On Earth Day, 1993, President Clinton, acknowledging that America needs to incorporate environmental values into economic and political decisionmaking, called on the Department of Commerce to develop methods for incorporating the use of natural resources into the U.S. national economic accounts, particularly the income and product accounts (e.g., gross domestic product (GDP) and gross national product (GNP)). Recognizing that "standard of living" involves many factors—among them per capita income, protection of our environment, rate of resource depletion, and enhancement of human resources—maximizing GDP, under current definitions, is clearly a limited surrogate for maximizing "standard of living," ''quality of life," or "human welfare." The Joint Economic Committee of Congress has already held hearings on the subject, and instructed the Congressional Budget Office to conduct a study (CBO, 1994) on the conceptual framework and information needed to revise the national accounts to incorporate environmental and natural resource concerns. The Environmental Protection Agency and the Department of Agriculture have begun pilot projects designed to identify potential problems involved in attempting to revise the national accounts in this fashion. The United Nations (UN) and the World Bank have for some time been engaged in efforts in this area, and have incorporated natural capital estimates in partially revised GDP accounts for a number of countries (Lutz, 1993; UN, 1993; UNEP, 1989). Thus, adjustments to consider natural capital might well represent the next major milestone in the continuing evolution of the national accounts.
At the National Research Council (NRC) 1993 workshop on "Valuing Natural Capital for Sustainable Development," considerable attention was focused on the national income accounts, which, in their calculations of GNP or GDP, make no or very inadequate allowance for natural resource use and environmental degradation. As a consequence, there is concern that national policy underprotects natural assets, leading to their excessive depletion and degradation, relative to what would be a more balanced national strategy. Some argue that if the national accounts reflected a more accurate picture of environmental values, there would be reduced losses on the environmental and natural resource side, and greater concern about the loss of future economic potential. Others believe that revising such economic balance sheets to handle
the multiplicity and complexity of natural assets is difficult, perhaps impossible, and that decision frameworks should recognize that a variety of strategies and accounts is needed to address issues of national welfare.
Incorporating environmental values and concerns in forecasting, assessing, and making policy for the nation poses formidable challenges. And yet, the incompleteness of the current national income accounts forces us to reexamine our definitions of economic progress and national prosperity. Revising these indices also demands consideration of our obligations to future generations—what sort of environmental legacy and natural resource endowment we wish to leave them. It also, finally, raises the question of international scale. The environment does not stop at our country's borders. Many of our most valuable resources, the atmosphere and oceans, are shared by a number of countries and may be considered global "commons."
The workshop sought to address some of these difficulties, explore the state of the methodology, and gauge the adaptability of the national income accounts. Basic theoretical issues were outlined, including the implications of the different concepts of "sustainability," work in progress was presented, and suggested approaches for future research were considered.
ENVIRONMENTAL VALUES AND THE NATIONAL ACCOUNTS
Rudimentary efforts to measure a nation's income and wealth go back several hundred years. But the comprehensive, systematic, and regular collection and analysis of national accounts statistics date from the first part of the twentieth century. In the United States, formal national income estimation began in the Department of Commerce in 1933, at the depth of the Great Depression. The timing was not entirely coincidental. The severity of the slump gave rise to theoretical studies and policy guidance, largely associated with the work of British economist John Maynard Keynes, designed to spur macroeconomic revival. The prevailing climate of anxiety inspired the launching of major statistical programs that would provide the means of tracking national economic performance. Among many experts contributing to this undertaking in the U.S. was the scholar Simon Kuznets, whose work in this field earned him the Nobel Memorial Prize in Economics in 1971.
Over the years, national accounting concepts have become more refined, anomalies have been corrected, and statistical detail has increased. The scope of national accounts has also expanded to comprise not merely income and product accounts, the components with which we are most familiar, but input-output transactions, national wealth statements, and financial flow-of-funds tables. (A useful collection of articles on the evolution of national accounts and some key conceptual issues appeared in "The Economic Accounts of the United States: Retrospect and Prospect," in the Survey of Current Business, 1971.) Many countries currently maintain elaborate systems of national accounts, and the UN Statistical Office has pursued efforts at standardizing accounting treatment and concepts among nations.
National Accounts Dilemmas
Although natural capital is the focus of this report, shortcomings in the scope and content of the GDP accounts, and caveats in their interpretation and use, have been addressed in scholarly writing and analysis over many decades. A recurrent issue during that time span, for example, has been the extent to which the GDP actually tracks human welfare. A frequent criticism concerns the treatment of "defensive" expenditures in the national accounts. For example, additions to the GDP from oil spill cleanup operations in Prince William Sound may seem unusual until we are reminded of the many inherently "unproductive" things we pay for to protect the gains, or overcome the indulgences, that rising income makes possible—spending on items like residential security systems, diet plans, and national defense. A third GDP dilemma relates to the adequate measurement of certain government activities, such as research. Finally, there is a recognized failure to deal with "voluntary" activities, such as unpaid household work, which, if purchased in the marketplace, would merit an addition to GDP. While there has been no theoretical reluctance to improve the accounts, the translation of principles into practice poses formidable challenges.
