Accounting for the Environment

International Monetary Fund. External Relations Dept.
Published Date:
January 1991
  • ShareShare
Show Summary Details

Measuring the effects of the interaction between the environment and development recalls the conundrum: If a tree falls in the forest and no one hears it fall does it make a noise? Increasingly, economists have been attempting to include not only fallen trees but also other natural resources in their calculations of national products and incomes. The reason is that current national accounting systems do not capture the value of natural resources adequately and, therefore, development strategies that rely on standard income accounting techniques may not result in sustainable development.

With the rise in awareness of environmental issues in the 1980s, attention is now turning to the need to better understand the value of environmental resources and services and to improve the current United Nations System of National Accounts (SNA) to account for the environment in income estimation. Improved measurement of economic performance should in turn lead to better economic decisionmaking. National authorities and multilateral institutions, such as the United Nations, the International Monetary Fund, and the World Bank, among others, are seeking to address this need. This article traces the rise of this consciousness and the work being done by World Bank staff in this direction.

In conducting economic analyses, measuring economic performance, and directing public policy, the Bank as well as its member countries rely heavily on the major aggregates shown in the national income accounts, compiled in accordance with the United Nations’ SNA. The current SNA, published in 1968, emphasizes GDP. The GDP is a useful measure, mainly of market activity (although it includes some estimates of nonmarketed goods and services), and is an important indicator for macromanagement of an economy. But it has increasingly met criticism, mainly because it takes no account of consumption of natural capital and, therefore, is seen to discourage the implementation of policies that result in sustained development. The SNA framework does not provide for measures such as a net domestic product (which takes account of depreciation of man-made capital). A few countries actually calculate an NDP, but this measure, too, does not capture the loss or depletion of national environmental resources.


The existing SNA is contained within a well-defined receipt and payment framework that generally relies on market prices. But some of the effects of degradation, pollution, and waste disposal, together with their repercussions on society, cannot be captured by market-based information and standard accounting techniques. Thus, the challenge is to capture these effects statistically and to link them to economic activities.

The deficiencies in the ability of the current national accounting framework to take natural resources and the environment into account arise, in part, out of an inconsistent treatment of man-made and natural capital. There are three specific shortcomings:

• Natural and environmental resources are not included in balance sheets; national accounts, therefore, represent limited indicators of national well-being, since they measure poorly—or even “perversely”—changes in environmental and resource conditions;

• Conventional national accounts fail to record the depreciation of natural capital, such as a nation’s stock of water, soil, air, nonrenewable resources, and wildlands, which are essential for human existence.

• Cleanup costs (e.g., expenditures incurred to restore environmental assets) are often included in national income, while environmental damages are not considered. For private firms, defensive environmental expenditures (i.e., measures to reduce or avoid environmental damage) are netted out of final value added. In contrast, such cleanup costs are considered as productive contributions to national output if they are incurred by the public sector or by households. The calculation of GDP is distorted in two ways—undesirable outputs (e.g., pollution) are overlooked and beneficial environment-related inputs related to environmental needs are often implicitly valued at zero.

Some national approaches to environmental accounting


Norway probably has the longest-running interest in environmental and resource accounting. The intent of the Norwegian system is not to provide a better measure of true income and make possible adjustments to GDP, but to assist the government in making decisions on managing resources that are economically and politically most important.

The resource accounts include petroleum, minerals (iron, titanium, copper, zinc, and lead), forest products, fish, and hydropower. Accounts for mineral resources exist for only a few selected years, while forest statistics have been available since 1970, and those for fish, since 1974. The environmental resources accounts are confined to land-use statistics, the discharge of selected air pollutants, and two water pollutants. The coverage of the Norwegian resource accounts could be expanded to cover other resources, but cost-benefit considerations have so far not encouraged such a move.

United States

Environmental and resource accounting in the United States has so far been limited to the collection of data on pollution abatement expenditures. Before 1989, the Bureau of Census undertook a survey of about 20,000 establishments in the manufacturing sector, while the Bureau of Economic Analysis drew its data from a survey of about 9-14,000 firms in both the manufacturing and non manufacturing sectors. Since 1989, only a sample of 600 firms has been surveyed.

