A decade has passed since Wasting Assets, a study of Indonesia by Robert Repetto and colleagues at the World Resources Institute, drew widespread attention to the potential divergence between gross and net measures of national income. This was by no means the first ‘green accounting’ study. Martin Weitzman, John Hartwick, and Partha Dasgupta and Geoffrey Heal had all conducted seminal theoretical work in the 1970s. But the World Resources Institute study demonstrated that data were adequate even in a developing country to estimate adjustments for the depletion of some important forms of natural capital and that the adjustments could be large relative to conventional, gross measures of national product and investment. The adjusted, net measures suggested that a substantial portion of Indonesia's rapid economic growth during the 1970s and 1980s was simply the unsustainable ‘cashing in’ of the country's natural wealth.

The present paper gives an overview of the theory of green national accounting. Three purposes of green national accounting (measurement of welfare equivalent income, sustainable income, or net social profit) and two measures (Green NNP and wealth equivalent income) are considered. Under the assumption of no exogenous technological progress, Green NNP is shown to equal wealth equivalent income if there is a constant interest rate or if consumption is constant. It is established as a general result that sustainable income [les ] wealth equivalent income [les ] welfare equivalent income, while Green NNP [les ] welfare equivalent income under no exogenous technological progress and a constant utility discount rate. Green NNP is shown to measure gross social profit rather than net social profit.

Theoretical issues arising in maximin and utilitarian programs are considered in order to shed light on the merits of various concepts of income and types of environmental accounting as guides for environmental policy. The accounting prices for sustaining an economy obey Hartwick's rule but are inconsistent with the principles of national accounting. Moreover, they would be formidably difficult to calculate. Green net national product is an approximate index of welfare in a utilitarian economy which maximises future discounted utility flows. These conclusions hold even if underlying conditions are non-autonomous.

JEL Codes: Q3, E2

For guidance in determining which items should be included in comprehensive NDP (net domestic product) and how they should be included, reference is often made to the linearised Hamiltonian from an optimal growth problem. The paper gives a rigorous interpretation of this procedure in terms of a money-metric utility function linked to familiar elements of standard welfare theory. A key insight is that the Hamiltonian itself is a quasilinear utility function, so imposing the money-metric normalisation is simply equivalent to using Marshallian consumer surplus as the appropriate measure of welfare when there are no income effects. The twin concepts of the ‘sustainability-equivalence principle’ and the ‘dynamic welfare-comparison principle’ are explained, and it is indicated why these two principles are important for the theory of national income accounting.

This paper is about net national product (NNP). We are concerned with what NNP means, what it should include, what it offers us and, therefore, why we may be interested in it. We show that NNP, properly defined, can be used to evaluate economic policies, but we also show that it should not be used in any of its more customary roles, such as in making intertemporal and cross-country comparisons of social well-being. We develop such indices as would be appropriate for making those comparisons. In particular, we show that welfare comparisons should involve comparisons of wealth. Writings on the welfare economics of NNP have mostly addressed economies pursuing optimal policies, and are thus of limited use. Our analysis generalises this substantially by studying economies whose governments are capable of engaging only in policy reforms. We show how linear indices can be used for the evaluation of policy reform even in the presence of non-convexities in the economic environment. The analysis pertinent for optimising governments are special limiting cases of the one we develop.

The literature on green NNP has widely interpreted NNP as ‘constant-equivalent consumption’. We show that this interpretation is wrong. It is the Hamiltonian that equals constant-equivalent utility. Since both theory and empirics imply that the Hamiltonian is a non-linear function of consumption and leisure, the Hamiltonian should not be confused with NNP.

Green income accounting models are designed to appropriately value changes in a country's natural resource (natural capital) base. However, green NNP is useful as a guide for domestic and international policy only to the extent that it accurately reflects the economic goals and policy options of policy makers. For example, international policy designed to slow natural capital depletion in a developing country is more effective if policy makers recognize the developing country's perceived income effects of the policy. Traditional green accounting models do not satisfy this criterion because they are based on the assumption that policy makers are either not concerned with the distributional consequences of policies, and/or are not limited in the instruments available to them. We present an alternative green NNP measure that reflects distributional goals and policy implementation. Using this measure, the depletion (accumulation) of natural capital stocks in excess of economically efficient rates may increase income.

This paper investigates the theory and practice of adjusting national income and product accounts for the stock and depletion of mineral assets. These green income adjustments can have a significant impact on the accounts of mineral-based developing economies and the macroeconomic policies that might be derived from these accounts. We propose that the popular methods used to adjust the accounts for the impacts of mineral assets and depletion are upwardly biased, and we present alternative stock valuation and depletion formulas that are empirically supported variants of the formulas currently in use.

This study estimates depreciation values for timber extraction activities in the Brazilian Amazon for 1990 and 1995. A generalized approach following Vincent and Hartwick (1997) is applied enabling us to calculate depreciation based on all three methods proposed in the literature. We also calculate Hotelling rents for timber as a whole and for mahogany alone. Apart from the expected differences in the results for each method, the results show substantially low depreciation estimates as a direct consequence of the high timber stocks and scarcity perception by economic agents due to the lack of property rights in the region. For mahogany alone we obtain substantially higher values. We conclude that if scarcity rents are not fully perceived, or they really do not exist in such huge supply conditions as our estimates may suggest, charges related to other forest services are fundamental to make environmental accounting an useful tool for planning in the Amazonian context.

The objective of the study is to construct forest resource accounts for the state of Maharashtra in India and incorporate the value of depletion and degradation of forest resources into the system of national accounts (SNA). The net state domestic product (NSDP) is adjusted for the depletion of the forest resources to obtain Environmentally adjusted net state domestic product (ESDP). The results show that the value added by forests is 3.56 per cent of NSDP and the value of depletion is 19.8 per cent of the estimated value added. The ESDP of Maharashtra is found to be 99.3 per cent of the estimated NSDP. The study has demonstrated that, although the existing database needs further improvement, forest resource accounting is feasible for the state of Maharashtra in India and can serve as an indicator of economy's performance.

An environmental accounting approach is adopted to adjust current measures of national income and net savings in SA for the value of net accumulation in timber and carbon stocks as well as for the value of water abstraction externality of cultivated forests. Results indicated that the said values missing from current measures of income and capital formation are substantial, amounting to about 0.6 per cent of NNP, on average over the study period. Potential VAD lost to agriculture due to water abstraction by cultivated forests was estimated at R104 million per annum, on average since 1981. This estimate, however, did not account for the social costs associated with potential losses of environmental services from affected ecosystems.

Previous empirical studies assessing sustainability have adopted the weak sustainability indicator (WSI) which measures changes in the stock of physical and natural capital. Whereas these studies have been retrospective, we use a dynamic general equilibrium model to investigate global and regional sustainability over the period 1985–2050, focusing on fossil fuel extraction. Our standard scenario predicts increasing WSI values suggesting global sustainability in all periods to 2050 as, despite a rising world oil price, the introduction of backstop fuels in 2010 provides an alternative source of energy. A more pessimistic scenario examines possible outcomes in the absence of energy saving technological change and backstops fuels. WSI values still indicate global sustainability, though economic growth is only half that predicted by the more optimistic model, and two countries—USA and Brazil—are unsustainable.