur but are not likely in the near term, a satisfactory compromise may be to make current or anticipated use the point of departure for cleanups, with discretion to choose alternative cleanup levels when other factors justify doing so. This would reorient the Superfund program toward the most significant risks while leaving decisionmakers the flexibility they need to deal with atypical situations.

The Danger of Oversimplifying Risk Assessment

Proposals to require land-use based remedies also tend to oversimplify the way in which land-use expectations should be reflected in risk assessment. In fact, as previously noted, they tend to overlook the entire distinction between risk assessment and risk management. Typically, such proposals specify a formula for the range of uses that is to serve as the basis for evaluating risk, such as "current and substantially likely" or "current and reasonably anticipated" uses. After this range of uses is determined, all further analysis of risk is apparently to be based on exposures associated with those uses.

[26 ELR 10361]

H.R. 2500, for example, states that remedial action shall be deemed protective if it protects against "realistic and significant risks" based on the "actual and reasonably anticipated future uses" of the site with "reasonably anticipated" defined to mean "uses that have a substantial probability of occurring."¹⁵ This would seem to direct the agency to ignore, in assessing risk, the possibility of any land uses other than those that exist already or have a "substantial probability."¹⁶ This conceals the existence of other land uses that are entirely possible but not certain enough to be considered substantially probable. For example, in an undeveloped area for which no specific development plans yet exist, this approach would ignore all possible uses except the current ones, even if it is likely that some further development will occur.

While uncertain uses should not receive undue weight, the possibility that they might occur should be taken into account. There is increasing recognition that evaluations of risk cannot be reduced to single numbers, and that risk assessments should instead inform decisionmakers of the range of potential risks and the likelihood of each.¹⁷ This is as true of land use as any other variable. Rather than choosing a single assumption about future land use based on some mandatory formula, risk assessors should evaluate the potential range of future uses and the probability of each occurring, when they evaluate the exposures that go into predicting risks. Current and likely future uses should receive the greatest weight in predicting risk at the site, but other potential uses should be taken into account as well. The analysis should also distinguish between sites for which future uses can be predicted with great confidence and those for which future use is highly uncertain.

"Selecting" Land Uses

Another common error is to assume that land use is selected in the process of designing the cleanup. Many discussions seem to anticipate that future land use will be "selected" in some process before, but in connection with, the remedy selection, and that the remedy then implements that decision. In most cases, however, the agency responsible for selecting the remedy has very limited control over future use at the site; in general, it will be able to do little more than choose the level of contamination left onsite (thus determining what kinds of uses can safely be conducted there), and perhaps establish some limited institutional controls restricting future use.

The process of determining what land uses will actually occur at the site involves a separate set of participants and occurs outside the remedy selection process. At best, the agency selecting the remedy can try to coordinate its process with those in which these other determinations are made.

More often than not, actual land uses are not "selected" by any single party or even any single governmental entity. Local governments, which tend to maintain most land-use controls, probably have the most say, but they can at most prohibit undesirable uses and guide market developments through their planning mechanisms, such as comprehensive plans. Even these governments, however, cannot ultimately control market forces, and they often find that economic pressures force them to change their original plans.

Moreover, local land-use decisions, especially those regarding hazardous waste cleanups, are often highly politicized and generally involve extended and unpredictable procedures. Local government officials for whom the cleanup is not necessarily the most important or most attractive issue may not give this decision the kind of priority that the remedy selecting agency would like it to. And local government officials may not be interested in facilitating a remedy that involves restricted use.

Most proposals for land-use based remedies involve establishing community-based advisory groups to make recommendations within the remedy selection process regarding the appropriate land use for the site. The precise role of those recommendations, however, is generally not clear, which may lead to misunderstandings in the future.

Community advisory groups can serve a number of functions. First, they may provide factual information as to what land uses are in fact likely or foreseeable at a site, although factual predictions and recommendations should be clearly distinguished. Second, they can shed light on the economic and social value attached to more stringent cleanup alternatives — information that should bear on remedial criteria such as cost-effectiveness. For example, if the community group shows no interest in preserving the area for potential residential use, there would be less justification for incurring the added cost necessary to clean up to residential levels.

Third, the community's recommendations can serve a political function. To the extent that obtaining broad "stakeholder" support is a goal of the program — and it certainly is under the Clinton Administration — community views need to be known. It is important, however, to distinguish the advisory group's political role from input it may provide on more objective matters, such as what land uses are likely and the level of economic benefit associated with any given cleanup level. Few if any proposals make this distinction, apparently lumping predictions about land use together with desires or preferences and failing to recognize that they play very different roles in the agency's remedy selection calculus.