With respect to natural resources and the environment, defects in our measurement practices have also been recognized for many years. A landmark 1972 National Bureau of Economic Research volume, The Measurement of Economic and Social Performance (NBER, 1972), anticipated a major need for linking resource and environmental quality trends to national accounts, recognizing that conventional practice "gives incorrect indications of changes in welfare . . . because it fails to allow for the disamenities associated with industrial growth, particularly pollution of air and water."
Overcoming these shortcomings in the treatment of nonrenewable and renewable resources, much less the more elusive phenomenon of environmental quality, is problematic. While the estimated depreciation of reproducible physical capital (plant and equipment) allows, albeit roughly, for the lost stream of future income that must be compensated for in order to keep a country's economy whole (" sustainable"), no such allowance has been made for changes in the stock of natural capital. Recognition of this anomaly in national income measurement practices has stirred increased research and debate about how best to reform nations' social accounts in a more environmentally sensitive direction.
In recent years, contributions to study of the problem have emerged from four quarters.
Some analysts, notably Herman Daly and his collaborators, acknowledge the defects of GDP measures that exclude resource depletion and environmental damage. They devote little attention to this particular problem, however, because they challenge the utility of GDP accounts in general.
Other analysts accept the inherent usefulness of a system of national accounts but, both through argument and actual case studies, recommend specific ways of embodying natural capital in the accounts. This group includes Robert Repetto, Henry Peskin, Salah El Serafy, David Pearce, and numerous other scholars.
International institutions, principally the World Bank and the UN Statistical Office, have incorporated natural capital estimates in partially revised GDP accounts for a number of countries.
Finally, a number of nations' official statistical organizations, including those in Norway and the Netherlands, have begun limited implementation of environmental and natural resource accounting reforms.
Studies by Repetto, Peskin and others have broken new ground in this area. Beginning with his 1989 World Resources Institute study, Wasting Assets, Repetto, at times in collaboration with others, has not merely critiqued prevailing accounting practice on conceptual grounds, but has attempted at least partial recalculation of some developing economies' national accounts to reflect net natural resource depletion. Thus, the 1989 study suggested that Indonesia's net domestic product (NDP) growth fell markedly below GDP growth once allowance was made for natural resource depletion. A more recent example of such estimates is a recalculation of Costa Rica's NDP, again reflecting not only the conventional, and relatively constant, annual depreciation charges against the country's GDP but also the rapidly rising estimated depletion of three major natural resource categories: forests, soils, and fisheries (Repetto et al., 1991). Such divergence between conventional and resource-adjusted accounts, particularly should it show signs of widening, could send a potentially significant message to policy planners concerned with economic growth. It should be noted, however, that even when natural resource depletion goes on at what may be deemed an unsustainable pace, the ultimate consequences for social welfare will be much affected by the nature of the investments into which the proceeds of natural resources liquidation are plowed. A country depleting, say, its phosphate stocks can squander the proceeds of the sales, unsustainably, on luxury cars for government officials or invest the proceeds in education.
Paralleling and often including the work of academic scholars, international institutions such as the World Bank and the UN Statistical Office have for some years been pursuing an active research program in resource and environmental accounting. In the case of the UN, that work represents a subset of the organization's involvement, over many decades, in developing, independent of natural resource and environmental concerns, a standardized global System of National Accounts (SNA). Two notable studies, published by the World Bank, are Environmental Accounting for Sustainable Development (World Bank, 1989) and Toward Improved Accounting for the Environment (Lutz, ed., 1993). The latter volume contains case studies which apply alternative resource and environmental accounting techniques formulated by the UN Statistical Office to Mexico and Papua New Guinea. For Mexico, whose statistical system is, as one might expect, far more robust and developed than that of Papua New Guinea, the adjusted NDP tends to be in the range of 90 percent of the conventionally measured NDP. But what is more significant is that Mexico's net investment, when corrected for natural resource depletion and environmental effects, is roughly half the conventionally measured estimate of net investment. To be sure, such revaluation experiments are far from comprehensive and methodologically definitive. Yet they offer valuable insights, exposing measurement problems and demonstrating opportunities and directions in further pursuit of such efforts, particularly in developing countries where natural resource sectors, in most cases, constitute a proportionately much larger share of the economy than in industrial countries. They may also have profound implications for economic policymaking. A nation pursuing an aggressive industrialization strategy because it produces increased investment and rapid economic growth by conventional
measures might reconsider that course of action if its national accounts revealed only modest investment and growth figures and rapid depletion of valuable assets.
A number of advanced countries have themselves actively engaged in experimentation which could, in time, lead to natural capital-adjusted modification of their official accounts. To date, however, these efforts have been exploratory and selective and have not followed any cross-country standardization rules that would permit international comparison. France, the Netherlands, Norway, and the U.S. have emerged as the principal players in the field, with France and the Netherlands emphasizing development of an integrated conceptual framework, and Norway and the United States concentrating on mobilizing extensive resource and environmental databases that are used for many independent accounts or policy objectives, but have not been incorporated into the main national accounts.
