New Directions in Water Resources Planning for the U.S. Army Corps of Engineers (1999)

Several Water Resource Development Acts since 1986 have mandated a variety of environmental protection and restoration activities for the Corps. The Corps' authority to determine if changes can be made in existing structures or operations to improve environmental quality rests in WRDA '86, Section 1135 ("Project Modifications for Improvement of the Environment"). This continuing authority requires 25 percent cost sharing with nonfederal partners and limits federal expenditures to $5 million per project. Total appropriations to this section are generally limited to $25 million per year.

The Corps was given an additional mission with the passage of WRDA '90. Section 306 states that "the Secretary [of the Army] shall include environmental protection as one of the primary missions of the Corps of Engineers in planning, designing, constructing, operating, and maintaining water resources projects." In a complementary action, Section 307(a) set the goals of "no net loss of wetlands" and "an increase in the quantity of the Nation's wetlands."

The Corps was authorized to implement projects for protection, restoration, and creation of aquatic and ecologically related habitats, including wetlands, in connection with dredging of authorized navigation projects in WRDA '92, Section 204 ("Beneficial Uses of Dredged Materials"). This legislation essentially states that the Corps can now use material formerly called "spoils" for beneficial ecological purposes. Nonfederal sponsors are responsible for 25 percent of the project costs and 100 percent of operations and maintenance. The annual appropriation limit is $15 million.

The Corps' environmental programs were broadened further with passage of WRDA '96. Section 206 authorizes the Corps to engage in aquatic ecosystem restoration projects that will improve environmental quality when they are in the public interest and are cost-effective. Unlike legislation that restricts work to existing Corps projects, this stipulates that the Corps can fix environmental damage created by others. Nonfederal sponsors must pay 35 percent of construction costs and accept 100 percent of operation and maintenance costs. The federal limit is $5 million per project and the annual appropriation will not exceed $25 million.

Corps Initiatives and Planning Guidance

With publication of a relatively inconspicuous "statement of new environ-

mental approaches" by the Office of the Assistant Secretary of the Army (civil works) on June 25, 1990, the Corps embraced environmental restoration as a primary project purpose. The document states:

Assistant Secretary of the Army for Civil Works, Nancy Dorn, reiterated this position in 1992 in a letter to Max Peterson of the International Association of Fish and Wildlife Agencies. She wrote: "Early in 1990, we announced an environmental initiative for the Corps water resources development program which included the establishment of fish and wildlife restoration as a priority project output. We have provided strong budgetary support for this initiative" (emphasis added).

Many of the Corps' customers view the accommodation of new ecological knowledge as a dramatic shift in program emphasis. Some fear that the emerging focus on ecological processes in water resources management will detract from the Corps' traditional missions in flood damage reduction and navigation. It is more accurate to view the incorporation of ecological knowledge into the Corps' portfolio as broadening, rather than abandoning, its traditional programs. The Corps project of the future (and some Corps projects today) will incorporate environmental protection and restoration as primary objectives of water resources planning, and frequently as a means of achieving other water management objectives.

The Corps' involvement in watershed-scale, environmentally oriented water resources management processes is generally motivated by a blend of federal interests, such as the need to maintain navigation or reduce flood damages while conserving endangered species or restoring fishery resources. These projects often cross agency jurisdictions and require interagency participation. Many involve the alteration of structures built to perform a narrower range of tasks. These structures, including dams, levees, and drainage canals, are currently undergoing operating criteria reformulation or structural modifications to support environmental objectives,

while maintaining the Corps' traditional program goals. Such reformulations are currently occurring in the Columbia and Upper Mississippi River basins and the Everglades.

These projects are similar to the multiple-purpose projects earlier in the 20th century but have fundamentally different goals. The older multiple-purpose projects sought to control natural hydrologic patterns. They were generally designed for regulating streamflows, storing peak flows, and releasing them during low-flow seasons to provide water for power, navigation, and municipal and industrial supplies. By contrast, the reformulated projects frequently seek to restore natural hydrologic patterns and ecological processes.

It has long been understood that the relationships between water and land resources must be considered in water planning. The importance of this principle is reflected in the titles of Senate Document 97 and the WRC's P&G and P&S, all of which included the phrase "planning for water and related land resources." These considerations are even more important as the Corps seeks to further integrate ecological concerns into water resources planning and management.

