The National Income and Product Accounts: History and Application to the Environment
Natural and social scientists concerned about natural resources and the environment have endeavored to take the measure of nature. Measures used for this purpose range from those used to monitor the state of major environmental indicators, such as air and water quality, to analytical measures of major environmental aggregates. For the most part, however, measures of the economic contribution of natural resources and the environment have lagged behind physical measures. The slow development of economic measures is due to two major factors. First, economic accounts generally record and measure activities that pass through the marketplace, while most of the activities that raise environmental concerns—from air pollution to appreciation of pristine wildernesses—take place outside the market. Second, the paucity of data and difficulties of valuation for most environmentally related activities make constructing economic measures much more difficult than is the case for market-related activities. The end result is that most nations produce detailed national economic accounts accompanied by vast quantities of useful data for market-related activities and little or no comparable data for nonmarket environmental activities.
The intuitive idea behind the desire to broaden the U.S. national accounts is straightforward. Natural resources such as petroleum, minerals, clean water, and fertile soils are assets of the economy in much the same way as are computers, homes, and trucks. An important part of the
economic picture is therefore missing if natural assets are omitted in creating the national balance sheet. Likewise, consuming stocks of valuable subsoil assets such as fossil fuels or water or cutting first-growth forests is just as much a drawdown on the national wealth as is consuming aboveground stocks of wheat, cutting commercially managed forests, or driving a truck.
History of Augmented Accounting
From the perspective of environmental accounting, the key point to recognize is that gross domestic product (GDP) is conceptually defined to include only the final output of marketed goods and services, that is, goods and services that are bought and sold in market transactions. This point is clearly stated in a comprehensive discussion of the National Income and Product Accounts (NIPA): "... the basic criterion used for distinguishing an activity as economic production is whether it is reflected in the sales and purchase transactions of the market economy" (U.S. Department of Commerce, 1954).
There are, however, important exceptions to this basing of the NIPA on market transactions. One is the exclusion of illegal activities such as drugs, prostitution, and illegal gambling; thus GDP will rise as gambling moves into the legal market sector. In addition, there are imputations for near-market services that are not recorded in market transactions. For example, there is an imputation for the services of owner-occupied housing so that these services can be included in output and income as are the rent and output associated with rental housing. There is also an imputation for the fuel and food produced on farms and consumed by the farmers themselves. A further large imputation is made for banking and other financial services furnished by financial businesses below cost in lieu of interest payments.
The key issue involved in environmental and other augmented accounts is whether to broaden the above boundaries and if so, how and how far. It has long been recognized that drawing the line at the limits of the market distorts the value of the NIPA as a measure of economic activities and well-being (see also Chapter 1). There is a vast and changing amount of productive nonmarket activity that produces goods and services quite similar to those produced in the marketplace. Commercial laundry services are reckoned as part of GDP, while parents' laundry services are not; the value of downhill skiing at a ski area is captured by GDP, while the value of cross-country skiing in a national park is not.
At the same time, while recognizing the importance of considering
alternative measures, it is essential to retain the conventional market-based accounts as a central component of our national accounts. These core accounts are of great importance for purposes of historical and international comparison and will continue to be a critical indicator for economic policy making. The objective of augmented accounting is not to replace the core accounts with a preferred new bottom line; rather, the emphasis is on developing alternative approaches and measures that can illuminate the diverse dimensions of economic activity.
Work on augmented accounting by official statistical agencies, as well as by individual scholars, has provided estimates for a wide variety of nonmarket activities for experimental augmented national accounts (see Eisner, 1988, for a comprehensive review of augmented accounting). Beyond the environmental arena, which is reviewed in the next section, significant examples of work to extend the accounts include the following areas:
- The value of home production and unpaid work
- The value of the services of consumer durables (similar to the imputation of services of owner-occupied housing)
- The value of research-and-development capital
- The value of leisure time
- The value of informal and home education
This work on extending the accounts is motivated by the idea that expanding the boundaries of the accounts would provide a better estimate of the size, distribution, and growth of economic activity and economic welfare than that offered by the current accounts.1
In revising and extending the U.S. NIPA, the Bureau of Economic Analysis (BEA) is following guidelines suggested by the internationally formulated and recommended U.N. System of National Accounts (SNA) (Parker, 1996 and 1991). These efforts have entailed both modifications in core measures, such as the introduction of separate current and capital accounts for government, and the development of satellite accounts, such as for research and development. Satellite accounts (sometimes referred to as supplemental accounts) expand the analytical capacity of the national accounts without overburdening them or interfering with their general orientation. Because they supplement rather than replace the core accounts, they can serve as a laboratory for experimentation and provide a means for applying alternative approaches and new methodologies.
The guiding principle in developing augmented accounts is to measure as much economic activity as is feasible, regardless of whether that activity is of a market or nonmarket nature. The goal is to achieve a better measure of final output—of what consumers in the United States currently enjoy in the way of goods and services, and of the accumulation of capital, of all kinds, that will permit the future production of goods and services.
In terms of current consumption, augmented output includes not merely what consumers buy in stores, but also what they produce for themselves at home; the government services they ''buy" with their taxes; and the flow of services that are produced by environmental capital such as forests, national parks, and ocean fisheries. The need to include nonmarket components arises because of the tradeoffs between market and nonmarket activity. For example, parents produce more in the market when they go to work, but they also have less time at home for child care and domestic services. Likewise, the resources used to provide government services that add to real consumption may reduce the quantity of services provided by businesses. In the environmental area, resources devoted to removing lead from gasoline and paint will lower conventionally measured consumption, but will raise the nation's human capital by protecting children from brain damage and other debilitating illnesses.
Similar issues arise in the measurement of national saving and investment. Conventional NIPA saving and investment measures include only tangible investments in plant, equipment, and inventories. This conventional picture omits the much larger intangible and human investments in education, training, research and development, health, and the environment. A complete set of accounts would entail full integration of comprehensive investment flows with comprehensive capital or wealth accounts. These accounts would then relate not only to current production of goods, but also to changes in the value of human capital; to the accumulation of knowledge and technical advances; and to the improvement or deterioration of the basic environmental capital of land, water, and air. Development of a complete set of capital accounts would thus give the nation a much more complete picture of how well the current generation is performing in its role as trustees of the nation's tangible, human, and natural resources.
Comprehensive accounts and environmental accounting provide information that can help governments set sound economic and social policies and aid the private sector in making productive investments. An important example is use of pollution abatement costs to estimate the impacts of regulation on productivity and output growth. Studies by Denison (1979) and by Jorgenson and Wilcoxen (1990) have provided valuable information on the relative importance of regulation, pollution
control expenditures, and other factors in the slowdown of productivity growth in the United States after 1973. In addition, these data have been crucial inputs to studies of the cost of air pollution regulation and the benefits and costs of controlling air pollution conducted by the U.S. Environmental Protection Agency.
Two major issues arise in the design of augmented accounts. The first is where to draw the line when extending the accounts beyond the boundary of market transactions. The dilemma is similar to that faced by the little boy who said, "I know how to spell banana, but I don't know where to stop." Should the line be drawn at near-market activities—for example, home-cooked hamburgers, depletion of oil and timber resources, fish caught and consumed by anglers, and services of consumer capital such as automobiles and washing machines? Or should the accounts extend to all private goods, such as educational investments and the value of visits to Yellowstone National Park? Should the accounts attempt to measure the value of leisure time? Should they extend to include public goods such as the value of clean air and clear water? Should they include international concerns such as the damages from ozone depletion and global warming? These thorny questions are taken up later in this report, but we note here that they are pervasive in the design of augmented accounts.
The second major issue is how to measure nonmarket activities. Measurement involves collecting data that will support estimates of both quantities and prices. While data on market activities are often costly to collect, for the most part the elemental data exist in the form of individual transactions in which someone buys a banana, a computer, or a haircut—transactions that are generally recorded. Nonmarket activities pose difficulties because the physical activities involved are generally not recorded, and there are no objective records of the valuations. An example of the difficulty is a consumption service such as swimming in the Atlantic Ocean. No one records how many times Americans actually swim in the Atlantic Ocean in a given year. More difficult is the valuation of swimming: since swimming is generally free, except in congested areas, we do not know how to value the swims. There are numerous techniques available for estimating both the quantity and value of nonmarket activities such as swimming, but they almost always require gathering additional data and involve complex imputations of value where no market data are available.
We must not, however, forsake what is relevant and important merely because it presents new problems and difficulties. The economic light is brightest under the lamppost of the market, but neither drunks nor statisticians should confine their search there. In extending the accounts, we must endeavor to find dimly lit information outside our old boundaries of search, particularly when the activities are of great value to the nation.
Developments at the Bureau of Economic Analysis
Over the last decade, BEA has taken a number of important steps in extending the core economic accounts and developing satellite and supplemental accounts (BEA, 1995a). Among the most important developments in the core accounts are the following:
- Improved measures of price and output. BEA has pioneered the use of improved measures of price and output, including the use of chain-weighted price and output indexes. These measures allow more accurate tracking of inflation and output than did earlier fixed-weight measures. This work has demonstrated that these state-of-the-art concepts can be implemented routinely in national statistical measures.
