_ _ _ _ _ _ _ __ - I 45 WORLD BANK ENVIRONMENT PAPER NUMBER 2 Sustainable Development Concepts An Economic Analysis FILE COPY Report No.:11425 Type: (PUB) Title: SUSTAINABLE DEVELOPMENT CONCEP Author; PEZZEY, JOHN Ext.: 0 Room: Dept.: BOOKSTORE NOVEMBER 1992 John Pezzey -4, m\ ~~ _ \- ., I A i l,-/ /' I A g 1- ~~~~K& A .,X' , ' " ' iLE C0M "fl RECENT WORLD BANK ENVIRONMENT PAPERS No. 1 Cleaver, Munasinghe, Dyson, Egli, Peuker, and Wencelius, editors, Conservation of West and Central African Rainforests/Conservation de la foret dense en Afrique centrale et de l'Ouest WORLD BANK ENVIRONMENT PAPER NUMBER 2 Sustainable Development Concepts An Economic Analysis John Pezzey This book is printed on recycled paper The World Bank Washington, D.C. Copyright 0 1992 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing November 1992 Environment Papers are published to communicate the latest results of the Bank's environmental work to the development community with the least possible delay. The typescript of this paper therefore has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The findings, interpretations, and condusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member counries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissenunation of its work and will normally give pernission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for cdassroom use is granted through the Copyright Clearance Center, 27 Congress Street, Salem, Massachusetts 01970, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contams an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C 20433, US.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France. John Pezzey is a lecturer in the Department of Economics, at the University of Bristol in the United Kingdom, and a consultant to the Environmental Policy and Research Division, in the Environment Department of the World Bank. Library of Congress Cataloging-in-Publication Data Pezzey, John, 1953- Sustainable development concepts: an economic analysis / John Pezzey. p. cm. - (World Bank environment paper; no. 2) Includes bibliographical references. ISBN 0-8213-2278-8 1. Sustainable development. 2. Economic policy. I. Title. II. Series. HD75.6.P49 1992 338.9-dc2O 92-35724 CIP Foreword The decade of the 1980s has witnessed a ment and Development. While the term fundamental change in the way governments and "sustainability" has been widely used development agencies think about environment since then, little attempt has been made and development. The two are no longer to translate this concept into an regarded as mutually exclusive. It is now analytical framework that can be used in recognized that a healthy environment is the development of "sustainable' essential to sustainable development and a economic policies. healthy economy. Moreover, economists and This paper attempts to analyze the planners are beginning to recognize that concepts of sustainable development, economic development which erodes natural sustainable resource use and sustainable capital is often not successful. Quite the growth in terms of conventional contrary. Development strategies and programs economic analysis, to examine why free which do not take adequate account of the state market forces may not achieve of critical resources-forests, soils, grasslands, sustainability, and to explain how policy freshwater, coastal areas and fisheries--may interventions may help or hinder the degrade the resource base upon which future achievement of sustainability. An growth is dependent. earlier version of this paper was Since its formulation, the Environment published as an Environment Working Department has conducted research and policy Paper and was widely distributed and work on these important issues. The quoted. I am pleased, therefore, to see Department's work has focussed, in particular, it reissued in revised form as a Bank on the links between environment and Environment Paper, so that it may reach development, and the implications of these an even wider audience. linkages for development policy in general. The objective of the Environment Paper Series is to make the results of our work available to the general public. The broad concept of sustainable development was widely discussed in the early 1980s, but was placed firmly on the international agenda with the publication of Our Common Future in 1987, Mohamed T. El Ashry the report of the World Commission on Environ- Director, Environment Department I v Table of Contents Abstract ix Summary xi Part I: Concepts 1 1. Introduction 1 1.1 The growing recognition of sustainable development as a policy goal 1 1.2 Purpose of this paper 1 1.3 Methodology used 2 1.4 Structure of the paper 2 1.5 Some issues to be addressed 3 2. Measuring the economy and the environment 3 2.1 Individual physical quantities 4 2.2 Weights and aggregates 5 2.3 Functional relationships; environmental productivity and amenity 5 2.4 Criticisms of the neoclassical paradigm 6 3. Definitions of growth, development, and sustainability concepts 9 3.1 The context of sustainability 9 3.2 Growth and development 10 3.3 Optimality 11 3.4 Survivability versus sustainability I1 3.5 Sustainable use of renewable resources 11 3.6 Maintaining the effective resource base 13 3.7 Deep ecology and non-instrumental sustainability 13 3.8 Intergenerational equality 13 3.9 Income distribution 14 3.10 Definition of a future generation 14 3.11 Sustainability as non-declining utility or non-declining capital 14 4. Optimal control and sustainability 16 4.1 The role of optimal control models 16 4.2 Optimality and sustainability 18 Part II: Applications 20 5. Economic growth and the environment-balancing consumption and clean-up expenditure 20 6. Non-renewable resources I: sustainability and the discount rate 22 vi Table of Contents 7. Non-renewable resources II: sustainability and environmental dependence combined 24 7.1 The model-cake-eating with enviromnental amenity or productivity 24 7.2 Relevance to policy-can environmental policy help sustainability? 26 7.3 Property rights and environmental policy 30 8. Non-renewable resources III: Ihe role of investment, and technological limits to growth 32 8.1 The model-capital growth with environmental amenity or productivity 32 8.2 Capital-resource substitution, interest rates and technological limits 33 9. Renewable resources: poverty, survival, and outside assistance 35 9.1 The model-corn-eating and subsistence consumption 35 9.2 Possible extensions 37 10. Income distribution and sustainable development 38 11. Are discount rates too high? 39 11.1 Discount rates and sustainability 39 11.2 Changing the demand for investment funds 39 11.3 Changing the supply of investment funds 41 11.4 Interest rates in developing countries 43 12. Information and uncertainty 43 13. Operationality: putting the ideas into practice 45 13.1 To what system should the sustainability criterion apply? 45 13.2 Is a separate sustainability criterion necessary in practice? 46 13.3 Can a sustainability criterion be made operational? 