There are several practical concerns regarding the role of such community groups. First, the preferences of a particular group may not necessarily represent the community as a whole; true "community preferences" may be arather elusive concept. Second, such groups do not necessarily speak for the local governments that actually determine the legally permissible land uses. An agency selecting a remedy would be unwise to proceed without confirming that the group's recommendations are consistent with the development that is actually likely to take place under applicable legal structures. Third, even local governments do not ultimately determine the kinds of uses that occur at the site: [26 ELR 10362] land users do. A community may wish to set aside an area as an industrial park, but if industry does not arrive, the site may have to be converted to some other use. Therefore, the agency selecting the remedy needs to assess objectively what kind of use is realistically likely at the site.

Remedial agencies can and should attempt to coordinate their efforts with the other processes through which land use is determined. Such coordination can attain clear benefits, most obviously when the land is currently undeveloped and future patterns of development are hard to predict. If the local governments and other community representatives agree, the agency may be able to conclude that some or all of the site will be used for industrial purposes and design the cleanup accordingly. Such efficiencies should be sought whenever possible. Closing military bases are a prime example of this kind of opportunity.

In other cases, the attempts of various government agencies to coordinate their efforts may be less successful. This may be due to conflicts between local interests and the rules under which CERCLA cleanups operate,¹⁸ an inability of the community to reach consensus or of the local government to take necessary political action, or simply a lack of realism in local planning about what kind of development is actually likely to occur. Recognizing these limitations is important to understanding that no single process can be assumed as the model by which all land-use based remedies will be established. Any system for selecting such remedies must provide enough flexibility to allow the process to develop case-specific results that work best at each individual site.

Implications of Land-Use Based Remedies

As a final note of caution about basing all cleanups on expectations about land use, it is important to confront the practical implications of such an approach. These considerations may give pause to anyone considering a mandate that cleanups address only current or immediately foreseeable uses. In particular, those who endorse land-use based remedies as a way to reduce the cost and complexity of cleanups may not fully appreciate that long-term regulatory involvement is generally required to implement them.

When land uses can be predicted with a high degree of certainty well into the future, remedies can generally be tied to the predicted uses with little difficulty. But when future uses are uncertain, it is necessary to address the potential that now-dormant risks will later arise. This requires imposing land-use controls, tracking risks over time, and responding as needed when those risks change.

Institutional controls are often proposed as the way to address long-term uncertainties of land-use based remedies. Institutional controls are legal or informational measures that will limit future activities at the site.¹⁹ It cannot be assumed, however, that effective institutional controls can be easily established in all or even most cases. And while institutional controls can be useful remedial measures, and probably can be relied on more than they have been historically, their success rate is not certain. Therefore, a healthy degree of skepticism about institutional controls is probably in order.

Generally, institutional controls fall into two categories. The first category is "governmental" controls established under the regulatory authority of a state or local government. Governmental controls, such as zoning or well-drilling restrictions, require independent action, outside the CERCLA remedy selection process, by another unit of government that may or may not be inclined to help complement the chosen remedy (e.g., local governments often prefer off-site disposal and may be reluctant to provide controls needed to support on-site remedies). Moreover, the local political processes through which such controls might be established are unpredictable. Even if controls are adopted, whether they will be retained and vigorously enforced may depend on the continuing support of the local governmental unit.

The second category of institutional controls consists of those based on private-property law. These controls, such as easements and covenants, require parcel-by-parcel negotiation and, if acquired, may be subject to common-law doctrines that limit their enforceability over the long term. For example, covenants are generally enforceable only by persons owning neighboring land. Easements can be enforced by persons not owning neighboring land, but under traditional common law such easements may not be transferred (e.g., by EPA to a state or local government), and they may not be enforceable against subsequent purchasers.

To be protective, all land-use based remedies should include a regular monitoring program to track land-use changes that might increase risks at the site. An ongoing enforcement role is also necessary to ensure that any implemented controls are complied with. Furthermore, institutional controls may simply fail (e.g., a zoning law prohibiting residential use may be repealed), or enforcement may be ineffective (e.g., when a new residential development has been established on a site intended to remain industrial). In such cases, the entire remedy may have to be reconsidered.

Basing cleanup actions on land use thus introduces new complications — especially the need for some kind of longterm regulatory presence. This is not necessarily a bad thing, but it does show that moving toward land-use based remedies has its costs as well as its benefits. It may be overly sanguine, therefore, to this approach as a cheap and easy way of making cleanups less burdensome.

We cannot have it both ways: we can focus on immediate risks and implement more stringent cleanups only if and when they are needed, but we cannot do so and at the same time expect to have "walk-away" remedies. Basing cleanups on land use will work only if we are ready to invest in the regulatory apparatus necessary to make such a program work over the long run.