- Improved investment accounts. BEA has moved to broaden the U.S. investment accounts in line with international standards by introducing estimates of government investment and capital and improving the estimates of depreciation and capital stocks.
- Improved international accounts. Recognizing the growing importance of services in the nation's economy, BEA has incorporated new information on international trade in services and revised estimates of foreign direct investment.
In its efforts to improve the national economic accounts, BEA has been proceeding in a prudent and conservative fashion, employing proven and consistent techniques. In its core national accounts, BEA employs the concept of Hicksian income (see Appendix A). Such production-based measures of income and output are useful for delineating market activity and should continue to form the basis of the core national accounts.
Market-based concepts are inadequate, however, for tracking the entire range of economic activity, market and nonmarket. The purposes of augmented accounting are to provide more comprehensive measures of output, saving, and investment; to ensure that the accounts treat economic activity in a consistent way when the boundaries between market and nonmarket activities change; and to provide information on the interaction between the economy and the environment so that natural and environmental resources can be more effectively managed and regulated.
Importance of Environmental and Natural-Resource Accounting
Environmental and natural-resource accounting has emerged over the last three decades in response to increasing awareness of the interac-
tion between the natural environment and economic activity. Growing concerns about resource scarcity were reinforced by the dramatic increases in energy and mineral prices of the 1970s. Many began to worry that the nation was rapidly depleting its precious stocks of subsoil assets. Further awareness resulted from documentation of the economic and social costs of environmental degradation and pollution in terms of human health and property values, reinforced by pictures of rivers and lakes on fire and serious oil spills.
A set of well-designed environmental accounts could overcome the shortcomings of the current market-based accounts. Indeed, the construction of environmental accounts is one element of the more general task of developing a set of comprehensive economic accounts that includes both market and nonmarket economic activity. This section reviews the primary shortcomings of the current accounts and explains why a well-constructed set of comprehensive accounts would have significant economic value to the nation.
Deficiencies of Current National Accounts
Efforts to develop alternative accounting approaches to supplement the standard market accounts with measures of changes in consumption and investment in natural resources and the environment have been undertaken in response to three perceived deficiencies in the way the conventional accounts treat natural resources and the environment.
First, as an indicator of economic well-being, the accounts sometimes behave perversely with respect to environmental degradation and changing stocks of natural resources. For example, cutting down the nation's dwindling redwood forests increases GDP, yet no account is taken of the loss of this precious asset because the nation's forests have not been included in the asset accounts. For similar reasons, when fishing activities increase the harvest of cod or halibut, the national accounts record the increased production and consumption, but omit the decline in breeding stocks and the costs imposed on future producers and consumers. And pollution abatement expenditures increase measured output—even when such expenditures serve only to offset environmental deterioration, and there is no net increase in current or future consumption. In these and many other examples, changes in production do not reflect genuine changes in economic well-being and may even result in economic harm or cost in the future.
Second, the standard national accounts are inconsistent in their treatment of different forms of wealth. For example, the NIPA include a full set of accounts of gross investment, net investment, depreciation, and the capital stock for produced, tangible producer capital. In contrast, natural
capital—such as oil and gas deposits, forests, soils, and underground aquifers—is largely omitted from the accounts. When a commercially grown tree is cut, the production cost of the tree is counted as a cost of production, but when a first-growth national forest is clear-cut, there is no parallel subtraction. As a factory ages, this is counted as a depreciation charge, but the accounts are not charged when an oil deposit is exhausted. Likewise, the national accounts nowhere reflect the occurrence of widespread deterioration or improvement in the quality of environmental assets such as air and water. The distinction between gross and net investment for reproducible capital is justified on the grounds that those investments which simply replace depreciated stock add nothing to economic well-being and that failure to subtract depreciation would yield income measures that might be unsustainable in the long run. The logic of this argument is equally applicable to environmental and natural capital.
The third and perhaps most important deficiency of the conventional national accounts is that they give a very incomplete picture of the full scope of economic activity. By focusing only on marketed outputs and factors of production, the conventional accounts neglect a large number of economically significant inputs and outputs that are not bought and sold in markets. In the environmental area, these nonmarketed inputs and outputs often include the free goods and services provided by environmental assets such as air, water, forests, and complex ecosystems. Many of these assets—such as recreational sites in Yellowstone Park, stocks of underground water and flows of river water in the Southwest, and public parks in New York City—are limited or fixed in supply. Thus they have economic scarcity value even though they may lack market prices. Because the conventional accounts omit such economically valuable but nonmarketed goods and services, they overstate the role of market inputs and outputs in economic welfare. They also fail to provide business, citizens, and policy makers with the full and accurate assessment of the state of economic activity that is needed for economic policy and rational environmental management.
Finally, it should be emphasized that the current NIPA do not focus on a conceptually appropriate definition of market national income and output. The most appropriate measure of national output in the core accounts today is real net national product, which measures the total net output and income accruing to residents of the United States, corrected for inflation. This differs from the measure currently emphasized, real GDP, in two ways. First, GDP includes depreciation or capital consumption, which exaggerates sustainable income by including in national income a sum that cannot sustainably be consumed. Traditionally, output measures have emphasized gross rather than net product because depreciation is difficult to measure accurately. Second, GDP excludes the net
factor earnings abroad of domestic residents, which is included in national product. Inclusion of net factor earnings abroad is desirable if output is designed to measure the sustainable consumption of the nation. The recent switch in emphasis from national to domestic product occurred because domestic product is more closely related to domestic output and employment. While the emphasis on GDP rather than net national product is understandable, the panel emphasizes that the latter is conceptually preferable as a measure of sustainable income.
Value of a Comprehensive Set of Accounts: Scorekeeping and Management
Economic accounting—whether it be business accounting or the accounting of a nation's economic activity—traditionally serves two major functions: it offers a way to track the economic performance of a business or a nation, and it provides an organized body of economic data that enhances the ability of an organization or a nation to manage its economic affairs. The principal reason for growing interest in natural-resource and environmental accounting is the belief that improved accounting for the contribution of natural capital will enhance the ability of the conventional accounts to serve both of these functions.
The NIPA are the major way nations keep score on overall, regional, or sectoral economic performance, past and present. The core accounts include production measures such as gross national product (GNP) and GDP, along with data measuring market incomes and a broad array of sectoral data. These core accounts are widely used to gauge a nation's economic performance over time and to compare economic performance among nations; they are an essential tool for assessing the state of the economy and formulating macroeconomic stabilization policy. For example, economic research has shown a close link between movements in GDP and changes in the unemployment rate, changes in tax revenues, and the federal budget deficit. Understanding the economy requires comparing current trends and movements in national output with those of various historical periods in order to forecast the future. This scorekeeping function of the national accounts is widely accepted in spite of many deficiencies in the measures of prices and outputs and the numerous interpretative problems introduced because the core accounts are limited to market transactions (see Hicks, 1940; see also Kuznets, 1948a, 1948b).
As discussed above, measures of augmented national income and product endeavor to extend the purview of the economic accounts by including a broader set of consumption, investment, and income measures. These augmented accounts give a more balanced view of the trends
of overall economic activity and provide more accurate estimates of trends in income, saving, and investment. More comprehensive systems that account for negative outputs such as pollution and positive outputs such as outdoor recreation will yield more meaningful indicators of economic performance. One valuable contribution of well-designed comprehensive measures is that they can eliminate anomalies in the national accounts. For example, according to conventional measures of economic performance, oil spills and earthquakes often raise GDP and appear to make the nation better off. Such anomalies would be redressed by appropriate accounting measures. Similarly, improved measures would correct the anomaly that a nation with abundant natural parks and recreational opportunities provided freely to its citizens appears to be worse off than a nation that provides recreation only through commercial theme parks.
A final set of scorekeeping measures relates to what is called "sustainable income." These measures address the question of whether the nation is setting aside sufficient tangible and intangible capital and new technological knowledge to ensure that future generations will have an adequate standard of living. Sustainable national income is defined as the maximum amount a nation can consume while ensuring that future generations will have living standards at least as high as those of the current generation. It turns out that ideal measures of sustainable income are closely related to current national income and product measures. Techniques for measuring sustainable income are discussed later in this section and in Appendix A.
The second function of economic accounting—providing data needed to manage economic activities—requires gathering a systematic record of all the inputs and outputs that characterize an economic system. The management function of accounts and budgets is widely used by both businesses and governments. While scorekeeping indices may tell a business whether it is profitable, details of accounts and budgets are necessary to help the business make better decisions and improve its profitability. Similarly, while government budgets are valuable summary indicators of the overall importance of the government sector in the economy and of government's net contribution to national saving, the most important function of the budgetary accounts is to help Congress and the executive branch direct the day-to-day operations of the federal government and the allocation of federal resources.