Even more demanding are efforts to incorporate protection or restoration of ecosystems in water resource planning. Since 1986, the Corps has considered many possible restoration projects. Some have been completed, and others are at various stages of planning. In an attempt to understand the breadth and implications of the restoration program on a Corps-wide basis, the Institute for Water Resources conducted a national review of projects in 1996. Fifty-two projects were selected from an 80-project data base of reports and fact sheets. The national review showed that 16 different districts and divisions were involved in restoration (IWR, 1996). Fourteen projects were authorized under Section 1135, twelve under the Upper Mississippi River EMP; fourteen under the Coastal Wetlands Planning and Restoration Act; seven under Flood Plain Management with Environmental Features (pre-1986 legislation); and five under the Beneficial Uses of Dredged Material Program. Among their objectives were restoring salt marshes, barrier islands, wetlands, stream channels, and rivers; reducing sedimentation and saltwater intrusion; improving water level control; utilizing available sedimentation; reducing flood damages; and preserving environmental resources. Project costs were in the range of $79,000 to $20 million, but the vast majority involved less than $5 million.

The evolving restoration program has had a significant impact on Corps planning. Certain projects have taken longer to complete because of new information requirements and new cost-sharing rules. Biologists and other scientists are being given a more prominent role in Corps project planning. It may take time to fully assimilate those developments within the Corps' culture.

Many problems attend the calculation of benefits of the Corps' environmental projects. The benefits from traditional Corps projects can typically be expressed in monetary terms, such as the dollar value of flood damages avoided or additional commerce transported. By contrast, outputs of ecological restoration projects tend to include intangible values such as endangered species protection,

aquatic ecosystem protection or restoration, as well as aesthetic values. These types of project outputs defy monetization and do not easily fit into a traditional benefit-cost valuation framework. This section reviews Corps initiatives and general concepts related to the valuation of the outputs from restoration projects.

Mainly through its Institute for Water Resources, the Corps has had many research projects directed at improving its ability to evaluate the outputs of proposed projects. The Corps' traditional objective has been to recommend project alternatives and projects that maximize net national economic benefits, expressing benefits and costs in monetary terms. Deciding which outputs to quantify and whether to describe those outputs in monetary terms has caused a great deal of discussion within the Corps. Some project outputs, such as hydroelectricity, readily lend themselves to being expressed in monetary terms. Others such as flood protection can be monetized if based upon physical damage to structures and contents, although the process of arriving at a dollar value can be complex. Outdoor recreation provides environmental amenities traditionally valued in monetary terms. The Corps' restoration projects typically aim to enhance or restore ecosystem services such as biodiversity, ecosystem resilience, food web support, and natural carbon storage. These ecosystem services generally do not have market values, though there have been recent attempts at monetizing them (Costanza et al., 1997).

The difference between outdoor recreation and outputs such as electricity and flood protection is that outdoor recreation is not bought and sold in the marketplace, so an estimate of its monetary value must be based on nonmarket valuation techniques (Freeman, 1993; Just et al., 1982).¹ These techniques are designed to estimate the maximum amount a consumer would have been willing to pay for the recreational experience, rather than do without. Procedures for placing a monetary value on recreation are detailed in the P&G. The Corps also produced more detailed documents on the use of other nonmarket valuation techniques, such as contingent valuation (Moser and Dunning, 1986) and travel cost analysis (Vincent et al., 1986).

Using these nonmarket valuation techniques, the Corps could place a dollar value on all project outputs resulting from changes in environmental amenities. Desired improvements would increase net economic benefits whereas undesired impacts would decrease net economic benefits. This analysis would allow the Corps to continue to pursue its historical objective of maximizing net national economic benefits, as well as to maximize the national interest (OECD, 1989; 1992; 1994).²

Over the past few years, Corps headquarters has sponsored the Evaluation of Environmental Investments Research Program (EEIRP), a joint research effort by the Institute for Water Resources and the Corps' Waterways Experiment Station. The program's work units were assigned specific tasks related to the overall objective of

developing a practical evaluation framework for projects with substantial environmental components. The tasks of the nine EEIRP work units are: (1) determining and describing environmental significance; (2) determining objectives and measuring outputs; (3) objectively evaluating cultural resources; (4) engineering environmental investments; (5) using cost-effective analysis techniques; (6) applying monetary and other valuation techniques; (7) incorporating risk and uncertainty into environmental evaluation; (8) setting up environmental data bases and managing information; and (9) developing the evaluation framework (a summary of the EEIRP's research is contained in IWR, 1997).