ENVIRONMENTAL ACCOUNTING CHALLENGES
It is clear that social accounting reforms embodying natural capital, if they are to materialize at all, must proceed in a step-wise, evolutionary, and pragmatic manner. To press for short-order, comprehensive, and defensible accounting treatment of natural resources and environmental assets is to both harbor illusions and probably doom the whole effort in its infancy. The success of policies aimed at valuing natural capital for sustainable development demands innovations in accounting acceptable to policymakers, experts, and a large lay public. Consider just three major stumbling blocks, all of which need to be overcome to ensure the viability of any environmental accounting framework.
Establishing definitional boundaries. While it may prove feasible to incorporate, say, changes in the stock of certain minerals or in ground-level ozone concentrations in an urban airshed, the outer limits of the scope of natural capital are quite blurred. Species extinction, encroachment on biodiversity, impairment of visibility, or noise pollution are legitimate examples of resource or environmental degradation; but for now, these effects may be beyond our ability to measure.
Estimating net physical depletion. Depending on the resource in question, quantifying the net volume depleted can prove to be a formidable undertaking, as Sadoff's paper (in this volume) shows in dealing with the tropical hardwood sector of Thailand. Even where the statistical basis is well established, as in the case of U.S. petroleum resources, the extent to which a year's oil use has been offset by new finds or upward revision of economically recoverable reserves is subject to statistical uncertainty and, frequently, to disagreement among estimators. Many depletable resources—for example, copper—are not truly depleted but become embodied in products from which they may at some point be reclaimed. Further, over some time scales, many seemingly renewable resources (such as old-growth forest, ground water aquifers, or top soil) can be lost to the benefit of society, meaning, in essence, that they behave as depletable resources.
Monetization. Full integration of natural resource and environmental quantities into national accounts requires monetization (i.e., assignment of dollar values) consistent with the valuation basis and principles governing conventional GDP measures. This task presents
multiple challenges. Some resource and environmental services are not mediated through market transactions and so require the application of imputed prices. Even in the case of items traded in markets, the choice of a price which appropriately captures the discounted present value of future income foregone by depletion of the resource stock is at best ambiguous. In numerous cases, there remains widespread disagreement, even among scholars familiar with the issues, about both the feasibility and desirability of imposing market-like valuation to environmental assets and processes.
ISSUES IN ENVIRONMENTAL ACCOUNTING
In response to the challenges posed by environmental accounting, workshop participants examined a number of conceptual, technical and philosophic issues. The discussion revolved around four distinct, though interrelated, topics: sustainability, substitutability, irreversibility, and intergenerational equity.
In addressing measurement and methodological issues raised by environmental accounting, experts in both the social and physical sciences have focused considerable attention on the concept of sustainability. This concept was the starting point for discussion at the workshop, where it was examined from a variety of different viewpoints, including those of the life and physical sciences, economics, and industrial, technical, and commercial fields.
Sustainability implies a process which can be continued indefinitely, a course of action that does not include within itself the seeds of its own end or defeat. "A sustainable path," according to the Brundtland Commission definition, "fulfills the needs of the present without compromising the ability of future generations to meet their own needs" (World Commission on Environment and Development, 1987). A more immediate way to put this is to ask whether we, as a world community, are living on interest or capital, and whether our children will inherit as much as we did from our parents. Different economists define economic sustainability in different ways. Solow, for example, identifies a sustainable path as one "that allows every future generation the option of being as well off as its predecessors" (Solow, in this volume). Pearce et al. (1989) suggest that "sustainable economic growth means that real GNP per capita is increasing over time and the increase is not threatened by 'feedback' from either biophysical . . . or social impacts,'' while "sustainable development means that per capita utility or well-being is increasing over time, or . . . that a set of 'development indicators' is increasing over time,'' and that the development, again, is not endangered by negative feedback. Sustainable development clearly does not presuppose a fixed composition of the capital stock; within limits, there is ample opportunity for substituting among various types of capital. Of course, these economic definitions of sustainability assume an anthropocentric frame of reference. They are concerned with sustainability of an ecosystem only so far as it impinges on human welfare.
Economic production involves a transformation of raw materials into finished goods, and the success of many economic enterprises is tied, to some degree, to the cost and availability of
raw material. Whenever a society or economy depletes its stock of resources it reduces its future productive capacity. Unless the stock can be renewed, or unless some substitution for the depleted stock can be made, the path is unsustainable. For a sustainable economy, the net "disinvestment" of natural capital or any other forms of wealth which contribute to economic activity, must be minimized or prevented. As noted already, this does not necessarily require replacing identical stocks, as long as nations and firms can renew or substitute for depreciated resources. Thus, in the view of some experts, economic sustainability need not require conservation or replacement of original resource stocks.