The detailed information embodied in the national economic accounts serves a similar function in the management of national economic policy. Even in countries such as the United States that have strong laissez-faire traditions, the economic accounts are an essential input to major economic policy and forecasting models that influence fiscal and monetary
policy. At a more detailed level, the data help businesses track their own sectors and forecast their sales and profits, and are useful for a wide variety of economic activities.
Unfortunately, the conventional economic accounts are sometimes deficient for management purposes because of their omission of those inputs and outputs that are not traded in the marketplace. Resource and environmental accounting expands the list of inputs and outputs so that policy makers are in a much better position to develop and analyze policies, especially those that involve interactions between the natural environment and the market economy. Economic accounts expanded to include resource and environmental activity are especially useful for the analysis of major environmental policies and programs that may affect large segments of the economy, such as those related to water allocation or global warming, or for the analysis of nonenvironmental programs that may have substantial environmental consequences, such as interstate highway programs. Without a comprehensive environmental and nonmarket accounting framework, each policy analysis requires data collection de novo. As a result, analyses of environmental programs today are extremely expensive and inefficient. A system of resource and environmental accounts linked with the conventional economic accounts can provide the inputs for a wide variety of policy analyses at relatively low incremental cost.
The remainder of this section describes in detail the primary benefits of a comprehensive set of accounts.
Comprehensive Accounts Give a Complete Picture of Economic Activity
At the most general level, comprehensive economic accounts provide a complete reckoning of economic activity, whether it takes place inside or outside the boundary of the marketplace. As suggested above, economic decision makers need to understand more than the conditions of the marketplace if they are to make sound decisions. Business clearly need and want to know about basic economic conditions in the world, the nation, their region, and their industry. Without such information, firms are flying blind. They run the risk of continuing unsustainable programs to the point of serious decline or even bankruptcy. States and localities similarly require comprehensive accounts of economic activity. Such comprehensive accounts need to include natural-resource and environmental accounting. A firm will pause, for example, before building a plant for which a fuel that is running into short supply would be required or locating in a region whose water supplies are severely limited. Companies may want to build in areas that have many amenities and high environ-
mental quality—conditions that result both from market investments by the public and private sectors and from well-maintained natural capital. Managers and stockholders want to know the unpriced environmental costs of their actions because society may eventually make them pay those costs. These kinds of questions are equally vital for states, localities, and foreign investors, as well as for individuals who face personal investment or locational decisions. In short, environmental accounts are an important tool for providing the information necessary to track economic conditions and make sound decisions.
Limiting the national accounts to market sectors can produce misleading information on overall economic trends. One important example is standard measures of national saving and investment, which include only investment in tangible capital such as factories, equipment, inventories, and houses. By omitting market and nonmarket investments in intangible and human capital, the current national economic accounts can underestimate national saving by almost 500 percent.2 Another example of misleading signals is in the treatment of the movement of people from unpaid to paid work. Because the unpaid work is not counted in national product while the paid work is, measured national output rises more than the actual amount of total national output of goods and services as labor force participation rises.
Similar issues arise within the natural-resource and environmental sectors. When companies discover large deposits of oil, gold, and other mineral assets, these deposits are not counted among the nation's investments or as an increase in its stock of assets. Similarly, although forests contribute greatly to the nation's well-being, only the timber value of forests is counted as part of the official national output. The value of hunting, fishing, and other forms of nonmarket forest recreation likewise is not counted as part of the national output, even though the total economic contribution of these nonmarket outputs probably exceeds the value of timber production.
The largest distortion in the environmental area probably arises in those sectors related to environmental quality. Economic studies reviewed in Chapter 4 of this report indicate that the nation is devoting more than $100 billion annually to pollution abatement and control expenditures. Yet virtually all the economic benefits from these expenditures are omitted from the national accounts. Even though investments in
clear air and water produce benefits in improved health of the population, improved functioning of ecosystems, enhanced recreational opportunities, and lower property damages, a large share of these benefits is not likely to be captured by the current market-based accounts.
Finally, studies indicate that extending the accounts to nonmarket consumption and investment would have a significant impact on estimates of income, product, and national wealth. Preliminary work on augmented accounting exclusive of the environment indicates that broadening the accounts to include comprehensive consumption and investment could easily double the reported net income and output and might increase reported net investment by a large factor.3 Similarly, as discussed in Chapter 4, corrections for environmental flows, particularly those involving nonmarket impacts on the health and safety of the population, could have major impacts on measured income.
Comprehensive Accounts Provide Much Useful Information for Public Policy and Private Decision Making
Environmental accounts would provide useful information for managing the nation's assets and for improving regulatory decisions. For example, improved natural-resource and environmental accounts can provide useful information on natural assets under federal management. Better information on the value of minerals on federal lands would be useful in determining appropriate royalty rates and leasing policies for resources not allocated through competitive auctions. Better information on the stumpage value of timber in national forests would be useful not only for accounting purposes, but also for better management of these forests and for decision making on the balance among timber harvesting, wilderness preservation, recreation, and other uses. Efforts to prevent overfishing have been hamstrung by the lack of reliable information on changes in fish stocks. In many of these cases, data are already collected by federal agencies, but incorporating these data into the consistent frame-
work of a set of national accounts would help regularize their collection and ensure consistency over time and across sectors.
In the case of environmental resources such as air and water quality, a comprehensive set of environmental accounts would provide useful information on the economic returns the nation is reaping from its environmental investments. The contrast between private and public investments is instructive. When a private company makes an investment in an automobile factory or a power plant, company accounts can be used to estimate the economic costs and benefits of that investment. Yet although the nation has allocated more than $1 trillion to environmental, health, and safety investments over the last quarter-century, there is no comparable set of accounts by which to reckon the nation's returns to those investments. Improved environmental accounts would also provide essential information for sound benefit-cost analyses in regulatory decision making. One of the most serious weaknesses in the U.S. environmental database is the lack of comprehensive and reliable data on actual human exposures to major pollutants. Better information on physical emission trends, human exposures, and the economic impacts and damages from air and water pollution would be valuable for expanded accounting measures of productivity. Hence, both the underlying information and the aggregate dollar values in environmental accounts would provide essential information for ensuring that our environmental regulations pass an appropriate cost-benefit test.
Investing in Comprehensive Accounts Would Yield a High Economic Return for the Nation
The federal government currently makes a substantial investment in collecting, analyzing, and distributing statistical data on the nation's economy. This information is valuable in part simply because we are curious about ourselves as a nation. We want to know what we are producing and consuming, to compare ourselves with other nations, and to assess the trends in economic activities. But provision of statistical data is also an investment in a public good. Having more complete, accurate, and timely data on economic activity requires the resources and data-collection abilities of the government. Data for economic accounts will not be provided by the private sector both because the private sector does not have access to the full range of administrative data available to government agencies and because there is little private economic profit in gathering and providing comprehensive economic accounts. Yet while the federal government invests heavily in the collection and distribution of economic data, it has to date invested very little in providing comprehensive economic accounts. And while many in the private sector have
attempted to construct such accounts, private researchers have neither the resources nor the data to do so. As a result, the United States today has not set of comprehensive economic accounts, public or private.
An investment in comprehensive economic accounts would benefit the nation because, as noted earlier, better information allows both the public and private sectors to make better decisions. In particular, improved data on the interaction between the economy and the natural environment would have substantial economic benefit for the nation. Many examples of such benefits can be cited. Here we mention but a few from different areas to suggest the range of benefits to be derived.
One important area in which environmental accounting has proved useful is productivity. Growth in productivity, measured as output per person-hour, declined sharply after 1973. One of the leading explanations for this decline was that increased health and safety regulations were imposing significant economic burdens on the nation's businesses. Preliminary versions of natural-resource and environmental accounts—particularly the estimates of pollution control and abatement expenditures prepared by BEA—were of great value for estimating the impact of regulation and pollution-control expenditures on productivity. It is now generally accepted among productivity specialists that environmental regulation is responsible for some of the slowdown in productivity growth; without the existing environmental accounts, it is doubtful whether such a clear understanding would have been possible. Similar studies have analyzed the effect of pollution controls on agriculture and on coastal waters.4
One of the most difficult problems in environmental policy has been comparison of the costs and benefits of environmental regulations. The nation invests substantial sums in cleaning its air, water, and land. These investments have yielded substantial benefits in the form of declining emissions of many pollutants and fewer violations of air quality standards. What is unclear at present is the extent to which the expenditures have produced commensurate economic benefits in terms of improved human health, higher crop yields, and reduced property damage. Recent studies indicate that there have been substantial net economic benefits from pollution control (see Chapter 4). But these studies have not provided sufficient detail to allow pollutant-by-pollutant or sector-by-sector estimates of costs and benefits. Improved accounting systems for the environment could help refine our estimates and regulatory tools so that our pollution control investments might be more effectively allocated. There are major stakes involved here. A 10 percent reduction in pollution
control expenditures due to improved information would amount to more than $10 billion per year in efficiencies for the nation.