46 14. Conclusions and suggestions for further work 48 References 50 Appendix 1 Definitions of sustainability in the literature 55 Appendix 2 Cake-eating model with no environmental effects 63 Appendix 3 Cake-eating with environmental amenity or productivity 64 Appendix 4 Capital accumulation with environmental amenity or productivity 67 Appendix 5 Corn-eating with a minimum subsistence level 70 Tables 1 Vector notation used in the economy-environment model 4 2 Possible definitions of growth, development, and sustainability 10 Table of Contents vii Flgures 1 Economic and environmental stocks and flows-a general model 7 2 A purely economic model 8 3 Optimality, sustainability, and survivability 12 4 A developing country model, totally dependent on renewable resources 17 5 Economic growth and the environment: a static model with clean-up expenditure 21 6 Possible trade-offs between output and environmental quality 23 7 A cake-eating model with no environmental effects 25 8 A cake-eating model with environmental amenity 27 9 A cake-eating model with environmental productivity 28 10 Sustainability, optimality, and government intervention 29 11 Effect of a lower effective discount rate on initial utility 31 12 A capital accumulation model with environmental amenity 34 13 A corn-eating model with subsistence consumption 36 14 Removing investment income tax lowers the interest rate 40 15 Tightening environmental policy may lower the interest rate 42 16 Increased saving lowers the interest rate 44 ix Abstract This paper attempts to analyze sustainability The suggested implications for policy are concepts, such as sustainable growth, sustainable that conventional environmental policies may development and sustainable resource use, in also improve sustainability, making a separate terms of the conventional neoclassical theory of sustainability criterion redundant in practice; and economics. It then tries to analyze why free that politically difficult short-term sacrifices may market forces may not achieve sustainability, be needed to reach optimal and sustainable and how policy intervention may help or hinder growth paths. "Conventional environmental sustainability. policies" need not always mean making the Several different definitions of sustainability polluter pay for externalities. More important is are reviewed. Most require that the 'quality of that property rights over the environment are life" should not decline over the long-term first defined and enforced, if this is possible. future. Many can also be interpreted in terms of A simple model with renewable resources maintaining an economy's capital stock. shows how population growth can threaten However, a relevant definition of capital stock sustainability, and how poverty and still has to be chosen, and this means judging environmental degradation can be linked, how significant, essential or substitutable are the establishing the case for development assistance. various natural and man-made resource inputs to The role of a more equal income distribution as the economy's production processes. part of sustainable development is briefly A number of simple models of the economy discussed. Giving environmental property rights and the environment are used to explore these to the poor may both reduce poverty and issues. One model uses comparative static improve the environment. This is true whether analysis to explain why different tradeoffs may the poor are the polluters (by being so poor that be made between consumption and they degrade their own land and cause floods, environmental quality at different stages of siltation, etc elsewhere), or the pollutees (by economic growth. Other models use optimal suffering air and water pollution, in cities). control theory to examine sustainability over If improved environmental policy alone is time in the context of both non-renewable and not enough to achieve sustainability, so that a renewable resources. Such models may not separate explicit sustainability policy is achieve much realism, but they should help to necessary, one must ask how it can be applied at clarify conceptual thinking about sustainable both the system and project level. At the system development. level it is fairly clear, at least in theory: The main results suggested by these simple aggregate constraints (either economic or models are that, if non-renewable resource regulatory) must be imposed to control the inputs are essential, then inadequate technical depletion of whatever resources have been progress and open access to environmental determined to be important for sustainability. resources may be the key factors that cause Such constraints should drive up the price of unsustainability. Government intervention, in the such resources to whatever level is necessary to .form of resource conservation subsidies or induce the required conservation efforts depletion taxes, is shown both to correct the throughout the system. Such efforts are open access problem and to improve equivalent to intergenerational compensation sustainability; conversely, government subsidies investments. An example of this kind of process for resource depletion aimed at encouraging is already under way with international development will harm sustainability. But agreements to reduce the use of CFCs. improving sustainability by slowing down Making sustainability operational at the resource depletion also may mean lower initial project level is much harder, even conceptually. levels of consumption and utility. System sustainability cannot be disaggregated x Absta into project rules in the simple way that system harmful overall effect on environmental optimality can be disaggregated into cost-benefit resources. This is an attractive and fairly rules for project appraisal. Many writers suggest operational concept, but it is not clear how that sustainability will be assured, if groups of projects should be defined, how 'intergenerational compensation" projects are compensation should be paid, and who should required for any group of projects that has a pay it, particularly in the private sector. xi Summary The paper attempts to analyze sustainability resources. Looking objectively at the concepts, such as sustainable growth, sustainable resource base may be more relevant development and sustainable resource use, in than notions of intergenerational terms of conventional neoclassical economic welfare, when studying poor theory. It then tries to analyze why free market developing country economies. forces may not achieve sustainability, and how (5) Many definitions of usustainable policy intervention may help or hinder development" explicitly reqluire sustainability. Many of the concepts and ideas attention to the needs of the cuirent appear in 'Environment, Growth and poor as well as to the needs of the Development" (World Bank Development future. Committee Pamphlet 14). [Section 1] Many different definitions of sustainability We set out a general model depicting can be interpreted in terms of maintaining the stocks and flows of economic and environmental economy's capital stock. However, capital stock variables such as capital, labor and natural can also be defined in several different ways, so resources (hereafter abbreviated to just a choice of definition is still necessary. The 'resources"). Each variable consists of many relevant choice requires a judgement of how different categories, and weights such as prices significant, essential or substitutable are the or natural resource accounting values are needed various natural and man-made resource irnputs to calculate aggregate stocks and flows of into the economy. [Section 3/Appendix 1] variables. General interdependencies are pointed The uses and shortcomings of abstract out, such as the effect of resource and pollution optimal growth models for analyzing sustainable stocks on output ("environmental productivity") development are discussed. Optimal growth and on welfare (environmental amenity"). models can never achieve much realism, but Criticisms of neoclassical assumptions inherent may be useful for clarifying concepts and for in the model are discussed. [Section 2] making general suggestions for policy in what is Dozens of different verbal definitions of a very diverse and complex field. Sustainability sustainability concepts are listed in an appendix. may be viewed as a constraint on the The general economy/environment model is used conventional optimality criterion of maximizing to suggest quantified, sometimes conflicting discounted utility, rather than as a replacement interpretations of these definitions. Key points for it. Providing an ethical foundation for a which emerge are: sustainability constraint requires that people are (1) The geographical and temporal seen as having separate preferences for private context for sustainability concepts and social choices. In practice governments may must always be made clear; be no more concerned about sustainability than (2) 'Growth" generally ignores the individuals. [Section 4] direct effect that the environment Comparative static analysis is used to may have on social welfare, whereas analyze rational tradeoffs between consumption 'development" takes it into account. and environmental quality at different stages of (3) The most common, although economic growth. Resource inputs are ignored, subjective, definition of 'sustain- and a given output is assumed to be divided ability", is that the welfare of future between consumption and pollution control generations should not be less than expenditure. It is possible to view a commonly the welfare of the current generation, observed pattern, that environmental quality first i.e. utility should be non-declining. declines and then recovers as industrialization (4) 'Sustainable resource use" focuses proceeds, as an optimal allocation of resources. on maintaining a stock of renewable One can then perhaps conclude that continued xii Summary environmental improvement is generally conventional environmental policy may also compatible with economic growth in the mature improve sustainability, making a separate stages of development. However, it is also sustainability criterion redundant in practice; and possible that environmental policy is inevitably that politically difficult short-term sacrifices may weak during early industrialization, and that be needed to reach optimal and sustainable truly optimal consumption-environmenttradeoffs growth paths. 