While conserving natural capital does imply absolute protection against net physical depreciation or depletion, such a rigidly conservationist view is not inherent in all variants of the sustainability concept. Biologic sustainability, for example, is not tied to preservationism; every single biological entity does not need to be conserved, indeed cannot be, in order for a biological system to qualify as sustainable. Any determination of biological sustainability depends on the scale of the population or ecosystem in question. Policies might protect the spotted owl population in the northwest, but fail to protect owl habitats on a national or international scale. Conversely, the bison population in the U.S. as a whole might be sustained, while an individual herd is wiped out.
Experts have experimented with a number of strategies to promote biologic sustainability. Compensatory mitigation is one replacement strategy that has been tried (Science, 1993). Policies of compensatory mitigation mandate ecologic restoration or creation to replace natural areas, primarily wetlands, lost to development. The limited success of such strategies is due not only to the difficulty of enforcing proper compliance with regulations, but also to the fact that even with the best attempts, the resulting "artificial" ecosystems have often proved flawed, lacking the complexity, adaptive properties, and dimensional depth of the natural ecosystems they were intended to replace. For example, artificial ecosystems are often constructed in relation to one set of environmental variables but may have lower resistance and resilience to changes in driving variables (such as global climate) than their natural counterparts.
Considerations of sustainability often draw a distinction between natural resources which are possible to replenish, such as municipal water reservoirs refilled by water treatment, and resources which cannot be replenished, such as fossil fuels or species near extinction. But this distinction is not always clear-cut. For example, forest lumber may be replenished, in the sense that the forest may be managed as a farm, with new tree crops planted regularly to ensure steady yield, but it may not be possible to similarly renew the forest's value as part of an ecological system. Wetlands replaced according to compensatory mitigation policies may fulfill established guidelines, but they do not reproduce the specific ecosystem of the lost wetland, and often fail to secure a habitat for that original wetland's species.
As noted above, in discussing replenishment one must consider the closely related concept of substitutability, in both its economic and environmental or biologic senses. Lost natural assets may be compensated for by other natural assets. For example, if a given industry is dependent on some particular resource which becomes increasingly scarce, it is possible that the industrial
processes may be reconfigured to rely on some alternative material or process. Whale oil lamps were replaced by alternate lighting methods and mineral-based oil replaced whale oil as a lubricant. (It should be noted that while this was an effective substitute for the economy as a whole, the whaling industry itself all but disappeared, and some whale species never recovered.)
A much more recent example of substitutability, based on a less direct aspect of depletion, involves the refrigeration and electronics industries that previously employed chlorofluorocarbons (CFCs) as a refrigerant and in some of their industrial processes. As these chemicals have become increasingly implicated in ozone depletion, the industry appears to be successful in finding substitutes, albeit at higher cost. In this case, it was stratospheric ozone, not the chlorofluorocarbons, that were depleted; CFC use led to the depletion of another natural resource. Not only can different natural resources substitute for each other, but it is also possible to substitute greater inputs of labor, reproducible capital, or renewable resources (e.g., solar energy) for inputs of a nonrenewable resource.
Whether or not a given material is an adequate substitute for a specific resource depends on the viewpoint and particular needs of the user, and on the available technology. The influence of advancing technology is readily apparent in the extractive industries. In mining, for example, due to technical advances in extraction and processing, ores of progressively lower grades are being profitably extracted. What were considered "tailings" or waste rock left over from mining operations in the past, would today be considered as source rock. Similarly, today's tailings may be valuable one day given future technologic advances. This has raised speculation that today's landfills may be tomorrow's ore bodies, in a twist on the ecological notion of composting (Adams, 1993). It also suggests that in some cases we may be underestimating the size of our depletable resource base.
Compensatory mitigation, as described above, is yet another type of substitutability, in which a man-made ecosystem is substituted for a natural one. But as has been seen, biologic substitutability is often more complex than economic substitutability. In an economic process, a resource may be limited to a small number of readily defined uses, perhaps even one single use, whereas a component of an ecosystem may play a vast number of complex roles, whose nature we can only begin to grasp. For example, while more than one species of bird can act as a predator of a certain insect, each of those bird species may be unique as to the precise niche it fills in the ecosystem. Not only is there no actual substitutability for species, but even for cases of near substitutability, the time scales that are involved often exceed many human lifetimes. Some have in fact argued that natural capital and man-made capital are complementary and only marginally substitutable (Daly, 1990).
One of the implications of an emphasis on sustainability is that irreversibilities—losses of resources that cannot be recovered—should be avoided wherever possible. An exception to this might include an irreversibility such as the extinction of a harmful microbe or virus, although there are those who would even argue for their preservation in the interests of future research. An irreversible effect implies that there is no going back, that what has been lost has been lost forever. The extinction of a species, such as the passenger pigeon, is an example of
a biological irreversibility, and natural species are lost yearly. The destruction of a unique ecosystem, which is beyond our technological capacity to recreate, is another. One of the dangers of irreversibilities is that, due to the limitations in our knowledge of the functioning of environmental processes, we often don't know what we have lost until it is too late. Such irreversibilities may result in incalculable, but potentially enormous, economic and social losses, as would be incurred, for example, through the extinction of plants that might potentially produce important pharmaceuticals.