Another potentially valuable application of environmental accounting relates to management of the nation's public lands. The nation's forests, rangelands, and waters provide a broad spectrum of valuable economic services. The federal government today reaps substantial revenues from timber harvesting, mining, and leasing of rangelands. A better set of accounts would probably indicate that current leasing policies are providing substantial subsidies. Between May 1994 and September 1996, mining companies patented claims on federal lands with an estimated gross mineral value of $15.3 billion, yet the charge to lessees for these claims was only $19,190. Because the accounting for federal mineral values was incomplete, the full resource value is not currently estimated. Similar subsidies are found in timber and rangeland (see Council of Economic Advisers, 1997). Improved accounts would help decision makers estimate the value of such federal assets and set more realistic prices for leases and patents.
Another area in which comprehensive accounts would be of great benefit is assessment of the costs and benefits of measures to slow greenhouse warming. Under the Kyoto Protocol of December 1997, the United States has undertaken to reduce its greenhouse gas emissions by 7 percent in the 2008-2012 period relative to 1990 emissions. The reductions are to include not only reduced emissions from industrial sources, but also the reductions resulting from carbon sequestration in forests. A comprehensive set of physical and economic accounts would provide the information base needed to estimate the carbon sequestration in forests. Current estimates are that approximately 200 million tons per year of carbon is being accumulated in forests. The nation would save $20 billion annually if a comprehensive set of measures and accounts verified this level of sequestration, if this sequestration could be used to offset industrial emission reductions, and if those industrial emission reductions cost $100 per ton of carbon. This is one of the most dramatic examples of the benefits of establishing comprehensive nonmarket physical and economic accounts.
Economists have developed a new view of the role of data collection, in which data are valuable because they allow better decisions to be made by both the public and private sectors. For example, better weather forecasting allows farmers to harvest their crops so as to reduce damage from frost. Another area that has been intensively studied is the value of better information about the science and economics of climate change. Governments and private firms, such as oil and coal companies and electric utilities, must cope with the enormous uncertainties in this area. Many of these uncertainties result from inadequate accounting of the costs of emission reductions and the potential impact of climate change in nonmarket
sectors. Recent studies have found that improved information in this area would have substantial value. For example, it is estimated that reducing uncertainties about the costs and damages of climate change by half over the next two decades would be worth more than $20 billion. 5 Clearly, as the United States and other countries grapple with the conflict between their international commitments and the domestic costs of emission reductions, improved information on the economic costs and benefits involved could greatly benefit the analysis.
Link Between National Income Accounting and Measures of Sustainable Income
In light of increasing environmental problems in many sectors, concerns have been raised about the sustainability of current patterns of economic activity in both developed and developing countries. What are the environmental and economic implications of continuing ''business as usual"? Will the current path of population, energy use, and growth of human settlements do irreversible harm to the natural ecosystems and life-support systems of the earth? Are we headed for economic overshoot and collapse if we continue to rely on today's technologies? In short, is our economy on a sustainable path? Economists have developed measures of national income and output that incorporate notions of sustainability. This section describes how the current national accounts are related to measures of sustainable consumption. A more complete discussion is contained in Appendix A.
Measures of national income take two fundamentally different approaches—one based on the idea of current production and one based on sustainable consumption. Those who originally constructed national income accounts were understandably concerned with obtaining accurate production-based measures of national output and income because much of their work took place in the shadow of the Great Depression. In the production-based view, tracking current production is seen as critical because it allows governments to take measures to stabilize the business cycle. As noted earlier, production-based measures usually rely on Hicksian income, which is the standard definition of net domestic or national product used in the national income accounts of virtually all nations today. The concept is production based in the sense that production in a given period is measured at market prices.
While standard production-based measures of income are useful tools for measuring current production, they do not directly address concerns
about the sustainability of current decisions. As suggested earlier, it is conventional in economic analyses to define sustainable national income as the maximum amount that can be consumed while ensuring that all future generations can have living standards that are at least as high as those of the current generation.6 Economic welfare, in this view, consists of per capita consumption of goods and services, both market and nonmarket. Consumption includes market items such as food, shelter, and entertainment; it also includes nonmarket items such as home-cooked meals or camping.7
What is the relationship between current measures of national output, such as net domestic product (NDP), and sustainable income? One of the most surprising results of modern economic theory is the output-sustainability correspondence principle (see Appendix A). This principle holds that under idealized conditions, net national product and sustainable income are identical. More precisely, when population is constant, when the national accounts include all stocks of capital and other dynamic features that affect production, and when markets accurately capture the entire social value of economic activity, NDP is an appropriate measure of sustainable income. In other words, the sum of total consumption and net capital formation is equivalent to the maximum sustainable amount of per capita consumption an economy can maintain indefinitely. Under idealized conditions, then, extending the NIPA to include comprehensive measures of consumption and net investment would make output and income more accurate indexes of sustainable income.8
The output-sustainability correspondence is of fundamental importance for guiding decisions about the design of the NIPA. However, important practical and theoretical qualifications to this principle must be emphasized. Augmented NDP will fail to measure sustainable income
accurately (1) if the list of consumption and asset categories is incomplete, (2) if there are technological changes or similar processes that are not captured in investment data, (3) if there are revaluation effects not calculated in the accounts, or (4) if there are market imperfections such as imperfect foresight. The significance of each of these issues is discussed in Appendix A. Even though the conditions under which the correspondence principle applies are quite stringent, the basic insight is of great value for the designing of environmental accounts.
Alternative Approaches to Environmental Accounting
General Issues in Environmental Accounting
Over the last quarter-century, official statistical agencies and individual researchers have responded to the deficiencies in current accounting approaches by developing alternative approaches and novel systems of accounts. The first approaches, by individual researchers, tested wholly new frameworks that often included major aspects of nonmarket activities. Later, official statistical agencies began to take incremental steps toward including some activities that are near-market in nature. The differences in approach generally reflect varying emphasis on the deficiencies discussed earlier, differing views on the functions of national accounting, and differences in what are considered the appropriate functions of official statistical agencies.
An important difference among alternative approaches is the relative importance assigned to economic as opposed to physical accounting. Many approaches emphasize the importance of developing economic accounts, which, as discussed in the last section, are useful for both scorekeeping and management. Other approaches emphasize physical indicators, stressing the development of detailed information on physical flows and human exposures and impacts. Such an approach would be emphasized, particularly by official agencies, when construction of economic aggregates depended heavily on controversial analytical methods and imputations and when collection and dissemination of objective data was the primary goal.
One major issue involved in decisions about how far to extend the boundary of augmented and environmental accounts concerns data quality. As the accounts move further away from the current market boundary line, the quality of the data becomes increasingly suspect, and the cost of obtaining the data becomes increasingly large. Such market or near-market data as volumes and values of petroleum reserves or timber stocks can be estimated with an accuracy reasonably comparable to that of mar-
ket data. On the other hand, obtaining data on nonmarket assets such as fishing stocks or the value of breeding potential is likely to be significantly more expensive. The data become even more fragmentary as one moves toward including environmental activities that have public-goods characteristics, such as the value of lower concentrations of particulate matter or improved visibility. Additionally, valuation sometimes involves highly complex and controversial approaches, such as use of survey questionnaires in which respondents are asked to place dollar values on hypothetical environmental conditions (an approach commonly referred to as "contingent valuation"; see Chapter 4). While private scholars might be willing to use back-of-the-envelope, or even seat-of-the pants, approaches, official statistical agencies are more reluctant to compromise their reputations with controversial and unproven methodologies.
The following subsections review approaches that emphasize physical accounting, the development of comprehensive economic accounts that include nonmarket activity and environmental services, and proposed approaches to environmental accounting developed by the United Nations.
One way to improve our understanding of the interaction between the economy and the environment is to supplement the accounts with improved physical information. Important examples are information on the state of the environment (e.g., ambient pollution levels and forest cover), the status of natural resources (e.g., reserves and resources of petroleum and natural gas), and the impacts of changing environmental conditions on human and ecosystem health (e.g., human exposures to different pollutants or pH levels in lakes). Such information can be arranged in a formal material-flow accounting system, such as that developed by Ayres and Knesse (1969). As demonstrated by Leontief (1970), the integration of such information with conventional economic data can be made quite rigorous by supplementing conventional input-output analyses with data on the flows of environmental pollutants. Sophisticated versions of such input-output matrices have been generated by Duchin and colleagues at New York University (see, for example, Duchin and Lange, 1993). This approach can be developed into a comprehensive input-output system that integrates economic and physical environmental information. An important example is the National Accounting Matrix Including Environmental Accounts (NAMEA) developed by Keuning and colleagues in The Netherlands (see Keuning, 1993). Similar physical accounting systems exist in Norway and France.
Most physical accounting efforts do not embed the information in an
input-output framework, but attempt to be more descriptive. These approaches take account of the environment by assembling large quantities of descriptive physical information, such as indicators of air and water quality, species counts, and area of forest cover. Typically, these informal accounting systems appear as national state-of-the-environment reports or in large physical environmental databases such as the STRESS system in Canada and similarly large databases maintained by several U.S. governmental agencies.