Superfund reform legislation could take some basic steps to strengthen the ability of private parties and government agencies to implement, and of government agencies [26 ELR 10363] to enforce, institutional controls. Both Rep. Oxley's reauthorization proposal and H.R. 248 contain such provisions. They would create a class of statutory easements exempt from the historic doctrines that tend to limit the effectiveness of such instruments. Under these bills, statutory penalties would be available for violations of such easements, and the easements should be enforceable via citizen suit.²⁰

Easements are only one tool, however, and in may cases are not necessarily the best. For example, in laying the groundwork for remedies focused on limited uses Congress might set up more effective procedures for imposing regulatory controls over contaminated sites that do not burden state and local governments with the responsibility for enacting and implementing such controls. Congress could authorize EPA to adopt use controls in limited areas through a process similar to rulemaking but more locally based. To address the obvious impacts on local residents, suitable consultation and public participation rights could be afforded.²¹

An Alternative Paradigm

A useful contrast to CERCLA's current approach is the corrective action program under the Resource Conservation and Recovery Act (RCRA).²² Because of the longterm permitting relationship between the facility and the environmental regulator, RCRA tends to place less emphasis on taking aggressive steps today to address possible future risks. Instead, the program puts priority on "stabilizing" sites; that is, taking the interim measures necessary to ensure that short- and medium-term exposures are adequately addressed. While a final remedy is anticipated, it is less urgent so long as the permitted facility stays in operation.²³

A comparable arrangement, using permits or some similar device, couldbe devised to clean up hazardous waste sites under CERCLA. Measures to address current and immediately foreseeable risks could be taken in the short run, with more complete steps deferred so long as the site remained in its current use. The permit would allow the current owner, and any purchaser, to use the property so long as it adhered to certain conditions. Presumably, purchasers would remain liable for the cost of any cleanup when the use changed. The permitting relationship could be used for long-term planning (for example, phasing into more permanent cleanups if a change in use were anticipated). The permit would serve, in effect, as the institutional control for the site, in place of the other, less reliable kinds of controls discussed earlier.

Significantly, a permit approach would also encourage currently contaminated property to be sold to users whose use would not trigger additional cleanup. This is in contrast to the pressures the current liability system inadvertently creates, resulting in development in currently pristine areas. While such an incentive might exist under other land-use based remedies, any such incentive would be more diffuse and less effective.

Another attractive feature of this approach is that the permitting responsibility could be shifted to the states. Just as states assume permitting authority today under RCRA²⁴ and the Federal Water Pollution Control Act,²⁵ states could incorporate schemes for this CERCLA cleanup permitting program. Some states already have similar arrangements for addressing contaminated groundwater.²⁶ These programs could be expanded to include more general oversight, presumably including oversight of any future cleanup. In these states, the federal involvement at the site could end relatively early. And in those states that did not want this responsibility, or that are uncomfortable with their current obligation to ensure that institutional controls are maintained and enforced, the federal presence could continue.

Conclusion

Moving to a land-use based remedial system is a step in the right direction, but may be more complicated than many people assume. First, decisions need to be made consciously — not surreptitiously — about the extent to which such a change should also signal greater reliance on use controls as alternatives to treatment and containment. Second, it needs to be recognized that simply predicting future uses, and requiring the remedy to protect those, and only those uses, will lead to wrong decisions in many cases; predictions and policy judgments are not the same thing. Third, risk assessments should not be based on current or likely future uses to the exclusion of other possibilities; they should be based on the full range of possible uses (giving appropriately greater weight to the most likely ones). Fourth, we need to recognize the complexity of local land-use decisionmaking and not assume naively that the remedial process will "select" the land use. Finally, we must confront the fact that land-use based remedies will require greater ongoing care and feeding than traditional permanent remedies.

Recognizing the need for long-term regulatory presence at these sites suggests an alternative paradigm entirely for the way we conduct cleanups. Rather than viewing them as one-time events, perhaps we should simply set up a system for ongoing oversight to monitor risks at contaminated sites and take action when they become excessive. Such an approach — possibly using a permit mechanism — would focus efforts on real, immediate problems while ensuring that latent problems are not forgotten.

1. 42 U.S.C. §§ 9601-9675, ELR STAT. CERCLA §§ 101-405.

2. Land-use based remedies are not only less expensive, but they may help attain the greatest economic potential from contaminated sites. To the extent that currently contaminated sites can be targeted for industrial uses in the future, it may be possible to reverse the "brownfields" phenomenon of driving businesses out of historically industrial areas and into pristine areas that might better be put to other uses.