Inherent in the fact that different experts employ competing concepts of sustainability is the diversity of viewpoints prevailing on the issue of intergenerational equity. While there is agreement on the need to include intergenerational considerations in environmental accounting, there still remains considerable dispute over the proper definitions and approaches. For example, what is the extent of our obligation to the next generation? Must we bequeath to them an inheritance of wealth commensurate in size to the legacy our own generation received? As Robert Solow has noted (Solow, 1992; included in this volume), we make trades with posterity: we consume natural resources, but substitute for them by investment in (reproducible) man-made capital and knowledge. We hope that the value of the latter more than makes up for any depletion of the former. For Solow, this "ethic of sustainability" defines each generation's obligation to its successor; if a generation adds to social capital in all its forms amply enough to maintain the aggregate value of all social capital, then it has made sufficient recompense for the nonrenewable resources it has depleted. However, measuring the value of natural resources, particularly as components of complex, interactive ecosystems, presents daunting challenges for such calculations. And many experts, including several in attendance at the workshop, question this conception of intergenerational equity, seeing it as anthropocentric and overly biased toward development now at the expense of conservation for the future.
If environmental economic approaches are to be successful, cooperative research is needed to address not only the conceptual, definitional problems described above, but also to overcome technical challenges in valuing natural assets and to devise suitable methods and practices for accurate environmental accounting. Four principal technical issues were identified at the workshop: monetization, externalities, data unavailability, and data accuracy.
First are the technical challenges of monetization, introduced earlier. To view natural resources as capital, and to incorporate them into economic analyses, economic value must be assigned to them. Many measurement techniques have been proposed, but all are fraught with difficulties. In general, among the many obstacles to assigning a monetary value to a natural resource is the question of what an asset is. If one defines a single component of a natural system as an asset (for example, forest timber), how does one define its role as part of the larger
asset that is the natural system (for example, the forest itself, with its myriad ecological, recreational, and aesthetic functions)? Of course, broadening the asset definition greatly complicates the problem of monetary valuation.
Second, externalities need to be incorporated in economic decisions. Externalities are costs (or benefits) that arise as a result of an economic process, but which are not borne (internalized) by the firm or system engaged in that process as in the case of environmental damage. These costs are generally absorbed by third parties or the public at large. In other words, externalities are costs (or benefits) that appear to be "free" to those controlling the process, but are costly (or valuable) to someone else. What is or is not an externality depends on accounting practices, and on legal and regulatory systems. For instance, if a manufacturing plant does not have to pay to clean up the air pollution resulting from the chemicals emitted from the plant's smoke stacks, then the creation of that pollution is an externality the plant imposes on society. The costs are not reflected anywhere on its books.
Agricultural activities, for example, involve numerous externalities. Farming may have damaging or beneficial effects on the off-farm environment but these are usually not accounted for in financial terms. On the one hand, farms preserve land from commercial development, thereby preserving limited natural ecosystems, and providing scenery of relative beauty. On the other hand, damaging effects may be created by farming, such as the sedimentation of neighboring streams and pollution problems associated with runoff of fertilizer or pesticides.
Many natural resources, although vital inputs to production processes, remain externalities. For example, a company may not treat water used in production as capital to be paid for and maintained, but consider only the costs of having it conveyed to the desired location. Common property resources, such as water and air, have been frequently exploited because of their unrestricted accessibility. In recent years, degradation of air and water has become a matter of public concern, and measures, such as the U.S. Clean Air Act amendments of 1990, have been taken to curb pollution. Deteriorating environmental conditions have also aroused growing concern in such urban centers as São Paulo, Mexico City, and Bangkok.
If a company is required to pay the costs of external damage, or is able to charge for external benefits, these externalities become internalized in the company's financial accounts. But finding ways of bringing externalities into the economy is difficult. Many externalities are not only physically difficult to measure but, even when measured, are troublesome to translate into monetary equivalents. We can force a utility to restrict its emissions through tradable emission permits or other policies, but we cannot accurately compare the costs of these restrictions against the damage to health these emission controls prevent, much less express such health damage in monetary terms.
Compelling individuals and institutions to reduce these external social costs by incurring some internal costs is not sufficient to yield the kind of environmentally sensitive national accounts reformers advocate. Comprehensive accounting reform requires that all the quantitative or qualitative deterioration of such natural capital as air, water, or soils would need to be charged against the nation's GDP. But at least the partial internalization of costs is a step in the direction of forcing some explicit economic manifestation of the effects of an eroding natural capital base. How would this come about? Take an electric generating plant whose degradation of the environment through its SO2 emissions fails to be treated as a charge against GDP, leaving the measure of NDP higher than it ought to be. Assume that the utility installs costly scrubbers
in order to comply with the U.S. 1990 Clean Air Act amendments mandating limits on SO2 releases. In an economy-wide sense (assuming full employment of resources), a given volume of electricity production requires shifting resources into the production of scrubbers at the expense of, say, widgets, whose decline signifies a level of GDP lower than would otherwise be the case.