'Conventional environmental actually lead to continually declining policy" need not always mean making the environmental quality as growth proceeds. polluter pay for externalities; more important is [Section 5] that property rights over the environment are Optimal control theory is applied to first defined and enforced. [Section 7/Appendix radical simplifications of the general 3] economy/environment model, in order to The third model looks at steady states of examine sustainability in the context of non- an economy which also uses accumulated capital renewable and renewable r esources. as well as resource flows to produce output (via Mathematical details are given in corresponding a Cobb-Douglas production function). appendices. The first model is of "cake-eating", Environmental amenity or environmental with exogenous technical progress in the productivity, combined with privately optimal transformation of a single non-renewable resource depletion, again results in socially resource into a consumption good. The optimal excessive resource depletion rates and lowered path shows steady growth of consumption (i.e. sustainability. A government conservation sustainability) only if the rate of technical incentive again slows resource depletion. It also progress exceeds the rate at which future utility raises the rate of return on capital (the interest is discounted; so people's concern for the future rate), because the resource price is driven up does affect sustainability. [Section 6/Appendix 2] and capital investment results in resource The second model is also of cake-eating, savings. Possible limits, imposed by the laws of but here an individual's utility depends not only thermodynamics, on capital substitution for on the rate at which he depletes his own resources (limits which the Cobb-Douglas resource stock, but also on the total resource formula does not recognize) and on stock owned by all individuals. This total technological progress are briefly discussed. resource effect is either direct (environmental Such limits, combined with the finiteness of amenity) or via the production function global resources of materials and energy, may (environmental productivity). In either case, ultimately constrain economic growth. [Section non-cooperative (privately optimal) resource 8/Appendix 41 depletion results in a "tragedy of the The fourth simple model is based on a cormmons": private rates of resource depletion single renewable resource ('corn"), where there are greater than is socially optimal, and the is exogenous population growth, no technical economy is less sustainable. progress in corn-growing and harvesting, no Government intervention in the form of environmental externalities, and a minimum resource conservation subsidies or depletion consumption level needed for survival. This is taxes is shown both to correct the tragedy of the clearly more relevant to developing countries commons and to improve sustainability. whose economies depend largely on renewable Conversely, government subsidies for resource resources. The optimal solution can be one of depletion, as often occur in reality, have the sustained growth of consumption and welfare, opposite effect. However, slowing down but only if two conditions both hold. The first is resource depletion also means lower initial levels that the resource growth rate exceeds the sum of of consumption and utility. The suggested (not the utility discount and population growth rates; proved) implications for policy are that if not, grinding along forever at subsistence Summary xiii consumption is optimal. The second condition is resource saving, and reduces total investment that the initial resource growth is large enough demand. On the supply side, there is a purely to feed the initial population; if not, people are moral case that people ought to care more for forced to eat resource capital (seedcorn) simply the future and lower their utility discount rate. to survive, and total depletion and catastrophe There is also an economic case: investment are the inevitable result. This model provides a supply will be too low (and interest rates too rationale for common development policies such high) if people are unaware of the ultimate limits as agricultural improvement, population control to growth, and hence have excessive and the need for outside assistance. Possible expectations of consumption growth rates in the extensions to include capital accumulation, non- distant future. [Section 11] renewable resources and environmental The importance of imperfect information externalities are suggested. [Section 9/Appendix is recognized but not analyzed; people mayr be 5] depleting resources unsustainably without We then discuss the roles of a more equal knowing it. Uncertainty about the future is also income distribution, and/or of meeting basic important, and the possibility of environmental needs, in sustainable development. To some catastrophes may justify greater environmental extent these may be separate issues, requiring protection as a form of insurance. [Section 12] separate redistribution policies. However, the Finally, we look at how a sustainability corn-eating model showed how poverty and criterion (if it is accepted as a social goal) may environmental degradation can be linked. Also, be made operational. First of all one must: be the allocation of environmental property rights to clear about what system level (species, the poor may both alleviate poverty and improve ecosystem, nation or planet) the sustainability the environment if there are both rich and poor criterion is to be applied to. Then one must ask classes in society. This is true whether the poor whether a separate sustainability criterion is are the polluters (by being so poor that they necessary in practice: it will be very difficult to degrade their own land and cause floods, apply, and in any case improving conventional siltation, etc elsewhere), or the pollutees (by environmental policies will generally improve suffering air and water pollution in cities). sustainability as an automatic side-effect. [Section 10] If it is decided that improved environ- Observable interest rates, as distinct from mental policy alone is not enough, and a unobservable utility discount rates, clearly affect separate sustainability policy is necessary, one sustainability, since they determine the relative must ask how it can be applied at both the weight given to present and future costs and system and project level. At the system level it benefits in discounting procedures used to make is fairly clear, in theory if not in pract:ice. investment decisions. We examine the case that Aggregate constraints (either economic or real interest rates are "too high" by looking at regulatory) must be imposed to slow down or the interest rate as the balance of demand and halt the depletion of whatever resources have supply for investment funds. On the demand been shown to be important for sustainability. side, it is argued that environmental policy may Such constraints will effectively drive up the lower demand and thus reduce the economy's price of such resources, to whatever level is interest rate. This depends on assumptions that necessary to induce the required conservation (1) "resource-using" investments are much efforts