Many federal agencies with which the Corps works, such as the EPA and the Fish and Wildlife Service, are often legally prohibited from considering economic values in their decision making procedures. Congress has given the Corps authorization (WRDA '86, Section 1135) to formally undertake environmental improvement and restoration projects without conducting formal benefit-cost analyses. The Corps has thus begun to consider maximizing other criteria, such as those featured in the Fish and Wildlife Service's Habitat Evaluation Procedures (HEP), as well as various measures of the environmental significance of resources, such as designation by preexisting legislative statute and subjective evaluation (Apogee Research, 1997). In both instances this usually results in adopting the specific legislative mandate of another agency as the project objective. Economic analysis in these cases typically takes the form of identifying the most cost-effective way of meeting a stated environmental objective and/or looking at the incremental cost of obtaining a marginal increase in environmental quality (Robinson et al., 1995).

The EEIRP reports identify three principal difficulties with the measurement of environmental outputs in quantitative but nonmonetary terms. The first is that most of the biological models available, like HEP, are principally designed to look at a single species. This orientation is not surprising given the endangered species mandate of the Fish and Wildlife Service, along with its traditional orientation toward hunting and fishing. However, the typical Corps project is aimed at general ecosystem restoration, and it is obvious that single-species models are often inadequate, even if the models are based on indicator species.

Second, examination of more general models for predicting detailed system-wide ecosystem effects suggests that the underlying science for developing easy-to-use portable ecosystem models is still immature. Progress is being made with less ambitious models such as the wetland evaluation technique, which considers a range of different functional relationships for wetlands. Problems with ecosystem modeling are also apparent in attempts to fully specify the linkages between environmental outputs of projects and human services (Cole et al., 1996). These linkages are important because they provide the interface between ecological and economic models, but they are not addressed at all in most available ecological models.

Third, for a project with multiple environmental outputs, it is not clear how to compare different outputs, much less how to make trade-offs when they conflict. A good example of this problem comes from the Everglades restoration project. Restoring the water flows in particular areas is beneficial to some endangered species but detrimental to others. Standard multi-attribute decision theory (Keeney and

Raiffa, 1976) explicitly or implicitly requires conversion of the different environmental outputs to some common unit so they can be consistently traded-off against each other. These trade-off issues become substantially more complicated when one moves from considering alternatives for a single project to attempting to make trade-offs on the larger scale of competing projects if the outputs are all denominated in different metrics.

The Corps is currently at a crossroads in defining its primary mission. The Corps has traditionally been charged to undertake projects in the national interest, however defined by Congress. One path open to the Corps is to continue to determine whether proposed projects are in the national interest and to recommend the best of these projects to Congress for funding consideration. If the Corps follows this path, the measure of the national interest must be revised to more appropriately account for projects' beneficial and detrimental environmental impacts.

The second path the Corps can take is to become a service provider to local, state, and federal agencies. This perspective is the logical implication of the "shared vision" concept put forth in the Galloway report and advocated in many recent IWR reports. Under this concept, local sponsors determine whether particular project alternatives are worth the costs to them. Local, state, and federal agencies then begin negotiations with the local sponsor to choose one of those alternatives. Within this paradigm, the Corps' role would be to: (1) facilitate cooperation of all relevant stakeholders; (2) provide technical expertise with respect to project design, construction, and maintenance; and (3) provide a large share of the chosen project's cost. Because the Corps would mainly provide a service to other agencies, a project's justification would come mainly from those agencies. The value of a project alternative is defined internally relative to each project, rather than reflecting an overall national objective.

A third option, which the Corps appears to be following, lies between the first two: adhering formally to the P&G but largely adopting the substance of the shared-vision concept. The Corps is doing this by embracing a broader concept of the national interest. The Corps sometimes adopts the objectives of other federal, state, and local agencies as the national interest it seeks maximize. This is coupled with a shift toward doing cost-effectiveness and incremental cost analysis at the project level as a substitute for, not part of, a comprehensive assessment of the benefits and costs of project alternatives. This more limited analysis is often sufficient for choosing among alternatives for a single project; however, it makes it difficult, if not impossible, to make objective comparisons across projects in order to decide which are the most desirable to fund.