Thus, even in the absence of a formal recasting of the national accounts along environmental lines—which would have dictated quantifying the deterioration of air quality prior to the installation of scrubbers and its improvement thereafter—the GDP penalty in this example at least alerts us to the economic implications of environmental protection. A similar point emerges from Crosson's paper (in this volume). Losses of soil quality which reduce farm production should be reflected in the national accounts. But if farmers compensate for soil quality losses by changes in farm practices entailing an increase in operating expenses, the soil-loss effect will show up as lower net farm income and GDP, assuming a full employment economy.
A third problem is the difficulty of characterizing and quantifying the natural environment. Even when there are applicable units of measurement, and when their rate of use can be charted, natural resource features are difficult to characterize. For example, it would be extremely difficult to calculate how much of a given resource exists within the environment, especially for resources such as air and water that are constantly depleted and created and used in different forms within different natural cycles.
This challenge exists even for so discrete a problem as determining animal populations, for which there are few accurate measures. Even in the case of species on the verge of extinction, whose numbers are so small that wildlife enforcement officials know individual animals, the exact population of the species may be unknown. Computing meaningful subsets of an animal population, such as the number of species within any given ecosystem, has aroused even more scientific controversy. Biologic inventories are necessary for proper valuation of natural resources, and some steps are being taken in this direction. While biologists have expressed Skepticism over the possibility of assigning a monetary value to biological entities, they have supported the creation and maintenance of biologic inventories as useful and valuable for research. The Department of the Interior is in the process of forming a new agency, the National Biological Survey, whose functions will include the inventorying, mapping and monitoring of biologic resources. Comparable efforts in other parts of the world are rare.
Physical accessibility to sources of key data poses another problem for information collection. Terrain and topography may prove difficult to traverse, unstable political situations may hamper research, and the costs involved in data collection may be prohibitive. For example, forest inventories are very expensive and very rare in developing countries, and few such inventories are performed. Because of such difficulties, relatively little data has been collected or research conducted. While remote sensing methods, such as Landsat, can contribute information about overall resource size, they may not be able to capture information about resource quality. For example, trends in total forest acreage can be obtained, but changes in forest density, ratios of new growth to old growth, or decline in health of trees are difficult to assess at present.
Even for simple inventories, much uncertainty is associated with the measurement process. When attempts are made to combine measurements, added levels of uncertainty are
introduced. Macroeconomic frameworks and indicators such as the national accounts deal with aggregates, and aggregation necessarily requires additive monetary units derived from many subsidiary measurements. In combining these figures, uncertainties associated with the measurements are also combined, in ways which are difficult to track. The cumulative or additive effect of these uncertainties, and the loss of information through aggregation, can only be guessed at. This opens the question of the extent to which the aggregate figures produced can be trusted as a basis for policymaking. The explicit inclusion of uncertainty figures as a major component of natural capital accounting frameworks may be one way to deal with this problem. Alternatively, a probabilistic approach for calculating uncertainties could be employed, with an aggregation or totaling of uncertainties implying a corresponding totaling of probabilities.
Finally, accuracy issues involved in measurement processes go beyond simple technological inadequacy. Even when we are able to gather data on a certain phenomenon, if these data do not fit earlier experience or preconceived ideas, they may be "filtered out" of our data bases. We measure what we are looking for and what we expect to find. We do not measure what we are not looking for (do not value in a scientific sense), and we do not always know how to deal with what we don't expect to find. We sometimes write off these "unexpected" results as "errors" or ''noise." For example, as Craig and Glasser have noted (in this volume), data revealing the Antarctic ozone hole were systematically removed from satellite data through numerical filters designed to remove data that were too far from expectation. Many other examples of this can be seen throughout the history of science. Consider, for example, the elaborate mathematical structures designed to reconcile observed planetary orbits with the socially constructed concept that the Earth was the center of the universe.
In addition to, and partially because of, the above-mentioned difficulties, there is much disagreement among experts as to the proper measurement procedures, and as to the method of incorporating these physical measurements into economic frameworks. For example, although the Netherlands has been one of the more active promoters of environmental accounting reforms, Dutch statisticians are wary of moving too aggressively in monetizing environmental resources or damage within economic frameworks, arguing that physical measurements and monetary quantities are too dissimilar to be validly compared (The Economist, 1993). Yet other statisticians maintain that without such comparisons, fundamental accounting reforms and related efforts such as cost-benefit analyses cannot be undertaken.
As noted above, all economic valuation techniques involve aggregation. Aggregation requires comparable units for the quantities being combined, giving rise to the problems mentioned earlier. Thus, experts debate the validity of assigning additive monetary values to such widely varying elements as air quality, species population, and ecosystem value. Some scholars assert that no legitimate comparison or aggregation is possible.