Physical accounting systems play an important role in accounting and policy formulation. They provide the underlying data for regulatory analysis and for development of the aggregates that underlie economic accounting. Moreover, they provide rich physical and intuitive measures of environmental impacts. At the same time, several factors complicate their use for policy purposes. First, the choice of appropriate physical units of measure is not obvious. Presumably, the units of measure should be relevant for some environmental-policy concern. Should a forest, for example, be measured in terms of its acreage, the volume of its timber, the variety of its biota (as evidenced by the number of available species), the stock of nontimber resources such as firewood and grasses, or the number of miles and acres of fishable waters? From the policy maker's point of view, the answer will depend on policy objectives: commercial timber management, firewood supply, recreational uses, erosion protection, species diversity, and so on. Additionally, when environmental assets have multiple uses, as in the case of forests, the units of the indicators are different (acres, cubic feet, number of species, cords of firewood, and miles of streams). The noncommensurate nature of the different attributes makes physical accounting rich in detail, but poor for making policy decisions and determining tradeoffs.
In all accounting systems, important questions relate to coverage, detail, and aggregation. In an effort to encompass the many policy issues involved, physical systems can easily become quite large and detailed. Of course, national accounting systems are also enormous data systems—but most of the vast data iceberg is under water, and only the monetary aggregates are visible in the published numbers. Indeed, large data systems are worth little for scorekeeping, modeling, or policy purposes unless they can be aggregated in such a way that they can be digested and understood. While economists often suggest that measures should be aggregated in terms of dollar values (or present values if there are streams of values over time), putting physical measures into a common unit of account often involves difficult valuation issues. In many cases (for example, protection of unique resources such as the wildness of Yellowstone or the visibility at the Grand Canyon), policy makers may be uncomfort-
able with aggregating these unique values and trading them off against mundane things such as guns or butter.
Aggregation of data in different physical units requires weighting the various measures in order to convert them to a common unit of account. Often this aggregation is accomplished by using a common physical unit of measure, such as weight, volume, or energy content. However, this approach is seldom sensible because the environmental impacts per unit of physical measure differ by orders of magnitude according to the substance and the pathway of human exposure. Compare, for example, the impact of 20 kilograms of plutonium and sulfur in different delivery vehicles.
The Dutch NAMEA system converts dissimilar pollutants to common units on the basis of their contribution to environmental themes such as global warming and acid rain. Thus, emissions of greenhouse gases today are commonly measured in terms of their CO2 equivalent or global-warming potential. While this approach is often sensible, it embodies hidden assumptions that may be highly controversial on close scrutiny. For example, the conversion of greenhouse gas pollutants to a common unit requires detailed scientific knowledge about the relative contributions of different gases, and if the contribution is nonlinear (as is almost always the case), the aggregation will be inaccurate. Sometimes, the aggregation includes hidden economic assumptions. For example, the usual approach to aggregating greenhouse gases is to take their contribution to global warming over 100 or 200 years, but not to discount them; this approach is generally flawed and may lead to inappropriate decisions (Reilly and Richards, 1993).
The above examples suggest that physical indicators are subject to many of the same pitfalls and difficulties that plague economic measures of nonmarket and environment activities. Physical accounting systems are most valuable for policy and scorekeeping purposes when overall environmental objectives and targets are clearly established. When the physical systems are highly complex and heterogeneous and are less closely linked to policy objectives, physical accounting is less useful for policy or accounting purposes. It is essential to emphasize, however, that detailed physical information remains an essential component of both economic accounts and environmental policy making.
At present, the United States places less emphasis on developing a comprehensive set of environmental indicators than do many other nations, especially in Europe. Recently, the need for a set of policy-relevant and scientifically based environmental indicators has received high-level attention. This point has been emphasized by the President's Council on Sustainable Development and the National Performance Review.
The development of improved environmental indicators is an important priority if the United States is to enhance its ability to evaluate and analyze environmental trends and to understand the interaction between the environment and the economy. To be useful for both policy-making and accounting purposes, these indicators should be designed to measure variables close to the area of ultimate concern; they should recognize the heterogeneity of environmental activities and damages, avoiding where possible simple national averages and recognizing the great diversity of the United States, particularly in areas where thresholds and non–linearities are important; and they should be capable of estimation through suitably stratified sampling, rather than requiring comprehensive population counts or inventories. From the point of view of environmental accounting, enhancing the accounts in a manner that is scientifically and economically sound will require considerable improvement in the underlying physical data.
Development of Comprehensive Economic Accounts
Comprehensive Measures of Income and Output
One approach to developing comprehensive economic accounts, and the one with the longest history, is the construction of comprehensive measures of national income or output to supplement the conventional national economic accounts. Many of these efforts have been broad-based attempts to address the general issues raised by the national accounts as discussed above, while others have focused primarily on the natural-resource and environmental components of the accounts.
Early efforts were part of the broader movement to construct more meaningful and comprehensive measures of economic welfare. These studies usually redefined the central concept of economic activity by extending both "consumption" and "output" to encompass large portions of nonmarket activity, sometimes including environmental activities. The first example of this approach is the Measure of Economic Welfare (MEW) indicator developed by Nordhaus and Tobin (1972). They defined an entirely new measure of economic welfare that included major new components such as leisure, nonmarket work, and imputations for the services of government and consumer capital. This measure also excluded activities that do not contribute to economic welfare, for example, commuting costs and regrettable necessities such as military spending. In addition, it subtracted an estimate of the environmental disamenities associated with urban activities.
A similar approach relying heavily on the concepts behind the MEW, Net National Welfare (NNW), was developed by the Japanese govern-
ment (Japan, Economic Council, 1973). A number of comprehensive measures of output for the United States, including many nonmarket activities and assets as well as environmental activities, were developed by Zolotas (1981), Kendrick (1987, 1996), and Eisner (1985), and by Jorgenson in association with Fraumeni (1987) and with Christensen and Jorgenson (1969, 1973).9 In addition, those emphasizing the ecological approach to economics have developed comprehensive accounts and attempted to value natural ecosystems (see Daly and Cobb, 1989, 1994; Costanza et al., 1997).
Targeted Approaches to Environmental Accounting
Comprehensive approaches are useful supplements to the conventional national accounts in that they can sketch the evolution of broad measures of economic activity. However, because most of the comprehensive approaches to measuring national output and income treat natural-resource and environmental measures in a broad-brush fashion, they do not provide many of the important details about particular sectors, environmental activities and assets, and interactions between the environment and the economy. Over the last two decades, many studies have taken a more targeted approach to environmental accounts, focusing on how the national accounts would be modified to incorporate the environment and offering estimates of economic activity in sectors that provide services of natural resources. One approach that treats the natural-resource and environmental sectors in more detail and illustrates many of the major issues involved is that developed by Peskin (1989a, 1989b). This approach was originally designed for the Measurement of Economic and Social Performance Project (sponsored by the National Science Foundation) and was recently adopted for the Philippine Environmental and Natural Resources Accounting Project (ENRAP) (see Peskin, 1989a, 1989b, for a description of the basic framework).
The ENRAP system is based on the principle that natural resources (including air, water, and land) have economic value because they generate valuable goods and services. Some of these services are marketed, such as commercial timber from forests, and these services are already included in the conventional market economic accounts. But many other valuable services—such as those associated with recreation, drinking water, and waste disposal—are not marketed, even though they have significant economic value.
The ENRAP accounting framework starts with the conventional economic accounts and the conventional distinction between sectoral inputs and outputs. Those nonmarketed services that are inputs or intermediate goods (such as erosion protection that enhances agricultural production or land and water that provide disposal services) are added to the input side of the accounts. Perhaps the most important modification is on the output side. Outputs include not only marketed outputs, but also nonmarketed goods and services that go to investment or final consumption. Outputs also include as a negative item the environmental damage resulting from pollution. From an analytical point of view, capital formation includes such items as net increases of natural capital in the form of changes in forest stock or mineral reserves and carbon sequestration by forests. Added final consumption includes the value of recreational services such as visits to national parks or recreational fishing and the value of changes in health status due to changes in air quality or drinking water.
In conventional accounting, the marginal value of a produced good or service equals its price. The ENRAP approach recognizes that in the absence of market prices, there is nothing to ensure that the value of a nonmarket service will equal the price of that service. Therefore, to obtain accounting balance with the introduction of nonmarket environmental services, ENRAP introduces a term on the input side of the accounts—"net environmental benefit." This entry is defined as the sum of the values of environmental input services (such as waste disposal), plus the values of positive output services (such as recreation), less any negative social damages (such as pollution damage) arising from the use of environmental inputs.
Another set of studies has focused on deriving estimates of economic activity in sectors providing services of natural resources or the environment, with emphasis on mineral fuels and forests. An important set of studies in this area has been undertaken by Repetto and colleagues at the World Resources Institute (Repetto et al., 1989). The principal thrust of this effort is to modify the conventional net national income and output by deducting estimates of the value of the depletion of natural resources such as forests, mineral stocks, fish stocks, and soils. The rationale for this modification is to ensure that reproducible capital and natural capital receive comparable treatment in the computation of net investment, net output, and national income. Expanding the boundary of the accounts also allows a more comprehensive definition of national saving and national wealth by including natural resources of minerals and forests, along with land and reproducible capital, in the definition of assets. More recently, the World Bank (1997) has provided estimates of augmented wealth, national output, and saving for a large number of developed and developing countries.