Much of the Corps' dilemma regarding its future direction relates to whether it is possible to place a monetary value on changes in environmental amenities. This dilemma was clearly evident when the Corps started using economic analysis. Hanemann (1992) provides an interesting history of the development and evolution of benefit-cost analysis for public projects in the United States. In practice, there are four distinct options for monetizing environmental amenities.

Under option 1, a monetary value is established by legislative or administrative fiat. Many fines are of this nature: for example, a certain dollar amount per gallon of oil spilled. Under option 2, monetary value is inferred from agency decisions. When one project alternative is chosen over another and those alternatives involved different quantities of the environmental amenity and different costs, for instance, it is possible to infer how much the agency values the environmental amenity. Under option 3, monetary value can be set by a committee of experts (Cropper et al., 1992; McFadden, 1976). ³ The use of this option always raises questions about who appoints the experts and on what basis those experts are asked to determine the monetary value. Finally, under option 4, monetary value is inferred from the trade-offs the public is willing to make with respect to the environmental amenity and other opportunities. One of these four options is always chosen. Thus, implicitly or explicitly, a monetary value is always placed on changes in environmental amenities in the policy making process.

For the Corps, option 1 (administrative fiat) would appear to be an undesirable and unlikely option for deciding among alternatives, although for a specific project this approach has some advantages. Likewise, relying on option 3 alone seems undesirable in principle, but there are clearly bureaucratic forces that will make it more likely to be the chosen option in practice. Option 2 is the only one (other than option 4) under which monetary value is not directly stated, but because the value can be readily inferred, use of option 2 raises questions about what the gains are (and to whom) by not making the monetary value explicit.

Although nearly always the most difficult option to follow, option 4 is the only option directly consistent with the Corps' mandate to maximize the net benefits of its projects to the public. Option 1 is consistent only if the monetary value set by administrative fiat is equal to the public value. Option 2 is consistent only if the projects undertaken (and not undertaken) are consistent with the trade-offs the public would make. This may be straightforward for a single, relatively unique project approved by the public's elected representatives, but the difficulties of consistent decision making become readily apparent when there are more alternatives. Option 3 is consistent only if the experts chosen and the decision making criteria they are given are consistent with trade-offs the public would make.

The Corps is effectively mandated to assist Congress in maximizing the net public benefits of Corps spending by developing projects that maximize increases in public welfare and, with the assistance of the Office of Management and Budget, submitting those projects to Congress for approval and funding. Fulfilling this mandate requires the use of benefit-cost analysis and the comparison of benefit-cost ratios. Effectively, this is the issue of choosing among alternatives for a particular project versus assembling a portfolio of different projects.

A discussion of the use of economic analysis in Corps projects could be framed in many ways. One useful way is in terms of four stages of a project's preapproval life: (1) choice of several project alternatives; (2) assessing benefits and costs of each project alternative; (3) choice of a project alternative agreed upon by the Corps and local sponsor to seek funding approval; and (4) the Corps, OMB, and Congress' determination whether the proposed project alternative is funded. Some project alternatives are never considered, and some project options initially considered do not have their benefits and costs formally estimated. For some projects, no alternative can be agreed upon by the local sponsor and the Corps as a candidate for possible funding, and some projects the Corps recommends for funding are never funded.

Preliminary questions that arise with any project proposal include: How was this particular option chosen? Was the process sufficiently inclusive of stakeholders and did it consider a broad range of options? Were promising options prematurely eliminated and, if so, on what basis? Project alternatives that pass this initial reconnaissance are formally assessed with respect to their potential benefits and costs. At this level of analysis, the issue arises of whether and various types of benefits and costs can be measured. Should everything be placed in monetary terms? If not, is it possible in both a technical and an agency procedural sense to make consistent trade-offs between monetized and nonmonetized benefits and costs? At a more detailed level, specific questions arise about whether the Corps is using the most suitable techniques for determining benefits and costs. Here questions about the appropriate level of expenditures to determine benefits and costs arise and can influence what can be measured with reasonable precision and how successful monetization of environmental benefits and costs is likely to be.

At a more fundamental level, one must ask what are being counted as benefits and costs. There are clearly problems when land purchased to provide nonstructural flood damage reduction is counted as a cost while the benefits of open space and ecosystem restoration are ignored. Given the desirability of reducing the time and cost of doing assessments, there is a clear need for easy-to-apply criteria to enable planners to stop a detailed benefit-cost analysis of a project alternative if it is clear that it has negative net benefits or is clearly inferior to another alternative.