throughout the system. These commoner than "resource-saving" investments; conservation efforts will be equivalent to (2) resource use results on balance in harmful providing intergenerational compensation in externalities. Using environmental policy to various ways. Moves toward this process are internalize these externalities therefore drives up already clear with international agreements on resource prices, shifts investment towards CFCs. xiv Summary Making sustainability operational at the same core ethic of intergenerational equity. project level is much harder, even conceptually. Choosing a sustainability criterion that is System sustainability cannot be desegregated into appropriate to a given policy context requires project rules in the simple way that system judgments on which natural and man-made optimality can be desegregated into rules for resources are significant inputs to production and cost-benefit analysis of projects. Many writers welfare, and on how essential and substitutable suggest that 'intergenerational compensation" they are. The notion that conventional projects should be required for any group of environmental policies may improve projects that has a harmful overall effect on sustainability is important. Suggestions for environmental resources. This is an attractive further work include more sophisticated models and fairly operational concept, but questions are of growth with renewable resources; more raised as to how groups of projects are to be analysis of how both poverty and environmental defined, how compensation should be paid, and policy can affect discount rates; more work on who should pay it, particularly in the case of the theory and practice of intergenerational private investments. [Section 13] compensation mechanisms at the project level; Conclusions include the following. Many and more work on uncertainty and sustainability criteria are derivable from the irreversibility. [Section 14] l Part I: Concepts 1. Introduction management are in many cases mutually consistent objectives." (World Bank 1988, 1.1 The growing recognition of sustainable pl). development as a policy goal 1.2 Purpose of this paper Sustainable development, and the interdependence of the economy and the But what exactly is meant by various environment, are increasingly important concepts sustainability concepts such as sustainable to policymakers around the world. The concepts development, sustainable economic growth, and grew out of the 'Limits to Growth" debate of sustainable resource use? Can one se ca non- the early 1970s (Meadows et al 1972, Cole et al renewable resource sustainably, or is the concept 1973), which discussed whether or not limited to renewable resources? Does continuing economic growth would inevitably sustainability necessarily imply a more equal lead to severe environmental degradation and distribution of income within the current societal collapse on a global scale. By the late generation, as well as between generations? Is 1970s and after much further debate (e.g. sustainability meaningful for developed as 'well Pirages 1977, Cleveland 1979, Coomer 1979), as developing nations, and at the global or l,ocal an apparent resolution of the problem was as well as national levels? Theanswers to these reached: economic development could be questionslarenot atdall clear. InsAppendix i we sustained indefinitely, it was held, but only if uave collected dozens of published definitions of development is modified to take into account its sustainability concepts. The diversity of and ultimate dependence on the natural environment. conflicts between these definitions is self- This broad concept of "sustainable evident, showing that sustainability is fast development" was first widely publicized by the becoming a 'motherhood and apple pie" World Conservation Strategy (IUCN, 1980). It concept, which everyone supports but no one has since become central to thinking on defines consistently. Indeed, 'it may only be a environment and development, and is espoused matter of time before the metaphor of by many leaders of world stature. Notable recent sustainability becomes so abused as to beoome examples are the report of the World meaningless" (O'Riordan 1988, p30). Commission on Environment and Development Certainly, using a sustainability concept (WCED 1987-the "Brundtland Report"), and without providing a fairly detailed definition can the landmark World Bank paper uEnvironment, easily lead to misunderstanding and confusion. Growth and Development" (World Bank 1987). This paper therefore sets out to: The Brundtland Report vigorously promotes the (1) categorize the various sustainability idea of sustainable development, which it defines definitions in formal terms; as: (2) analyze the circumstances which may "Sustainable development is development result in the various concepts of that meets the needs of future generations sustainability not being achieved; without compromising the ability of future (3) analyze policies that might achieve generations to meet their own needs" sustainability (henceforth we will use (WCED 1987, p43) and the World Bank is sustainability" as an umbrella term to now committed to promoting sustainable cover a number of concepts; the development and to the proposition that: precise concept intended will be made "economic growth, the alleviation of clear when necessary). poverty, and sound environmental 2 Part l: Concepts A particular feature is that the paper tries or build realistic simulation models of to provide a comprehensive view of sustainable development, quite apart from the sustainability which is applicable to developed as difficulty of collecting the necessary data, until well as developing countries, and to non- concepts are better defined and understood by renewable as well as renewable resources. looking at simple models. Even if somehow our models could be 1.3 Methodology used much more complete and realistic, they would still not be immune from criticism of a quite We do all this by building abstract models of different kind from the above remarks by Ayres. optimal economic growth and development, Some writers question not whether neoclassical using concepts from conventional methods are tractable, but whether it is even ("neoclassical") economic analysis. Some of appropriate to use neoclassical concepts in the these concepts are more or less measurable (such context of sustainable development and as output, consumption, natural resources', and intergenerational equity. We discuss this in capital) and some are inherently non-measurable Section 2.4, after we have expounded our (such as utility and social welfare). The models general neoclassical model. are analyzed mathematically to explore how Whether or not it is necessary to have these concepts relate to each other and how they models of sustainable development at all is yet grow over time. In particular we wish to another matter, one we defer till Section 7 for distinguish circumstances where economic reasons that will become clear. One good reason development and environmental protection are to try is surely to limit the kind of complementary, from circumstances where terminological abuse that would debase trade-offs have to be made. sustainable development into a useless phrase, as However, all simple economic models, O'Riordan (1988) fears. particularly optimal growth models, have several defects, as Robert Ayres has pointed out: 1.4 Structure of the paper "Often the specified conditions are far The structure of the paper is as follows: from realistic, and the practical value of * Section 2 sets up a general quantitative the exercise is slight until a great many model of the economy and the simplified models, based on different environment, and discusses possible assumptions, have been examined and the objections to the neoclassical results compared. Even then, the truly assumptions built into the model. generalizable statements are rare-and * Section 3 uses the general model to always subject to modification as a result define the various sustainability of analysis of the next such small model. concepts, stressing the difference Regrettably, academic economists not between survivability and improve- infrequently generalize too freely from the ment, and the common idea of results of ultra-simplified models." (Ayres sustainability as maintaining capital. 