Poorly defined property rights or open access will depress the price of a resource, and such underpricing of a resource means it will tend to get wasted. This point is borne out by both contemporary and historical experience. In some developing areas, ill-defined or poorly
enforced property rights (such as in the frontier regions of Amazonia) may encourage destructive and unsustainable land-use practices, inimical to the long-term environmental services the land could provide. Similarly, nineteenth century U.S. resource policy distributed land, water, mineral rights, and timber resources to private individuals for negligible prices, diligence requirements, and lease terms, which promoted rapid growth, but also encouraged wasteful, inefficient and frequently destructive mining, lumbering and farming practices. These are all examples of "externalities," as discussed earlier.
While the problems involved in measurement and integration of natural resources into economic accounts are difficult to grapple with, the thorniest issues confronting environmental accounting are philosophical ones. Ideological, moral, and even metaphysical values are embedded in the questions considered here. These ethical dilemmas cannot be avoided, nor would it be advisable to proceed with valuation procedures without examining the underlying conceptual concerns involved.
Another issue central to both environmental and economic valuations is the common tradeoff between short term economic growth and poverty alleviation on one hand, and long term sustainable development, on the other. It is difficult to calculate social values for resources that are not traded in conventional markets, for example, "scenery." Scenery is not bought and sold, though it does have economic impact—just ask a real estate developer or tourist industry employee. Yet, despite such obvious social and even economic impacts, it is often difficult to assign an economic value to such an intangible quality of the environment.
In the absence of public preferences as revealed in markets, valuation can be attempted through alternative approaches, such as "contingent valuation." This technique is based on assessing what people would be willing to pay in order to preserve a given resource. Such "willingness to pay" estimates are derived from interviews and questionnaires. Different users of a resource naturally differ in how much they are willing to pay to preserve it. Even individuals who have no direct contact, and expect to have none, with a given natural resource, are frequently willing to pay to preserve it for a variety of reasons ranging from patriotism to altruism. Certain unique natural features such as those found in the National Park System, for example, inspire such a public response.
Concomitant with willingness to pay is the flip-side of the concept, that is, "willingness to accept." This is basically a question of what people are willing to accept as compensation for environmental damage, that is, what compensation they would demand as recompense for the destruction of a particular natural resource. The advantages of such a measurement technique are that it can be used to calculate values for things which have been deemed invaluable. Through questionnaires eliciting willingness to pay or accept, numbers are produced which can be used in value calculations. The problems with such measurements, however, are associated with the highly subjective nature of the data source and the fact that the value can change dramatically over time and between one community and another. What a given individual answers to such a question may be very circumstantial and contextual. Opinions may vary over time, with social circumstances, and with education; even when they can be pin-pointed through questionnaires, they may change rapidly. Additionally, the phrasing of the question, even its placement within the questionnaire, may strongly affect the outcome.
Another problem arising here is the issue of different groups of users with different sets of values. A rainforest, for example, may be valued very differently by vacationing foreign
tourists than by local villagers fighting for subsistence. Also, since tourists are usually foreigners, a money valuation derived from willingness-to-pay methods will have to be calibrated for different currencies, and relative wealth. Setting dollar amounts for groups of people of different economic backgrounds may be difficult. For a variety of reasons such as these, contingent valuation and related survey techniques are sometimes viewed with distrust.
Furthermore, some people are uncomfortable with monetizing natural assets held in common, such as oceans or national parks. They reject such valuations as morally repugnant or anthropocentric. In fact, some physical scientists suggest that environmental accounting is a flawed concept because it deals with issues, such as preservation of wetlands, from the point of view of affected humans. Environmental accounting might allow us to consider the value of a wetland to bird watchers as well as developers, but cannot allow us to consider wetlands preservation from say, the "perspective" of hundreds of thousands of migratory birds whose feeding ground along a major flyway is threatened by human activity.
ISSUES FOR FURTHER STUDY
Seven issues were identified at the workshop for urgent attention and further study: criteria for selecting appropriate natural capital assets for economic valuation; actually selecting them; developing techniques for environmental accounting appropriate for national policymaking, firm-level decisionmaking and complex dynamic systems analysis; developing nonmonetized inventories for selected natural assets; tracking irreversible processes; identifying points of leverage for future action; and, finally, determining priorities for revising the national income accounts and other economic indices versus alternative means for characterizing the broader societal consequences of economic decisions with significant resource depletion and environmental degradation impacts.
A crucial first step is selecting particular natural resource or environmental assets for prospective incorporation into economic indices, or for consideration in some other fashion appropriate for policymaking. Clearly, both the technical challenges and the policy implications of environmental accounting will vary if the initial work concentrates on resources such as mineral stocks for which monetary values are easily derived and stocks are fairly well known, rather than on "free goods" such as air and water, or interactive processes such as maintenance of biodiversity or wetlands preservation.