Having performed a benefit-cost assessment for each project option, one then needs to look at how the Corps determines its preferred alternative, how the local sponsor determines its preferred alternative, and how conflicts between the Corps and the local sponsors' choices are resolved. There are three obvious sources of potential disagreement between the Corps and the local sponsor. The first concerns the possible divergence between national and local benefits; the second, differences in financial constraints related to funding a particular project alternative; and the third, differences the two parties face with respect to the characteristics an accepted project has for relevant political funding sources. There are two other interesting issues to consider. The first is whether a "nonfundable" project option should be actively considered even if it is obvious that the project alternative is the one that maximizes net benefits. The second is whether changes in the share of the

project to be funded by the local sponsor and the Corps should be allowed formally to influence the nature of any compromise among project alternatives.

When the Corps provided all or almost all of the funding for a project and its evaluation, it seemed obvious that national objectives should prevail over local ones. The shift in financing responsibility toward local sponsors introduces the possibility of a substantial divergence between national and local interests. With project financing now a shared obligation, it is not clear how to balance national and local objectives.

The Galloway report (IFMRC, 1994) typifies the current dissatisfaction with the role of the NED objective. That report argues for a coequal objective system in which environmental quality (EQ) has equal status with NED (essentially a return to project accounting with the P&S framework). The basis of this argument is that "unquantifiable environmental and social values" cannot be given adequate consideration in the analysis of projects. This argument, however, is increasingly rejected by economists working on environmental, health, and safety issues (Arrow et al., 1996), who contend that the benefits and costs of government projects and regulations should be measured in monetary terms and argue that benefit-cost analysis should play a major, but not necessarily decisive, role in government policy making.

Since the early 1960s, four key changes have occurred in the treatment of environmental values in benefit-cost calculations. First, there has been substantial progress in measuring a large class of environmental benefits that, prior to the introduction of multiple-objective evaluation in the P&S, were treated as "intangibles." Second, the movement toward monetization has encountered opposition from groups who argue that assignment of monetary equivalents to environmental projects is unethical. Third, there is much greater recognition of the conceptual and practical difficulties of developing nonmonetary environmental quality (EQ) measures comparable across project alternatives and projects. And fourth, there is a greater commitment to using nonmonetary EQ measures.

The Corps is currently required to assess the economic benefits and costs of water projects under the P&G. It is allowed to consider other factors contained in other accounts specifically related to environmental, social, and regional impacts not adequately captured in the NED calculation. The P&G contain an evaluation framework similar to that in the P&S. Each successive planning guidance document has moved more unquantifiable factors from the environmental account to the NED account. The role of the environmental account has not changed substantively over time, but the P&G formally took the position that only consideration of NED was necessary, whereas previous guidance documents contained language more consistent with a final multiple-objective trade-off.

At one level, the basic economic guidance provided in the P&G remains largely sound, but the technical guidance on the application of particular techniques in particular circumstances should be updated. Since the early 1980s there has been a tremendous increase in the use of various economic techniques to place a monetary

value on nonmarketed goods (Braden and Kolstad, 1991; Cropper and Oates, 1992; Freeman, 1993). Those techniques are now on more solid theoretical ground.

The two most intensely debated issues with respect to monetizing changes in environmental amenities are the inclusion of so-called passive-use, or existence, values in economic analysis, and their measurement by a survey method known as contingent valuation. These issues have drawn extensive comments both inside and outside the economics community, which is not surprising considering the stakes involved. In response to the controversy over the use of contingent valuation in natural resource damage cases, the National Oceanic and Atmospheric Administration (NOAA) put together a blue-ribbon panel cochaired by U.S. economists and Nobel prize winners Kenneth Arrow and Robert Solow. The NOAA Panel concluded that: "CV studies can produce estimates reliable enough to be the starting point for a judicial or administrative determination of natural resource damages including passive use values" (Arrow et al., 1993).