Because of a strong lobbying effort by environmental groups, the calculation and publication of a measure called “gross sustainable productivity” was made a requirement for the Commerce Department (Public Law 101-45, June 30, 1989). Also, the bill requires the Secretary of State to instruct the United States representatives to the Organization for Economic Cooperation and Development, the United Nations, and the multilateral development banks to seek revisions to the current accounting systems to take into account the depletion or degradation of natural resources. One problem with the proposed measure is that it has not been clearly defined, and even if agreement on the definition were reached soon, sizable resources might be required for data gathering and estimation of the new measure.


For a considerable time, French experts have been trying to set up a system known as “patrimony accounting.” The system is expected to be comprehensive, consisting of seven levels, starting from specific resource data at level one to aggregate welfare indicators at level seven. The intent is to analyze and describe the natural environment in its three basic functions: economic, ecological, and social.

While a lot of thought has gone into the development of patrimony accounting, very limited resources have been available for its implementation, which is why the process has been slow. So far, it appears that the empirical side of this approach has mainly concentrated on the establishment of resource accounts similar to those established in Norway.

These deficiencies point to the need for an accounting framework that addresses the preceding concerns and permits the computation of measures such as an environmentally adjusted net domestic product (EDP) and an environmentally adjusted net income (ENI). Such measures would attempt to better account for the depreciation of both man-made and natural capital, exclude relevant categories of defensive environmental expenditures, and estimate damages to the environment as a result of economic activities.

Taking into account depreciation of natural capital and calculating an EDP would likely result in a lower level of measured income and perhaps also a lower growth rate. While other methodological approaches could be used, the results of the World Resources Institute are illustrative of how adjusted national income figures might differ from traditional ones. They calculated depreciation for oil, timber, and top soil for Indonesia and found that the growth rate of the adjusted NDP for the period 1971-84 was only 4 percent compared to GDP growth of 7.1 percent.

There are differing views on whether defensive expenditures should be deducted from GDP in order to come closer to a sustainable income as defined below. Generally, progress toward a new accounting framework has been slow. Reasons for this include conceptual and measurement difficulties, as well as uncertainties about the benefits of gathering additional data relative to its costs.

Background to SNA

The foundations of the current approach to national income accounting were laid about half a century ago, when the world’s population and the size of the world economy, as measured by GDP, were much smaller. Consequently, the emphasis on natural resources and the environment was much less at the time, and it is understandable that better treatment of natural resources and the environment in the SNA was not a major concern. Yet, some of the early literature that provided the intellectual underpinning of the SNA anticipated the importance of natural capital and helped to define broad concepts of “true” income that were not limited to output derived from man-made capital

Sir John Hicks defined the concept of income as follows: “The purpose of income calculation in practical affairs is to give people an indication of the amount which they can consume without impoverishing themselves” (John R. Hicks: “Value and Capital,” Second edition, Oxford University Press, Oxford, UK, 1946, p. 172). The same basic idea holds at the national level. True income is a practical guide to the maximum amount a nation can consume without depleting its stock of assets in the future. This would be true to the extent that allowance is made for depreciation of capital (or productive assets), broadly defined. Clearly, GDP, or even NDP as computed under the existing SNA, is not consistent with the spirit of the above definition by Hicks, and hence the urgent need to re-evaluate the SNA.

The Bank and SNA

Based on its work since 1983, the Bank has actively encouraged the inclusion of environmental issues in the current revision of the SNA and proposed as an interim measure that a set of environmental “satellite” accounts be created to accompany the SNA framework. This proposal was accepted by the SNA expert group meeting in January 1989. The revised “Blue Book” of the SNA (expected to be issued in 1993) will include a discussion on this issue and justify the need for satellite accounts that permit income computations that take environmental concerns into account.

The Bank has surveyed the experience of industrial countries with various environmental and resource accounting approaches to see what lessons could be learned for developing countries to better deal with environmental issues in accounting (see box).

Some of the Bank’s current research in this area, carried out jointly with the United Nations Statistical Office (UNSO), is being conducted in two developing countries—Mexico and Papua New Guinea. The UNSO Framework being used in these studies is a system for environmentally adjusted economic accounts (SEEA), which derives EDP and ENI. It tries to integrate environmental data with existing national accounts information, while maintaining SNA concepts and principles as far as possible. The challenging empirical question is the extent to which it will be possible to actually value environmental assets and flows of services from them.

By expanding the SNA to SEEA, the total coverage of productive assets has been increased by adding environmental assets (such as soil, wildlands, and biodiversity) as stores of wealth, provided they are linked to economic activities. Similarly, additional costs related to the environment (e.g., cleanup expenditures by government) are also included in the system. These costs are directly related to productive activities and the generation of value added, and include:

• imputed charges for the depletion of minerals and other natural resources; and

• costs of degradation of land, water, air, and so on, as a result of productive activities.