Some Common Issues in Environmental Accounting
In principle, it is economically sound to adjust conventional national output and income measures for final nonmarket consumption provided by the environment and other activities, as well as for the net capital accumulation in nonmarket assets. Adding nonmarket consumption and investment will produce a more accurate measure of sustainable income. Two decades of work on environmental accounting has shown, however, that there are significant obstacles to the construction of accurate estimates of augmented national income and output. For practical accountants, the most daunting obstacles are empirical and data problems involved in estimating quantities of stocks and flows and providing monetary valuation; these problems were discussed briefly above. There are also conceptual issues. To illustrate the challenges involved in including nonmarket activities, we discuss three such issues here—treatment of depreciation, treatment of pollution abatement expenditures, and issues of valuation.
According to standard national accounting conventions, the value of an asset is its market value, which, under competitive conditions, will equal the present value of the net returns from that asset. In this report, depreciation (or depletion for subsoil assets) is defined as the expected change in the present value of the returns on an asset due to aging. Under this approach, changes in present value due to changes in expected interest rates, expected prices, or expected physical flows are called revaluations. These definitions are the same for environmental assets, but for those assets the physical flows include both public and private services and damages, such as value of drinking water and adverse health effects. From an economic point of view, depreciation depends on the expected decline in the value of the economic services of an asset and not solely on its physical condition. For example, a forest might increase in value even though it had a declining volume of timber production if there were an increase in production of other goods and services.10
Asset values and depreciation can be estimated by calculating the discounted present value of the stream of net returns, market and nonmarket, from environmental capital. This calculation is complicated because it requires estimating future returns, capital lifetimes, and discount
rates. A number of simplified methods have been proposed to overcome these difficulties, many of which are reviewed in detail in the next two chapters. Particularly for nonreproducible and renewable assets, alternative approaches give quite different answers, so caution must be used when applying those approaches to environmental assets. In examples of the value of net investment in subsoil assets for the United States, the results sometimes have differing algebraic signs under different approaches. These results emphasize the difficulties of ensuring precision when moving beyond the traditional boundaries of the marketplace.
Pollution expenditure accounting.
One early suggestion for improving the conventional economic accounts, especially with respect to their neglect of environmental deterioration, was to treat all pollution abatement expenditures as ''intermediate" expenditures in the national accounts. As a consequence, pollution abatement investments, governmental municipal sewage treatment expenditures, and defensive consumer outlays for pollution control would not be counted in GDP. The idea of deducting environmental protection and similarly defensive expenditures from GDP has a long history, but has never been formally adopted by national accountants, presumably because of difficulties in drawing the line between defensive and nondefensive outlays (see Nordhaus and Tobin, 1972; Juster, 1973).
One way of rationalizing a proposal to treat pollution abatement expenditures as intermediate inputs is to assume that these expenditures are just what is necessary to keep the stock of environmental capital intact—that is, to assume that the expenditures are exactly sufficient to offset any pollution and other environmental degradation. While convenient, this assumption is unlikely to be realistic for any particular time or sector. It seems likely that until the early 1970s, environmental quality in many areas was deteriorating, while since then it has improved in many areas.
Additionally, current measures of pollution abatement are defective because they are poor estimates of the true cost of environmental regulation. For example, many plants met regulations under the original Clean Water Act without making any expenditures on abatement equipment. They simply ceased producing certain highly polluting products, such as bright, highly coated writing papers. Other examples include the elimination of certain oil-based paints and of leaded gasoline. In these cases, abatement expenditures were zero, although true economic costs were positive. In other cases, abatement expenditures overstate true opportunity costs because of the difficulty of separating accounting costs into pollution abatement and other costs. Moreover, many pollution abatement activities are voluntary and are not in response to policy. For ex-
ample, a major component of the U.S. pollution abatement expenditure series is expenditures associated with sewer hookups and septic tanks for newly constructed housing. Such sanitary practices have a history that long predates the environmental movement, and recent activity reflects local laws, building codes, and zoning ordinances regarding pollution. If expenditures associated with such conventional practices are not excluded from pollution abatement expenditure series, the use of such series to estimate the costs of regulation or explain productivity changes can yield very misleading results. How large might this overestimate be? One study has shown that about 20 percent of reported pollution abatement expenditures in the United States did not originate in federal regulatory policy. In some industrial sectors, nearly all reported expenditures predated federal regulations. 11
Traditional economic accounts use market prices to value intermediate and final output. For near-market or nonmarket activities, economists rely on alternative approaches that use proxies for market prices or develop alternative methods that impute values indirectly. One example is firewood collected by households on their own property; here, the appropriate value would be the market price of equivalent commercially sold firewood. Life becomes more complicated when the goods and services have no market equivalent or have public-goods characteristics. In these cases, values are often imputed (1) by using surveys, (2) by unbundling the commodities and valuing component parts, or (3) by looking at behavior that reveals consumer valuation of the commodities. These three techniques are exemplified by contingent-valuation surveys, hedonic regressions, and the travel-cost method, respectively. 12
A key feature of the appropriate design of augmented accounts is that prices or values should always be measured by the value of the marginal or last unit of the good or service consumed. That is, the value of bottled water is not the average value, but the value of the last unit drunk, which will be significantly lower than the average; the difference between average and marginal value is called consumer surplus. This convention of using the marginal value is employed throughout the national economic accounts, so adhering to this approach will ensure that environmental goods and services are valued consistently with market goods and services. One of the difficulties in adopting valuations from the existing
environmental literature is that many valuation studies calculate the average rather than the marginal values (that is, they add in consumer surplus, which is appropriate for economic accounts). Use of average values will tend to overstate the total value of an item relative to market goods and services.
There is an interesting parallel here between valuation issues for environmental services and difficulties in measuring the cost of living. Recently, a group of economists undertook a comprehensive assessment of the adequacy of the Consumer Price Index (CPI) (Advisory Commission to Study the Consumer Price Index, 1996). The commission collected a wide variety of studies and investigated whether the CPI accurately measures the trend in the cost of living. The commission concluded that the CPI has a significant upward bias, primarily because of an inadequate treatment of quality change.
Criticisms of the CPI revolve primarily around the difficulty of measuring nonmarket services. That is, a major criticism of the CPI is that it measures the prices of the market goods that consumers purchase rather than the prices of the nonmarket services these goods deliver. Thus the CPI measures the prices of automobiles, electricity, and hospital days, not the costs of travel, lighting, or delivering a baby. To estimate the service prices for consumer purchases would require—in a way directly parallel to valuation of environmental goods and services—imputing or calculating the values of nonmarket services. There are no deep theoretical problems involved in the estimation of these nonmarket values, but they present measurement problems in practice because transaction prices for the services are almost never observed. Many of the difficulties that arise in correcting the CPI for quality change are analogous to the difficulties that arise in valuing nonmarket environmental goods and services. Indeed, correcting the CPI for quality change would probably be easier than estimating the proper values of environmental goods and services because the CPI includes primarily private goods, while valuation of environmental services involves public goods as well.
Extensions of the U.N. System of National Accounts: System of Integrated Environmental and Economic Accounting
A great deal of work outside the United States has been devoted to developing physical and monetary accounts for natural resources and the environment (see for example Uno and Bartelmus, 1998). Because of the diversity of approaches and controversies about alternative methodologies, however, no international consensus has been reached on the appropriate model for establishing a uniform system of environmental accounts. Therefore, it was decided for the 1993 SNA to treat environmental ac-
counts as satellite accounts. Environmental accounts would thereby serve as a tool for expanding the analytical capacities of the national accounts without changing the core accounts, thus complementing rather than substituting for the traditional national accounts (see United Nations, 1984, 1991, 1993).
The various approaches were compiled and synthesized in the United Nations System of Integrated Environmental and Economic Accounting (SEEA) (United Nations, 1993). Unlike the SNA, the SEEA has not been adopted as an international standard and should be viewed as a set of proposals for environmental accounts.
The SEEA is a highly flexible framework encompassing approaches that range from reorganizing the current accounts to building a full set of household and nonmarket service accounts. The basic framework envisions adding environmental flows in a series of steps or versions. Version I of the SEEA reorganizes the traditional national accounts to highlight environmental and natural-resource flows. Version II is a restatement of the expenditure-accounting approaches describing the monetary and physical flows and stocks. Version III links the physical information of version II with the monetary data of version I. Version IV imputes environmental damages to obtain a more comprehensive measure of output and includes the depletion of natural resources and environmental pollution costs. Version V, which has not been extensively discussed, considers more radical extensions, such as extending the production boundary in the household sector and introducing environmental services as an output.