In drawing this conclusion, the panel first considered and rejected (e.g., Rosenthal and Nelson, 1992) arguments that lost passive-use values should not be counted as economic loss.⁴ The key implication is that if all passive-use values can be reliably measured, then there are no "unquantifiable environmental and social values" left in an environmental account to be traded off against the NED account. After hearing substantial public testimony and submissions, the NOAA panel concluded that passive-use values could be reliably measured with contingent valuation. They further recommended guidelines to help ensure the reliability of the application of contingent valuation techniques. These recommendations have considerable implications for the economic analysis of a wide range of Corps projects because they are very expensive to implement.⁵

Acknowledging the difficulties in monetizing the benefits and costs of environmental improvements, the committee ultimately concluded that the failure to do so actually does more harm to the environment. The arguments in support of this position are simple: harm done to ecosystem services does not reduce the NED estimate, and improvements do not increase the NED estimate. At present, harm to the environment either acts as a constraint on the feasibility, or mitigation (partial or full) is required after harm reaches some level. With respect to ecosystem restoration, multipurpose projects are clearly the most impacted by the failure to put changes in ecosystem services into monetary terms, as there is no way to compare

multipurpose projects without expressing the outputs of those projects in a common metric. As some outputs of those projects are already in monetary terms, a monetary metric is the natural method to use. For purely environmental projects, the committee believes that existing Corps authorizations that do not require a benefit-cost assessment provide sufficient flexibility for these projects to come forward to Congress for its decision.

Three aspects of the Corps' analysis of projects that have changed are relevant to discussion in this section. First, Corps projects on average have become smaller and more localized. Second, explicit environmental elements have become more common in Corps projects and often go well beyond simple mitigation requirements. And third, the number of alternatives considered in the decision process has increased. The first aspect of the change in Corps projects suggests that the costs of conducting a comprehensive, rigorous analysis for each project may be quite large in relation to the project's costs. Indeed, these costs may exceed the net benefits of even project alternatives with extremely favorable benefit-cost ratios. The second suggests that the valuation of changes in environmental amenities will play a larger and hence more controversial role in choosing among project alternatives. The third aspect implies that the costs of doing a rigorous analysis for the purpose of selecting a project alternative are increasing. Each of these aspects of Corps projects suggests the need to develop a formal set of procedures for evaluating small projects in a cost-effective, standardized way and to consider their impacts at a larger level, such as the river basin.

The first step in assessing the benefits and costs of a project alternative is to enumerate them clearly in physical/biological terms. This is consistent with the NEPA process and can provide the inputs to the EQ account where environmental amenities are involved. For example, a recent study (Cole et al., 1996) categorized benefits in typical Corps projects as arising from the following sources:

1. Direct uses in production:
2. • resource input in navigation and hydropower production;
   • increased food and fiber production, and commercial and industrial production;
   • water input for industrial processes and municipal/residential water supply;
   • commercially harvested fish, wildlife, and natural products.

2. Direct consumptive and nonconsumptive uses:
3. • aquatic habitat-based consumptive recreation (fishing, swimming, boating);
   • amenities and aesthetics (visual and cultural benefits);
   • water-enhanced, nonconsumptive recreation (picnicking, bird viewing, camping).

3. Indirect uses:
4. • flood storage and conveyance;
   • sediment retention;
   • wind and wave buffer;
   • pollution uptake and detoxification.

4. Passive, nonuse, or option values:
5. • values associated with knowing an ecosystem and its services (e.g., biodiversity) are intact.

6. Some of these project outputs have market prices (e.g., electricity), though many do not. In cases where market prices are not available, nonmarket valuation techniques need to be used.

   Nonmarket valuation techniques attempt to measure the public's maximum willingness to pay for a project output that is not normally bought and sold in the marketplace. At the individual household level, this is the maximum amount that a household could be asked to pay for a project output and still be as well-off as before the project output was provided (Just et al., 1982). The nonmarket valuation techniques (Freeman, 1993) available to the Corps, using ecosystem services as an example, can be summarized as follows:

   1. Factor income/productivity approach. Under this approach, benefits and costs can be evaluated by determining the contribution of the change in ecosystem services to the value of goods that are sold directly in markets (where those ecosystems serve as an input to marketed goods). Classic examples are the contribution of wetlands to commercial fisheries and of water quality to agricultural production.
   2. Travel cost analysis/averting behavior/household production function. This approach looks at how use of a nonmarketed good changes as the cost of a marketed good necessary to use the nonmarketed good changes. From this relationship, one can infer how much the person would have been willing to pay to use the nonmarketed good. The classic example of travel cost analysis is a recreation site where participation falls as the cost of getting to the site increases.
   3. Hedonic pricing (property/wage). This approach examines how the price of a good changes as its characteristics change. The classical examples here are to examine how housing prices change with respect to proximity to a lake or to look at how wage rates change with the level of on-the-job accident risks.
   4. Contingent valuation. This approach surveys consumers about trade-offs. Based upon their responses, the values that the public is willing to pay in order to obtain changes in ecosystem services are inferred.