Such costs, in addition to the depreciation of man-made capital, are deducted from GDP to arrive at the EDP. This EDP does not include damages that are unrelated to productive activities (e.g., natural disasters, naturally occurring erosion, and so on), but they, nevertheless, affect well-being through changes in assets (as a store of wealth). Further, as tentatively suggested in the UNSO Framework, these costs would be reflected in an environmentally adjusted net income (ENI). To arrive at ENI, the following five items would be subtracted from EDP:

Related reading

A volume on environmental accounting was published by the Bank under the title: Environmental Accounting for Sustainable Development and edited by Yusuf Ahmad, Salah El Serafy, and Ernst Lutz, $10.95. A survey of industrial countries’ experiences, prepared by Henry Peskin with Ernst Lutz, under the title “A Survey of Resource and Environmental Accounting in Industrial Countries.” as Environment Working Paper No. 37 (August 1990), the UNSO Framework paper “Integrated Environmental and Economic Accounting Framework for an SNA Satellite System,” by Peter Bartelmus, Carsten Stahmer. and Jan van Tongeren, and a paper by Mohan Munasinghe and Ernst Eutz entitled “Environmental Economic Evaluation of Projects and Policies for Sustainable Development.” Environment Department Working Paper No, 42, January 1991, are all available from the Environment Department, World liank, Washington, DC 2W33, USA.

• environmental protection expenditures of government and households, which are treated as final expenditures in the SNA;

• environmental effects on health and other aspects of human capital;

• environmental costs of household and government consumption activities;

• environmental damage from capital goods that are discarded; and

• negative environmental effects in the country caused by production activities in other countries (negative entry), and negative environmental effects transferred abroad (positive entry). In the case of beneficial effects transmitted across borders, the signs would be reversed.

Conceptually expanding the SNA to an SEEA as outlined above is relatively easy; the difficult part would be to produce actual estimates. Where the estimation work is undertaken, the satellite accounts will provide the countries with considerable flexibility in calculating EDP, ENI, or other selected aggregates. Aside from the explicit suggestions contained in the UNSO Framework (and the forthcoming UNSO handbook), countries may wish to adopt certain specific methods and compute their own indicators.

It should be noted that while flexibility at the experimental stage is desirable, there is obviously a trade-off with comparability across nations, which is needed for cross-country analyses.

One of the options with which countries may experiment is the user cost approach. Under this approach, the revenue from the sale of a depletable resource (net of extraction cost) is split into a capital element, or user cost, and a value-added element, or “true” income. The capital element is considered asset erosion and is therefore not included in GDP. A formula for determining the size of the capital element has been proposed by the Bank’s Salah El Serafy. According to this approach, the larger the current annual exploitation of the resource is in comparison with the known reserves, the larger the user cost and the smaller the income component. Complicating issues regarding this approach are the size of the discount rate to be used in the formula and the size of the proven reserves (which may actually increase over time, as in the case of oil during the last few decades).

In estimating the depreciation of capital in a broad sense, it is already clear from the initial work on Mexico that all kinds of assets could be considered, but with various degrees of difficulty. For example, capital that might be considered includes subsoil assets, forests, fisheries, soil, water, air, biodiversity, and historical monuments. However, data availability and reliability of estimates in such cases will clearly be very difficult and may not be consistent for the various categories.

The Bank is of the view that, at a conceptual level, there is a need to have a broad framework that can handle all kinds of capital (or productive assets). At the practical level, a pragmatic approach would be appropriate, which would suggest seeking initial agreement for appropriate treatment of certain expenditures (e.g., those incurred by governments) to protect the environment and marketable subsoil assets. One could leave other, more difficult areas to be incorporated after further research. In the meantime, while this work proceeds, and until firm conclusions have been reached, users of conventional national income aggregates, such as GDP, should keep the limitations in mind and use supplementary environmental information as far as possible to obtain a more balanced picture.

The World Bank has already published one compendium volume on environmental accounting (see box on related reading). It is planning to prepare a second containing the UNSO framework, the results of the case studies, the survey of environmental accounting approaches in industrial countries, and other relevant papers. Based on the experience of the case studies, the UNSO plans to further develop its draft handbook into a useful guide for countries wishing to deal better with environmental and natural resource issues.

Other Resources Citing This Publication