1978, p v; see also Koopmans 1977, p265) * Section 4 discusses the uses and shortcomings of optimal growth We quote this at length so that the reader methodology in analyzing economy/ will not be disappointed by what follows. The environment models and sustainability. assumptions are far from realistic; no numbers * Sections 5-9 all use radical appear; results are only generalized simplifications of the general model: usuggestions"; one cannot go on from this paper * Section 5 looks at economy/ to say whether or not Burkina Faso (or environment tradeoffs in a static wherever) is developing sustainably, and if not, world, where resource inputs are why not and what policies would make ignored, and a given output is divided development sustainable. But we contend that it between consumption and pollution is impossible to construct sustainability indices control expenditure. Part I: Concepts 3 * Section 6 (with mathematical details in 1.5 Some issues to be addressed Appendix 2) assumes a single, non- renewable resource input (cake") and To whet the appetite, the following issues will a given rate of technical progress, and be addressed, in addition to those inherent in the compares sustainability and optimality above section descriptions: as criteria for allocating resource * Is sustainable development a process consumption over time. or a state? (Sections 3, 10) * Section 7 (Appendix 3) builds on * Does sustainable resource use require Section 6 by allowing output or utility that every resource stock must not to depend on the total resource stock, decline? What is the relevance of and looking at the effects of natural resource accounting? Is government policy instruments on sustainable resource use a means or an conservation and depletion. end? (Sections 3, 9) * Section 8 (Appendix 4) introduces * What distinguishes a case where man-made capital, and discusses the economic growth and environmental substitution between capital and improvement are mutually consistent resources and possible limits on this objectives from one where they are substitution. not? (Sections 5, 7) * Section 9 (Appendix 5) assumes a * How much can sustainable single renewable resource ("corn") development be furthered by merely and a minimum consumption level defining and enforcing property rights needed for survival, and discusses the over the environment, and how much interaction of poverty, resource do such rights need to be changed? degradation and outside assistance. (Section 7) * Section 10 addresses the role of a * Will lower discount rates lead to too more equal income distribution and of little or too much investment for meeting basic needs as part of sustainable development? Is investment sustainable development. good or bad for the environment? * Section 11 asks whether discount rates (Sections 6,8,11) can be too high to achieve sustainability, and distinguishes the 2. Measuring the economy and moral and economic cases for lower the environment discount rates. * Section 12 briefly looks at the way Here we set out a formal and fairly general imperfect information, particularly model of the economy and its surrounding inherent uncertainty about the distant environment. To address sustainability issues it future, can affect sustain-ability. is necessary to include the environment, even if * Section 13 looks at whether or not nearenotinterestheienvironentalevesf sustainbility onceptscan be we are not interested in environmental issues. for sustainability concepts can be their own sake. If instead we use a conventional translated into operational procedures 1960s model of economic growth, in wlhich at the economy or the project level, output is produced from just capital and labor mechanisms for aintergenerational inputs and is freely disposed of after use, we compensation". have little reason to suppose that sustainability * Section 14 draws some conclusions should ever be a problem (unless savings rates and suggests directions for further are insufficient to maintain the capital stock). work. The resulting complexity of our model is One prominent issue that is hardly touched daunting, but is needed in Section 3 in order to Oneproinentssuetatishrdlytuched analyze the dozens of different definitionzs of upon is sustainability and international trade. We anaylity listedifferenditions of hope to address this in a later paper. sustainability listed in Appendix 1. We ithen revert to much simpler models for later sections, but the greater realism and complexity of this 4 Part I: Concepts section should be borne in mind throughout. The or values and the resulting aggregates or totals importance of aggregation problems is in Section 2.2, and some general functional emphasized by desegregating output, labor, relationships in Section 2.3. natural resources etc into several different classes. The number of classes (m) may vary, e.g. ml different goods, m2 different resources, 2.1 Individual physical quantities etc. Table 1 sets out the fairly obvious vector notation that is used in the model. In the model there are ml types of output, with Note the difference between the vector and total outputs g divided into consumption c, aggregate scalar measures, for example with investment in physical capital (machines, etc) k, regard to natural resource stocks. Non-declining investment in technology and human capital i, f means that every resource stock is constant or and clean-up expenditure x: growing; non-declining S means only that the aggregate resource S, computed using a vector Q = - + k + i + x of 'resource weights" y such that S = y.s, is conserved; individual resources (e.g. plant or Investments jk accumulate to form physical animal species) may be declining, perhaps to capital stocks k, and investments i accumulate to form stocks of technical knowledge 1. We assume that all outputs are 'goods" in the sense Table 1 that they do not represent costs or defensive Vector notation used in the expenditures. We thus abstract from the frequent economy-environment model criticisms that GDP measures include such items in practice (see for example Daly 1988). £ (vector) means {cl,c2, . ..,c}; There are m2 types of natural resources also written {cj}; j 1,.. ,m (usually abbreviated to just uresources"). And if p Resource stocks a are classified into L, the first is also a vector of m numbers, mn, resources which are renewable ('alive"), then and L, the last (m2-m2.) resources which are D D.C (= C say, a scalar) means non-renewable, thus: E pjcj; j = 1,...,m s = {(as} - min c (scalar) means min {c5}; j = 1,...,m The corresponding resource flows are * non-declining c (vector) means similarly classified into renewable and non- c 2 0, for all j = 1,..,m and for renewable components: r = {r,,r). Now many renewable resources are of no all time direct instrumental value to man, either as * non-declining C (scalar) means sources of amenity or sources of economic C = d( I pjcj)/dt 2 0, for all production. However, the intricate cycles of food, energy and nutrients within ecosystems time ensure that many non-instrumental species and resources are vital to the existence of other extinction. The ultimate purpose of resource species that are directly valuable. Our measure accounting is to measure not just s but also y, so of renewable resource stocks s must therefore that we can say something meaningful about be extended to include not only the valuable aggregate resources. Here we ignore the species (for example, trees in a forest) but also problem of just how difficult such measurement the resources necessary to support them (for is in practice, and assume that we can define all example, soil and bacteria to provide nutrients, the individual quantities needed to describe the insects and birds to provide pollination). economy physically in Section 2. 