   5. A fifth approach, which effectively draws from the other four approaches, is known as the benefit transfer method. This method takes monetary estimates from other studies, which have valued similar ecosystem services, makes appropriate adjustments to account for differences in circumstances, and applies that estimate to the project alternative being evaluated.

Each of these approaches is appropriate for particular types of nonmarketed goods under particular circumstances. The key is to apply the right tool given the resource constraints on the assessment. It appears to the committee that the Corps should probably pursue the benefit-transfer approach (Brookshire and Neal, 1992) in analyzing its small and medium-sized projects.

Three key difficulties with the benefit transfer are that: (1) it can only be as good as the original studies it is based upon; (2) the factors on which adjustments should be made are not often well understood; and (3) there is not a great deal of experience with conducting benefit transfers for nonmarketed goods other than health effects and outdoor recreation. As the Corps undertakes a large number of multipurpose projects that have similar outputs, they could routinely commission studies to obtain original estimates with respect to the types of outputs they are frequently called upon to value. The Corps can also sponsor original research to look at the other two issues in benefit transfer.

Other issues relating to economic analysis of environmental restoration projects include the measurement and discounting of environmental outcomes and how cost-sharing rules apply to environmental projects. Corps ecologists rely frequently upon the Fish and Wildlife Service's Habitat Evaluation Procedure (HEP) to quantify outputs of environmental programs. Output measures used in HEP are referred to as "habitat units" obtained by multiplying affected areas of habitat for a selected species by a habitat suitability index for that species. As such, habitat units reflect both the quantity and quality of habitat as an output for the selected species. For example, if a 100-acre parcel of wetlands is restored as habitat for a particular migratory fowl, and the quality of that habitat is assigned a suitability index of 0.6, the output would be 60 habitat units for that particular fowl. Corps guidance, such as the Evaluation of Environmental Investment Procedures: Interim Overview Manual (Harrington and Feather, 1996), defines these habitat units as important alternative metrics for assessing restoration project benefits. Indeed, it would appear that the purpose of environmental restoration projects is solely to produce HEP units in the most cost-efficient manner. The difficulty with HEP and similar methods is that they capture only a part of the national interest. They focus on habitat aspects of ecosystems and then only on a select species. Public preferences may be stated in more holistic views of which ecosystems should be restored.

Even if a perfect measure of environmental effects was found, other issues would be at stake. Regardless of whether the theoretical habitat units are general ecological goods or species specific, one must still recognize that they can be counted across three dimensions: time, quality, and surface area. Furthermore, there are many potential trade-offs across the three dimensions. For example, consider a case in which the goal is to increase the quantity of a specific fish species that is popular with anglers. A single planning alternative could produce a large quantity of small fish in a short period of time (perhaps all in one year) and in a highly confined area. A different alternative might produce the same number of fish, but they could be larger specimens available over longer periods of time, perhaps many years, and spaced

across a wide geographic area. The two alternatives might be equal in one unit of measurement, but may not be equal in terms of production costs and preferences to the anglers.

There are instances where cost-effectiveness analysis of supplying some desired number of HEP units is appropriate. A common instance is where protection of an endangered species is called for and the particular HEP model is calibrated to that species. In this instance, the Corps is helping to fulfill a legislative mandate and there is no need to conduct a benefit-cost analysis, only to accomplish the mandate at the lowest cost, subject perhaps to a spending constraint.

The interesting and challenging task for the Corps is to link outputs of ecological models, in whatever units are appropriate, with economic benefits. In doing so, the Corps should be careful not to confuse HEP unit maximization with actual restoration of the original ecosystem.

Conventional Corps planning studies discount all future flows of monetized benefits to present values using a single interest rate mandated by the Office of Management and Budget. Several issues arise regarding the choice of a particular discount rate and how it is chosen. One of these issues is the following: What is an appropriate choice when a current change in an environmental amenity is being traded off against a distant future change in that same amenity? Other federal agencies, such as the EPA and NOAA, are starting to consider this issue, as should the Corps. Coordination of a consistent federal policy on this issue would be desirable.