Since version IV has received the most international attention, it is the focus of the discussion here. Version IV treats environmental degradation and depletion as subtractions from net product. In effect, both depletion and degradation are viewed as sources of depreciation of natural capital. We focus here on two examples of how the SEEA differs from alternative approaches. One is in the estimation of depletion of natural resources such as petroleum, which is valued at market values or sometimes at replacement cost. A second is the cost of environmental degradation—such as water pollution—which is treated either as "costs caused" or "costs borne." Under this distinction, degradation can be valued in terms of either the costs to the sector if it were to eliminate the degradation (costs caused) or the damage to producing or affected sectors due to the degradation (costs borne). For the most part, when implementing the SEEA, researchers have relied on the costs-caused approach. These depletion and degradation estimates are subtracted from conventionally defined value-added to derive environmentally adjusted net income measures.
It is useful to highlight the fact that the SEEA relies heavily on costs in its design of environmental accounts. Although it is common practice
today, the use of restoration-cost estimates to measure environmental degradation and replacement cost to measure natural-resource depreciation is an inconsistent and inappropriate practice. The appropriate approach is to measure the market value (along with the relevant value of nonmarket impacts) of any change in the services of these environmental assets and of the change in the stocks of these assets. As is discussed in Chapter 3, for example, the appropriate valuation of depletion of petroleum stocks is the change in the market value of oil in the ground.
Use of the SEEA methodology can lead to inappropriate results. Suppose that the environment is initially clean and that the market and nonmarket damages from emitting a few grams of dust are very small. The cost of maintaining the clear environment by reducing those last few grams of dust might be enormous. Thus, the use of restoration costs as a measure of pollution control benefits (or damages) can lead to a significant overestimate of benefits. Paradoxically, if restoration costs are used to measure damage, the clean economy may be shown to be more environmentally damaged than the dirty society.13 With respect to the degradation estimates, the authors of the SEEA recognize the theoretical difficulties involved in using cost-caused or replacement-cost data. While leaving the door open for the use of valuation estimates based on damages or costs borne, the authors are skeptical of the practical use of valuation techniques and of similar imputation methods.
A second notable feature of the SEEA framework is its adherence to conventional SNA sectoring or production boundaries. The close adherence to SNA concepts is an important advantage since it helps ensure consistency with the core accounts. Also, since the SNA framework has been widely adopted, the close adherence of the SEEA to the SNA will help ensure international comparability. This consistency comes at a price, however. The most important shortcoming arises because of the omission of nonmarket services and investments. There is no place in the SEEA system (at least with versions I through IV) for the amenities provided by the environment in the form of recreation, health impacts, erosion control, or disposal services. The SEEA in effect equates the term "nonmarket" with "noneconomic."
As a result of the omission of environmental services, the economic link between the economic value of an environmental asset and the services it provides is broken. Thus, while a forest can have economic value
in the SEEA framework, this value comes only from the commercial products of the forest, such as timber, and not from other forest services, such as watershed protection, recreational services, and carbon sequestration. By neglecting nonmarket assets and services, the SEEA also limits the coverage of household production. Some household production—such as the production of nonmarketed firewood—has both substantial economic value to households and serious environmental consequences due to the pollution from the smoke.
A third key feature of the SEEA is the treatment of natural-resource depletion. In conventional accounting, net investment is measured as the change in the value of the stock of an asset between two periods. The SEEA does not employ the usual definition of net investment, but focuses only on natural-resource depletion. Under the SEEA, when resources are depleted, there is a deduction from net output; but when resources are discovered, there is no increment to net output. Hence, even though the stock of petroleum reserves is constant over time, the SEEA would be recording a series of deductions from output and income to reflect petroleum production. The SEEA logic is that discovered resources are not really additions to the stock; they merely represent a shift from the nonproduced, noneconomic stock of assets to the nonproduced, economic stock. If, however, the stock of petroleum were valued in terms of its market value or discounted services, additions and depletions would be treated more symmetrically. Discoveries would increase the value of the stock, while depletion would decrease its value.
Environmental Accounting in Other Countries
As noted earlier, environmental and natural-resource accounting has been extensively developed in countries outside the United States over the last quarter-century. As in the United States, these augmented accounts represent an attempt to cast light on the interactions between the economy and the environment. In other industrialized countries, three main areas of concern have been identified:
- Depletion—Some countries have been concerned about the depletion of scarce natural resources. Particularly in northern Europe, where North Sea oil and gas resources constitute a significant fraction of natural assets, policy makers want to determine the extent to which nonrenewable resources are being depleted.
- Degradation—Many countries have been concerned about the degradation of the natural environment through pollution. Pollution not only renders the air, water, and soils less productive, but has health impacts and degrades people's enjoyment of the environment.
- Protection—Countries adopt numerous measures to protect or restore the environment. These activities include pollution abatement and control expenditures, research into cleaner technologies, ecological taxes, and fiscal incentives for environmentally benign production patterns. Policy makers are interested in the economic costs and impacts of such environmental protection measures.
Each of these three major facets of the interaction between the economy and the environment corresponds to a different aspect of economic policy and requires different data sets, concepts, and classifications. A comprehensive environmental accounting system addresses each of these sets of issues. Environmental accounting in the United States has to date addressed primarily the issue of mineral depletion and (up to 1995) the cost of environmental protection. Approaches being applied in other countries involve analyzing different parts of the interaction, such as material flows into and within the economy; recycling; the costs of meeting environmental targets; ecological taxes; and emissions of various pollutants into the air, water, and soils.
As noted earlier, other countries have adopted BEA's approach of keeping their environmental and natural-resource accounts in satellite or supplemental accounts; the core national accounts have not been modified to reflect environmental and natural-resource changes. This approach has been endorsed by the European Commission (European Union, 1994:5):
The development of a "greened" GNP, although having a certain appeal ..., raises a number of difficult methodological questions which rule it out as a realistic option for the foreseeable future. Therefore what is needed—as a first step—is an approach which makes environmentally interesting parts like resource depletion and environment degradation, firstly in the form of physical indicators, later with the help of available techniques transformed into monetary value, still—however—keeping the various building blocks of such a system of integrated environmental and economic accounting separate, a so-called satellite approach.
Thus the European Union has decided to take the same approach as the United States: to focus on creating multiple, integrated data sets that track the interaction between the economy and the environment. This approach emphasizes the multidimensional nature of the interaction, rather than attempting to create a single-number modified GDP.
Table 2-1 provides a summary of the major sectors that have been studied in environmental accounts of various high-income countries. This table refers to published studies by official statistical agencies comparable
TABLE 2-1 Development of Environmental and Natural-Resource Accounts in Major Industrial Countries (synoptic table illustrating areas in which countries are working)
to BEA. In addition, extensive work in other areas has been undertaken by private research institutes. To a considerable extent, the focus of the environmental and natural-resource accounting of each country reflects its own national priorities and policy concerns. Therefore, Canada and the Scandinavian countries have highlighted forestry accounts, while densely populated Holland has focused more intensively on pollution of air, rivers (particularly the Rhine), and soils.
Some of the environmental accounts are quite close to the existing national accounts; this is particularly the case for the mineral accounts, which are conceptually included in existing national wealth accounts. Other accounts consist primarily of disaggregating existing transactions, such as those concerned with pollution abatement and control expenditures. Another set of accounts represents an extension of existing input-output systems to include physical flows of pollutants along with the purchases and sales of goods and services.
The above review indicates that the principles and practices of environmental and natural-resource accounting are well developed in major industrial countries. Countries are concerned about the interaction be-
tween the economy and the environment, particularly as regards the extent of resource depletion and environmental degradation, as well as the economic costs of environmental protection. Other countries follow the U.S. practice of analyzing environmental linkages in satellite accounts, which provide useful data for both management and scorekeeping without changing the core national economic accounts.
U.S. Integrated Environmental and Economic Satellite Accounts
History of Environmental Accounting in the Commerce Department
Many of the issues considered in the current discussion about expanding the traditional NIPA were involved in the earliest decisions about designing the accounting framework. From the beginning, those within BEA who constructed the NIPA considered aspects of what is now called environmental accounting. It was decided at the outset to focus primarily on an accounting framework whose boundary encompassed market transactions. Interestingly, in early efforts, depletion of mineral assets was a deduction from national product for obtaining net output. This practice was discontinued and depletion removed because the approach was thought to be asymmetrical in subtracting depletion without adding additions.
As the idea of augmented accounting began to emerge in the early 1970s, BEA came under pressure to expand its accounts to include significant nonmarket activities, with an eye to improving the accuracy of the accounts as a measure of economic well-being. At that time, BEA was not inclined to develop augmented accounts because it believed that imputations of the volume and values of nonmarket activities would be subjective and based on unproven methodologies and would lead to a deterioration in the accuracy of the national accounts.
BEA's initial concern was with the failure of the accounts to treat pollution abatement expenditures consistently. Specifically, if some part of final consumption were devoted to pollution control, GDP would not be affected, but if the business sector devoted the same level of resources to pollution abatement, conventionally measured GDP would fall. Rather than making major conceptual changes to the national accounts, BEA recommended that a separate series on pollution abatement expenditures be developed to interpret movements in national output, rather than to change the definition of national output itself. Work began on the development of such a survey in 1971, and preliminary results were published
in 1973. Ironically, this first foray of BEA into environmental accounting was eliminated in the budget cuts of the 1990s.