1, the weights There are m3 pollution stocks p and flows 5 ("disposal"). Part I: Concepts 5 Lastly, there are m4 groups of people. We 2.3 Functional relationships; environmental make no distinction here between population, productivity and amenity numbers of households and numbers of workers, and denote the population groups by the vector Most relationships between variables in the -. model properly be describedL by vextor Note two major simplifications in the relationships. For example, the population above: imports and exports are omitted, and growth of one species (an element of L) would natural resources are not subdivided into depend on the stocks of many other species and materials and energy. on inanimate resources and pollution (other elements of s, s, and p). However, for the rest 2.2 Weights and aggregates of this paper, there is no point in using vector and matrix algebra to describe these Note: assuming that all these weights relationships, so below we shall just describe exist-particularly weights for aggregating scalar relationships between aggregates; buit this individual utilities into social welfare-bypasses is already a simplification. A dot over a symbol enormous theoretical and practical problems. represents a time derivative, and a + or - albove There are m, goods prices g, hence Q = C an independent variable shows whether or not + Ik + Ir + X where: increasing that variable increases or decreases * aggregate output Q = g.q the dependent variable. All functions are * consumption C = g._ assumed to be continuously differentiable with * capital investment Ik = g.i respect to their arguments. Figure 1 depicts * technology investment I, = g.i aggregate relations in our general economy- * clean-up X = g.x environment model. "Stripped-down" versions * capital K = g.k of this Figure appear in several later sections to * technology T = g-l illustrate the more restricted models used there. The production of output depends on There are m2 resource values v = {IL,vl capital, labor, resource inputs, technology, and hence also on the 'state of the environment". The * aggregate resource stock S = y.s environment, which consists of the stocks of * renewable resource stock Sa = y.L resources S and pollution P, can clearly be * non-renewable resource aggregated in many different ways, and we often stock S. = L,, find it convenient to use El to represent: the * resource flow R = v.r environmental aggregate that affects production. * renewable resource flow R. = L.,, E2 will be a different environmental aggregate * non-renewable resource that affects amenity values, or we may also stock R. = LS- sometimes ignore this difference between There are m3 'toxicity coefficients" or environmental aggregates and simply use an all- other measures of waste harmfulness h, hence purpose measure E of environmental quality. * aggregate pollution stock P = h.p The dependence of production on the * pollution flow D = 1_. environment is hereafter called 'environmental Where convenient, we may also use m2 productivity". resource environmental values e to compute: + +.+ + + - overall environmental quality E=e._-P. Q = Q(K,L,T,R,S,P), There are m4 measures of labor productivity z, per capita consumption w, per which is simplified to capita utility u and class utility weights f, hence * aggregate population N = ..l. +++++ * labor input L = z.l Q(K,L,T,R,El) in Figure 1 * consumption C = w.! = g-c * social welfare U = f.u The growth of both capital and technology equals gross investment minus 6 Part I: Concepts depreciation (the depreciation coefficients ak It is worth noting in passing that it has and 6, are unrelated to the social utility not always been accepted that environmental discount rate 6 in Section 3): resources have a significant role to play in models of economic growth. Until the general K = Ik- dkK T = If - 6,T reawakening of environmental awareness in the late 1960s, economic output was The growth of resources equals natural considered to require only capital, labor and growth (which is zero for non-renewable technology inputs (with technology recognized resources) minus resource extraction: as a separate input only in the mid-1950s). + - Figure 2 represents the kind of model used S = G(S,P) - R then, and Burmeister and Dobell (1970) give a good summary of growth models of this type. The growth of pollution stock equals the In Section 3 we examine what objective rate of waste disposal minus an assimilation function(s) the optimal control process is term which represents both natural meant to maximize, and what constraints it is assimilation and the ameliorative effects of meant to observe. clean-up expenditure: In Section 4 we then discuss the extent to + + which these functional relationships can be P = D - A(P,X) analyzed using optimal control models, and what results can be drawn from such analysis. The growth of the human population However, we must first address some depends on natural growth, and in a complex fundamental criticisms of the very form and way on consumption income: concepts of our model. N = N(L,C) 2.4 Criticisms of the neoclassical paradigm Labor productivity can be affected by The above model incorporates a wide range of consumption and the environment assumptions from neoclassical thinking that many have challenged, both-generally and in + + - discussions specifically on intergenerational L = L(C,S,P) equity and sustainable development. These criticisms focus particularly on the form of the Lastly, "utility" or usocial welfare" utility or social welfare function U(.), but also depends not only on consumption2, but also on the production function Q(.). on the state of the environment. This is the (a) Many writers on sustainability "envronmental amenity" effect: question the inbuilt assumption that the functional form U(.) is + + - determined exogenously, i.e. that U = U(C,S,P), which is simplified to tastes and preferences appear when + + we are born and are not formed by U = U(C,E2) in Figure 1. culture, education or advertising. For example, WCED (1987, quote 4) In practice, the boundary between the holds that "perceived needs are amenity and productivity effects of the socially and culturally determined", environment is not always clear because the and that: boundaries of commercialization is variable. If "The changes in human attitudes I refrain from strolling in a public park that we call for depend on a because the day is smoggy, that is an amenity vast campaign of education, debate effect; but if I refrain from paying to enter an and public participation." (WCED amusement park for the same reason, that is a 1987, p23). productivity effect. r Environmental Quality E _= Capital Technology Natural c 0 Stock ~~Stock Resource l: __ Storcks | asou Pollution I Renewable Non- Stock Resources Renewable L 3 T Sa Resource es Environmiental S I Productivity' Technology IR IF n Services _ _ ResourceI Waste Output + Fws 11 ~~~~~~~~~~~~~~~~~D I Services I~~~~~~~~~Sevie sumptlon HOUSEHOLDS c Population - N I o Capital Investment 'K C Envirotnnlt-t Technology Investment IT Amenity Clean-up Expenditure X 8 Part I: Concepts Figure 2 A purely economic model Capital Technology Stock Stock K T Technology Servlces T Capital Services K Capital Investment ' PRODUCTION T ieLabor | .2 ~~~~~Services Output J C on sump tion HOUSEHOL g// C IA Populatlon N! Capital Investment IK / tL_ uttity Technology Investment IT /' ' Part I: Concepts 9 (b) A related point is made by the generations, and offer them increased evolutionary school of economics (see man-made capital and technological for example Norgaard 1984, 1985). knowledge (enough to lead to This school holds that the forms of increased production, even though economic functions change irreversibly natural resources will be depleted) as over time, in a waythat depends upon compensation? Page answers no:: the the future path of the independent natural resource base cannot be 'justly variable C, E, etc. This holds for both acquired" through human labor in the utility and production functions. way that machines can. Therefore we (c) Another questionable assumption in are not morally free to treat natural U(.) is that individuals derive resources as mere factors of happiness from absolute, rather than production that can be depleted and relative, levels of individual substituted for by man-made capital. A consumption and environmental related criticism is that the pure time quality. If position relative to others is discounting, as measured by the utility all that really matters (as suggested in discount rate a in the following some exploratory work by Easterlin section, has no moral justification. 