3. See National Oil and Hazardous Substances Pollution Contingency Plan, 55 Fed. Reg. 8666, 8710 (Mar. 8, 1990) [hereinafter National Oil and Hazardous Substances].

4. The NCP preamble does, however, indicate that when there is uncertainty about future uses, it may be appropriate to evaluate risks under two or more alternative scenarios "to estimate the consequences if the land is used for something other than the expected future use." Id.

5. Memorandum from U.S EPA Assistant Administrator Elliott Laws, Land Use in the CERCLA Remedy Selection Process 2 (May 25, 1995) (OSWER Directive No. 9355.7-04).

6. Id. at 6.

7. H.R. 2500, 104th Cong., 2d Sess. (1996) (creating new § 121(f)(1) of CERCLA); S. 1285, 104th Cong., 2d Sess. (1996) (creating new § 121(a)(1)(B)(i) of CERCLA); H.R. 228, 104th Cong., 2d Sess. (1996) (creating new § 121(b)(2) of CERCLA).

8. H.R. 2500, 104th Cong., 2d Sess. (1996) (creating new § 121(f)(1) of CERCLA); S. 1285, 104th Cong., 2d Sess. (1996) (creating new § 121(a)(1)(B)(i) of CERCLA).

9. § 102, H.R. 2500, 104th Cong., 2d Sess. (1996) (creating new § 121(f)(1) of CERCLA).

10. Id.

11. To avoid this result, EPA could interpret the statute to allow more flexibility. For example, "based on" could be read to mean "taking into consideration." To avoid the uncertainty created until that interpretation was tested, however, it would be preferable for the statute to provide flexibility on its face.

12. § 402, S. 1285, 104th Cong., 2d Sess. (1996) (creating new § 121(a)(1)(B)(i) of CERCLA).

13. Id. § 401 (creating new § 101(42) of CERCLA).

14. § 502, H.R. 228, 104th Cong., 2d Sess. (1996) (creating new § 121(b)(2) of CERCLA).

15. § 102, H.R. 2500, 104th Cong., 2d Sess. (1996) (creating new § 121(b)(5), (d) of CERCLA).

16. Similarly, S. 1285 specifies that risk assessment evaluate "the likelihood that potential exposure will occur based on the actual or planned or reasonably anticipated future use." Since the bill defines "actual or planned or reasonably anticipated" to mean actual uses, zoned uses, and others with a "substantial probability of occurring," this could require risk assessors to ignore uses that are a real possibility but do not have a "substantial likelihood" of occurring.

17. See, e.g., Memorandum from Carol Browner, EPA Administrator, to Assistant Administrators and others, transmitting policy on EPA Risk Characterization Program (Mar. 21, 1995) (on file with author). The attached policy states that "Information should be presented on the range of exposures derived from exposure scenarios." Id. at 2.

18. See 42 U.S.C. § 9621(b), ELR STAT. CERCLA § 121(b). For example, CERCLA contains a preference for on-site rather than off-site remedies. In contrast, local governments often prefer to have hazardous waste removed from the community.

19. The term "institutional controls" is sometimes used to refer as well to passive engineering methods, such as fences. The definition is consistent, however, with the discussion of institutional controls in the preamble to the 1990 NCP. See National Oil and Hazardous Substances, supra note 3, at 8666, 8706. For a discussion on institutional controls, see David Coursen, Institutional Controls at Superfund Sites, 23 ELR 10297 (May 1993); and John Pendergrass, Use of Institutional Controls as Part of a Superfund Remedy: Lessons From Other Programs, 26 ELR 10109 (Mar. 1996).

20. Massachusetts has already adopted a similar statute. MASS. GEN. LAWS ANN. ch. 21E, § 6 (West 1995).

21. Other steps could be taken to further limit intrusion on local prerogatives. For example, the use of this regulatory authority might be made available only when the relevant local government had been given an opportunity to adopt controls consistent with the remedy and failed to do so. The local government might even be given a veto over the federal control, although in that case it would be appropriate for the locality to have to bear any additional cleanup cost that resulted.

22. 42 U.S.C. §§ 6924(u)-(v) and 6928(h), ELR STAT. RCRA §§ 3004(u)-(v) and 3008(h). For a general discussion of the current corrective action program and issues that the Agency is seeking comment, see Corrective Action for Releases From Solid Waste Management Units at Hazardous Waste Management Facilities, 61 Fed. Reg. 19432 (May 1, 1996).

23. See id.

24. 42 U.S.C. § 6926, ELR STAT. RCRA § 3006.

25. 33 U.S.C. § 1342(b), ELR STAT. FWPCA § 402(b).

26. See, e.g., N.H. CODE ADMIN. R. ENV.-WS 410.18-.29 (1996) (groundwater management permits).