Second, a feasible early step is to experimentally incorporate measurements for one particular resource for which ample data exist, and the value of which may even be determined from existing markets. The natural resource that most closely fits this description is the nation's mineral wealth. Mineral and ore deposits have been well surveyed and inventoried since the establishment of the U.S. Geological Survey and the Bureau of Mines, and are being exploited by a relatively robust and mature industrial sector. Although there are problems in assessing the present value of prospective commercialized resource flows, the monetary values of many minerals can be derived from existing markets, and need not be subject to serious dispute. Economic issues regarding extractive industries and their sustainability have been extensively explored. There are established cases of substitutability between different minerals in different industries, although the elasticity of substitution between classes of rare and common minerals
is debatable and varies over time as a function of technology and societal need. Selected minerals (excluding those like the precious metals gold and silver whose prices have been subject to wide swings over time) would thus seem to be an ideal initial resource to be incorporated into an accounting framework.
Indeed, at the Woods Hole workshop, Carol Carson, Director of the Commerce Department's Bureau of Economic Analysis (BEA), reported that the Bureau had already begun such calculations. In the April 1994 issue of its Survey of Current Business, the BEA recounts the results of its initial efforts along this line. However, while the use and depletion of mineral stocks might offer the readiest initial opportunity for revising the national accounts, many scientists and policymakers, including a number of workshop participants, have expressed reservations about such an approach. They cautioned that the results of calculations which included only minerals would present a distorted view, and hamper efforts to achieve a more comprehensive valuation of natural capital flows. The most urgent arena for implementing environmental accounting, these experts maintain, is precisely those areas for which no clearly defined markets and price structures exist. They suggest immediate work on essentially nonmarket goods—such as air and water—and on interactive processes—such as maintenance of biodiversity. This line of argument, then, demands primary attention be given not just to those areas most easily amenable to monetization, but to assets not easily monetized or even measured.
Even after targeting priority considerations for revised environmental accounting procedures, considerable technical challenges remain in devising environmental valuations. Techniques are needed to update the national income accounts or other official economic indicators so that national debate and policy decisions accurately reflect the importance of natural assets. The nation requires continual improvement in methodologies by which people can assess these benefits and damages. New systems of measurement must remain flexible and dynamic so that they can cope with the constant new problems posed by measuring complex, dynamic, interactive natural systems.
There is considerable disagreement about whether assigning monetary values to any, many, most, or all natural processes or assets is wise or appropriate. Some workshop participants believe that the research community needs to develop alternate tools, such as nonmonetized inventories of natural assets, in order to equip policymakers with accurate data in areas where monetary valuation is either inappropriate or unavailable.
Two broad conclusions emerged from these intense and at time contentious workshop discussions. First, that incorporating environmental and natural resource values into economic recordkeeping and policymaking is an arduous task, fraught with technical and conceptual difficulties. It will require considerable ingenuity and effort to solve these problems. Future directions in the development of environmental accounting include the identification of productive avenues of research, the promotion of innovative and significant work, and the furtherance of interdisciplinary communication and cooperation, all of which should be encouraged.
The second conclusion was that, despite the difficulties, efforts must continue. Although it is probably premature to alter the official national accounts, attempts should be made to experiment with new environmental accounting methodologies as they are developed and refined. These revised numbers can then be disseminated side-by-side with the conventional accounts.
National policymakers, and the data sources on which they base their decisions, must soon begin to take into account resource depletion, ecological balance, and the sustainability of economic development. The nation's future economic strength and environmental integrity depend on it.
Adams, Robert McC. 1993. Smithsonian Horizons: Secretary Adams' comments. Smithsonian 24:6.
Congressional Budget Office. 1994. Greening the National Accounts, U.S. Congress. Washington, D.C.
Daly, H. E. 1990. Toward Some Operational Principles of Sustainable Development. Ecological Economics 2(1):1-6.
Environmental Accounting for Sustainable Development. UNEP-World Bank Symposium. 1989.
Environment and the Economy: A Special Report. 1993. Science 260:1883-1909.
Integrated Environmental and Economic Accounting. 1993. United Nations, Department for Economic and Social Information and Policy Analysis, Statistical Division, Studies in Methods. Series F, No. 61.
Lutz, E., ed. 1993. Toward Improved Accounting for the Environment, UNSTAT-World Bank Symposium. Washington, D.C.
National Bureau of Economic Research. 1972. The Measurement of Economic and Social Performance. Washington, D.C.: National Bureau of Economic Research.
Repetto, R., W. Magrath, M. Wells, C. Beer, and F. Rossini. 1989. Wasting Assets: Natural Resources in the National Income Accounts. Washington, D.C.: World Resources Institute.
Solorzano, R., R. de Camino, R. Woodward, J. Tosi, V. Watson, A. Vasquez, C. Villalobos, J. Jimenez, R. Repetto, W. Cruz. Accounts Overdue: Natural Resource Depreciation in Costa Rica. Washington, D.C.: World Resources Institute.
The Economic Accounts of the United States: Retrospect and Prospect. 1971. Anniversary Issue of the Survey of Current Business 51:7. Part II.
The Price of Everything, the Value of Nothing. The Economist. July 31, 1993. Pp. 63.
World Commission on Environment and Development (''Brundtland Commission"). Our Common Future. 1987. New York: Oxford University Press.