1974), changes in aggregate variables Criticisms of time discounting are: also C, S and P would have little effect on made by Parfit (1983) and are aggregate utility U. Similar problems reviewed by Markandya and Pearce occur if change relative to expectations (1988a, 1988b). is what matters (Rescher 1980). These criticisms can only be listed here, (d) Many authors would criticize our since by definition they cannot be incorporated utility function for leaving out in our neoclassical model. The questions of important, if inherently unquanti-fiable substituting capital for resources arise again in variables such as 'cultural disruption Section 3.11, but there the discussion focuses and social instability" (Barbier 1987) not on whether such substitution is morally and "basic freedoms" (Pearce, Barbier allowable, but on the extent to which it is and Markandya 1988). physically possible. (e) Yet more fundamental criticisms of our neoclassical functions are that the 3. Derinitions of growth, developmient, trade-offs or substitutions that they and sustainability concepts allow may be psychologically impossible or morally inexcusable. 3.1 The context of sustainability Page (1983) sets out these criticisms at length. The existence of a lifferengthabl. Th CExitnceiof a We can now use the concepts from the above differentiable an Cae functi ileos economy-environment model to attempt precise tuchat wercang makehow menl etradf neoclassical definitions of growth, development consuchpasworing w outd h much extra and sustainability concepts. Table 2 lists a consumption we would require to . compensate for a substantially selection of definitions, inspired by the increased risk of cancer due to higher quotations collected in Appendix 1 (most of the radiation levels in the environment, references below are to these quotations); cl[early many more definitions are possible. Page questions whether or not it is Sustainability has been applied to a vast osesible. toranser, sche questions array of situations, ranging from the conditions shensibly Morenovr the qurrengestions for success of a World Bank agricultural w hether orit th curent eneration development project to the problem of creating behasteg to attem g tionswet conditions for the improvement of the situation ubehal. ofte gehergtiton as yet of the whole human race in the "further future" the natural envirorinent for future (Kneese and Kopp 1988). Clearly an appropriate criterion for sustainability will depend very 10 Part I: Concepts much upon the context, especially if it is to be Table 2 gives three possible definitions of used operationally. An attempt is made in development, two of development as a process, Section 3.11 to show how diverse sustainability one as a state. Many other definitions are criteria can be derived from unifying general possible. In this formal context we cannot concepts such as 'maintaining the stock of represent ideas, such as in Daly (quote capital" or uensuring non-declining utility". 2),Georgescu-Roegen (quote 1) or Boulding Three remarks apply to almost all (1988), that development is qualitatively sustainability criteria: different from growth. (a) They are long term criteria. Although asustaining economic growth" by using skillful macroeconomic Table 2 management to avoid short term cycles Possible definitions of growth, development, of unemployment, inflation, and trade and sustainability deficits is clearly of prime policy See Section 2 for most notation. Additional importance, it is not our concern here, and we assume throughout that all notation: 5> 0 is social discount rate; t is time; of prdcto are - nl T > 0 is lifespan of a generation; and subscripts factors bn = 'basic needs"; sub = 'subsistence' es = employed. aclgclyssanbe (b) Most criteria derive from a common "ecologically sustainable" school of ethical principles regarding All survivabilit and sustainability concepts intragenerational and/or intergene- applyfor all time. rational fairness or justice. The biggest Economic growth = increase in Q or C single inspiration for this school has Development (process) I = increase in U = U(C,S,P) undoubtedly been the work of Rawls Development (process) H = increase in w Development (state) MI =miin w > Wb. (1971). However, other ethical views Optimal path = path maximizing of intergene-rational justice exist and should not be overlooked; Pearce f0U(t)e t'dr (1983) and d'Arge (1989) give useful analyses. Survivable growth = min w > w,,b (c) Sustainability criteria are mostly Sustainable growth I = non-declining Q (or C) mathematical inequalities and are Sustainable growth H = positive and non-declining therefore constraints, rather than maximizing criteria like optimality. ~~QIQ or CIC mnaximizing criteria like optimality. Survivable development = min w > w,b and 0 < G. We explore this difference in Sect- Sustainable development I = non-declining U ion 4.2. Sustainable development l = non-declining U, and with miin w increasing Sustainable development HI = non-declining U, and with miin 3.2 Growth and development & > w,, and maxw < w, Sustainable resource use I = non-declining & Economic growth is uncontroversially defined Sustainable resource use U = non-declining S. Sustainable resource use Ml = non-declining S as rising aggregate consumption C or output Q. Sustainable resource use IV = non-declining & and non- As long as the average propensity to consume increasing p (C/Q) is constant the distinction is unimportant. Lifetime sustainability = instead of non-dcining X Note that growth is measured in value, not (whatever X is), non-declinig physical units: a growth of economic output does not necessarily mean a growth in physical f 1X((ea1'd; throughput of materials and energy. The t problems start in defining development to make i.e. present value of X for up for the shortcomings, identified by Redclift ge.nrtion at time t with a fim (1987) and countless others, that growth ignores horizon stretching T years into environmental quality and other social factors, the future. and also ignores the distribution of income. Part I: Concepts 11 3.3 Optimality However again, this is not the end of the matter. One important advantage of survivability The conventional optimality formula of is that it is objective. One can in principle make maximizing the present discounted value of physical calculations of the minimum amounts of utility is widely accepted, although it does make food, shelter, clean air and water needed to' keep sweeping assumptions: for example, that an economy of a given number of people alive, intertemporal preferences are consistent (Strotz, or, somewhat more comfortably, provided with 1956) and well-known into the distant future. It "basic needs"; whereas improvements in utility can be applied to either growth, if U = U(C), or the 'quality of life" will always be non- or development definition I in Table 2, if U = market value judgments and thus much harder to UCSP Teissencpmeasure operationally. For poor, natural U(C,S,P). There is a semantic problem in that resource-based countries the minimum sonme writers regard "optimality" as defining the suppliesmust be provided the ecosystem. (For ultimate social goal, and would incorporate any other countries this is less clear: many developed other constraints such as sustainability into their countries use exports of manufactures and definition of optimality. This is understandable, services to buy necessary food and raw material but confusing for our purposes; here we restrict imports, althotigh to suggest that every country "optimality" simply to mean present value could do this would clearly be a fallacy of maximization. The time horizon chosen can be composition.) We can then calculate, subjject to finite, although we stick to infinity here. Most detailed scientific debate, the minimum size and dynamic optimization models ignore composition of the ecosystem needed to environmental amenity and assume U = U(C), guarantee survivability. but there are plenty of models that do include amenity as U = U(C,S) or U(C,P) (for 3.5 Sustainable use of renewable resources examples see d'Arge and Kogiku 1973, Vousden 1973, Forster 1973, Maler 1974, Lusky 1975, In natural resource-based economies, Becker 