In 1972, Congress directed the Secretary of Commerce to study the effect of the costs of the Clean Water Act on manufacturers. In response, BEA and the Bureau of the Census developed an environmental statistics program. BEA formed an Environmental Studies Staff within the Office of the Director. Several activities were initiated to support the planned pollution abatement expenditure series. In particular, BEA added a number of questions on pollution abatement to the November 1973 Plant and Equipment Survey of companies, and in 1974 the Census Bureau began surveying about 19,000 manufacturing establishments with regard to their pollution abatement expenditures. In response to the success of BEA's efforts to develop pollution abatement expenditure data, Congress approved funds that allowed for expansion of BEA's environmental program. In 1977, BEA established the Nonmarket Economics Division, which consisting of three branches—the Abatement and Control Expenditures Branch, the Unit Costs and Emissions Branch, and the Measures of Economic Well-Being Branch (Bureau of Economic Analysis, 1987).
The first two of these branches were involved primarily in reconfiguring the data on costs that were already contained in the national economic accounts. The Measures of Economic Well-Being Branch—in a significant departure from conventional national income accounting—focused on the deficiencies of conventional income and output measures as measures of social well-being. The Economic Well-Being Branch conducted research on a number of issues that would arise if the traditional accounts were expanded to better reflect societal well-being. These studies involved the value of nonmarketed household work, the value of services associated with governmental capital and consumer durables, the investment value of education and training, and the value of the discovery and depletion of minerals (Bureau of Economic Analysis, 1982). Methodologies developed in the work on minerals accounting contributed directly to the development of the mineral accounts in Phase I of BEA's IEESA.
Although interest in environmental and resource accounting was growing outside the United States, the Measures of Economic Well-Being Branch was abolished in 1981. The Nonmarket Economics Division was cut and renamed the Environmental Economics Division; the work of this new division was confined to the generation and analysis of pollution abatement expenditure data. In 1995, the balance of the program, along with the pollution abatement expenditure survey, was abolished to meet the budget cuts of the 1990s. In the meantime, BEA initiated work in the IEESA system in 1992, but this work, as noted earlier, was stopped by Congress in 1994.
Overview of the Integrated Environmental and Economic Satellite Accounts
Work on the IEESA began in earnest in 1992 and was accelerated when President Clinton emphasized its importance in his Earth Day speech in 1993. The essential framework for the IEESA, along with a proposed framework for future study, was set forth in two articles in 1994 (see Bureau of Economic Analysis, 1994a, 1994b). As envisioned by BEA at that time, the work plan would have three phases:
- Phase I, Overall Framework and Prototype Estimates for Subsoil Assets—The first phase involved establishing the overall framework and process for developing prototype satellite accounts for subsoil assets such as oil, gas, and nonfuel minerals. The focus was ''on proved reserves, the basis for valuation is market values, and the treatment given mineral resources—which require expenditures to prove and which provide 'services' over a long timespan—is similar to the treatment of fixed capital in the existing accounts" (Bureau of Economic Analysis, 1994a:48-49). The Phase I report of these two articles presented a preliminary view of the framework of U.S. environmental accounts, along with numerical estimates of the values of additions, depletion, and stocks for major subsoil mineral resources.
- Phase II, Renewable Natural Resources—The second phase "calls for work to extend the accounts to renewable natural resource assets, such as trees on timberland, fish stocks, and water resources" (Bureau of Economic Analysis, 1994a:49).
- Phase III, Environmental Assets—The third phase "calls for moving on to issues associated with a broader range of environmental assets, including the economic value of the degradation of clear air and water or the value of recreational assets such as lakes and national forests" (Bureau of Economic Analysis, 1994a:49).
Since publishing its first report in 1994 in the two above-mentioned articles, BEA has ceased further work on environmental accounting in response to the congressional stop-work order.
Summary and Conclusions
The last quarter-century has seen an increasing awareness of the interactions between human societies and the natural environment in which they thrive and upon which they depend. This awareness has been dramatically heightened by concerns about resource scarcity, environmental degradation, global environmental issues, and the possibility that the
economy is not sustainable. The combination of this increasing awareness and recognition of the primitive state of environmental data has led to a widespread desire to broaden the nation's economic accounts to include natural resources and the environment. The idea of including natural-resource and environmental assets and services in the economic accounts is part of a movement to develop broader economic indicators. It reflects the reality that economic and social welfare do not stop at the market's border, but extend to many near-market and nonmarket activities.
BEA has studied augmented accounting since the early 1980s. It began work on the U.S. version of environmental accounting, the IEESA, in 1992. Congressional concerns about environmental accounting were raised shortly after the first publication of the U.S. environmental accounts, and Congress requested that work on the IEESA cease until the methodological issues had been reviewed. In response to the congressional mandate, the Commerce Department asked the National Academy of Sciences to undertake a review of environmental accounting, and this report is a response to that request.
The NIPA are the most important measures of a country's overall economic activity. From the perspective of environmental accounting, the major point to recognize is that GDP is conceptually defined to include the final output of marketed goods and services—that is, goods and services that are bought and sold in market transactions. While recognizing the need to consider alternative measures, it is important to retain the core market-based accounts, which are of great value for historical and international comparisons and will continue to be a critical indicator for much economic policy making.
Work on augmented accounting in official statistical agencies, as well as by individual scholars, has yielded estimates on a wide variety of nonmarket activities for experimental augmented national accounts. The guiding principle in extending the national economic accounts is to measure as much economic activity as feasible, regardless of whether it takes place inside or outside the marketplace. Augmented national economic accounts are designed to provide better measures of final output—of what consumers in the United States currently enjoy in the way of goods and services, and of the accumulation of capital of all kinds that will permit the future production of goods and services.
A set of well-designed environmental accounts can overcome the recognized shortcomings of the current market-based accounts. They can provide useful information for managing the nation's public and private assets, for improving regulatory decisions, and for informing private-sector decisions. The collection of data on comprehensive income and output is an investment that would have a high economic return for the nation. There are many examples of the benefits of comprehensive eco-
nomic accounts. These include better estimates of the impact of regulatory programs on productivity, analyses of the costs and benefits of environmental regulations, management of the nation's public lands and resources, and assessment of the costs and benefits of taking steps to slow global warming.
Augmented national accounts can also provide valuable indicators of whether economic activity is sustainable. The national accounts have a close relationship with measures of sustainable income, since the usual measure of NDP corresponds to the highest sustainable level of per capita consumption under idealized conditions.
The nation's measures of national income and output can be improved by including all consumption and net investment to obtain augmented income and output measures. Among the currently omitted items that need to be added are nonmarket consumption, such as home production and final environmental services, and nonmarket investments, such as changes in the value of resource stocks and investment in human capital.
Over the last quarter-century, official statistical agencies and individual researchers in the United States and abroad have responded to the deficiencies in current accounting approaches by developing alternative approaches and new systems of accounts. An initial general approach is to supplement the accounts with improved data on physical flows. Physical accounting systems are valuable for policy purposes when overall environmental objectives and targets are clearly established. They are an essential component of both economic accounts and environmental policy making. At present, the United States has invested little in developing comprehensive environmental indicators. The development of improved environmental indicators is an important priority for enhancing the nation's ability to evaluate and analyze environmental trends and track the interaction between the environment and the economy. From the point of view of environmental accounting, enhancing the national accounts in a manner that is scientifically and economically sound will require considerable improvement in the underlying physical data.
A second approach to environmental accounting is the construction of comprehensive measures of national income or output to supplement the conventional GDP and NDP accounts. Many efforts in this area have been broad-based attempts to remedy general issues raised by the national accounts. Other studies have introduced augmented environmental accounts with a more targeted approach, focusing on how the national accounts would be modified to incorporate the environment and offering estimates of economic activity in sectors providing services of natural resources.
Because of the diversity of approaches and controversies about appropriate approaches, no international consensus has been achieved on a
uniform system of environmental accounts. The 1993 SNA entailed developing environmental satellite accounts as a way of expanding the analytical capabilities of the national accounts without changing the core accounts. In this proposal, environmental accounts would complement rather than substitute for traditional accounts.
The principles and practices of environmental and natural-resource accounting are well developed. Countries are concerned about the interaction between the economy and the environment, particularly the extent of resource depletion and environmental degradation, as well as the economic costs of environmental protection. Other countries follow the U.S. practice of analyzing environmental linkages in satellite accounts, which provide useful data for both management and scorekeeping without changing the core NIPA.
Intensive work on the IEESA began in the United States in 1992. As envisioned by BEA, the work plan would have three phases: Phase I, which involved establishing the overall framework and developing prototype satellite accounts for subsoil assets such as petroleum, gas, and nonfuel minerals; Phase II, which would extend the accounts to renewable natural-resource assets, such as trees on timberland, fish stocks, and water resources; and Phase III, which would extend the effort to issues associated with a broader range of environmental assets, including the economic value of the degradation of clear air and water and the value of recreational assets such as lakes and national forests.