1982, Krautkraemer 1985 and 1986). sustainability can often be reduced to somewhat simpler and more operational concepts of 3.4 Survivability versus sustainability sustainable resource use, such as those listed in Table 2. If a country is very poor, with virtually We tn moeeofno capital or non-renewable resources,, it is We then come to the question of survvablimty totally reliant on its renewable resource base. versus sustainability. There are some optimal Simple survival will then matter far more than growth modellers (e.g Kemp et al 1984) who concerns about environmental amenity, and regard sustainability as meaning simply that survival will depend upon the sustainable use of consumption iS kept above some subsistence the resource base. But does this then mean that minimum. However, most other definitions every resource must be conserved, as in understand sustainability to mean sustaning an definition I in Table 2? Does it mean that the improvement (or at least maintenance) in the conversion of large forests in North America quality of human life, rather than just sustaining and Europe to farmland in earlier centuries was the existence of life-see Allen 1980, quote 4; unsustainable? Such absolutism seems Brown et al 1987; Clark 1986; Markandya and unnecessary from an anthropocentric point of Pearce, quote 2; Repetto 1985, quotes 1 and 3; view, as Repetto (1985, quote 2) and WCED Tietenberg 1984; and WCED 1987, quote 1. (1987, quote 5) recognize. Trade-offs between Simply requiring that the future exlsts may be different resources can in principle be calculated manifestly unfair, since it would allow an using appropriate weights, ending up with Table opulent present followed by a Spartan but 2's definition II of sustainable resource use; survivable future. This is why we choose the although the theoretical and practical problems distinction between survivability and of resource accounting methodologies to find sustainability set out in Table 2 and illustrated in these weights or 'resource values" are legion Figure 3. Our standard definition of sustainable (Repetto and Magrath, 1988; Ahmad et al, development will be definition I-non-declining 1989). A slightly different approach appears in per capita utility-because of its self-evident the work of Barbier, Markandya and Pearce. appeal as a criterion for intergenerational equity. 12 Part I: Concepts Figure 3 Optimality, sustainability, and survivability Formal distlnctions between OPTIMAL SUSTAINABLE and SURVIVABLE development paths of welfare W(t) over time t are: OPTIMAL W(t) maxlmlzesj W(t)e tdt (PRESENT VALUE) fo: SUSTAINABLE W(t) Is such that dw W2 0 tor al time dt -SURVIVABLE W(t) Is such that W 2Wmin for all time EXAMPLES Welfare W (>(a) Path might well be OPTIMAL but Is NOT SUSTAINABLE wndn and also NOT SURVIVABLE time t Path might wel be OPTIMAL, but Is NOT SUSTAINABLE although It Is SURVIVABLE Wmin…-- t W ' c) Path might well be NON-OPTiMAL but it Is SUSTAINABLE and SURVIVABLE Wndn - - - - - - - - - -_- t Part L: Concepts 13 Some wording, for example the argument in accumulation and technical progress. Depending Pearce (quote 2) for 'constraints which set on how we define "value", and provided we resource harvest rates at levels no higher than work in a general equilibrium context, we show managed or natural regeneration rates", might in Appendix 3 how this latter concept can be seem to imply that individual resources should made exactly equivalent to our standard be protected absolutely. However, in Pearce et definition of sustainability as non-declining al (1988) it becomes clear that the concern is to utility. The contrast between the two approaches conserve the total stock of natural resources, is discussed further in Section 3.11. although there are several bases (prices, values, physical measures) for calculating the total 3.7 Deep ecology and non-instrumental stock, and natural threshold and irreversibility sustainability effects severely limit the tradeoffs that can be allowed between different resources without Definition I of sustainable resource use also threatening sustainability. Technical limits to follows from a deep ecology ethic, which holds substitutability are also discussed in Sections that other species have an inherent right to a 3.11 and 8.2. Allen (1980, quotes 2 and 3) sustained existence, independent of their points out that different resources will need instrumental value to man. This point of view is protecting in different countries. This seems a reflected in O'Riordan (quote 2)-who calls this fruitful area for further analysis. notion simply 'sustainability", which causes Further questions arise if a poor country confusion-Turner (quote 5) and WCED (quote has non-renewable resources, or is suffering 6); it is also discussed in Pearce (1987). It is a from pollution. Turner (quote 4) seems to point of view that many scientific studies of dismiss the concept of sustainability for non- sustainable ecosystems adopt, although usually renewables, but this ignores or at least assumes shrouded in the notion of preserving the sever limits on the roles of technical progress or 'scientific value" of species and ecosystems. capital accumulation, as is shown in Sections 6 The now widely-recognized concept of and 8. In any case, an extended definition of "existence value", wherein people "place a sustainable resource use would be necessary: value on the mere existence of biological aid/or definitions III and IV are some of the many geomorphological variety and its widespread possible variations. distribution" (Krutilla 1967), is partly relevant here, since it involves no direct use of 3.6 Maintaining the effective resource base resources. Nevertheless its anthropogenic origin would probably make us classify existence value Yet another concept of sustainable resource use, as an extension of 'enviromnental amenity". is that of maintaining the economy's effective resource base (Howe 1979, quote 1). This was 3.8 Intergenerational equality originally framed for non-renewable rather than renewable resources, and although Howe Another fairly absolute concept is that of strict attributes the concept to Page (1977), there is an intergenerational equality, the criterion that important difference between the two writers logically follows from the Rawlsian maximin here. Page proposed the criterion of a constant criterion of justice. Solow (1974b), Hartwick real price index for virgin materials, which (1977), and the many papers based on these two makes no allowance for capital accumulation or all use a constant (not rising) consumptiori path technical progress which may in principle reduce as their criterion for intergenerational justice. the amount of materials needed to produce a They show how, given certain assumptions, given amount of economic output. In contrast, constant consumption can be maintained by Howe focuses on maintaining the economic following Hartwick's Rule: all rents from productivity of the whole resource base, rather depleting non-renewable resources are invested than the physical stocks of individual or in reproducible (man-made) capital which aggregate resources. This is achieved by substitutes for resource inputs in the production balancing resource depletion with capital function. The assumption that colastant 14 Pan 1: Concepts consumption ensures intergene-rational equity horizon of T years to come, rather than utility at ignores any environmental amenity effects that just one point in time. This is reflected in the may result from natural resource capitaP being lifetime sustainability criterion in Table 2, replaced by man-made capital. taken from Riley (1980). We do not need to use Intergenerational equality features hardly at this more sophisticated criterion for our simple all in discussions on sustainability, so it does not models below, but it is important to bear it in appear in Table 2. It has limited appeal because mind. the present generation lives on for a finite time into the future, and so may itself prefer a future 3.11 Sustainability as non-declining utility or with growing welfare to one with constant non-declining capital welfare. Nevertheless Hartwick's Rule may be useful for achieving sustainability in some In conclusion to Section 3, unless otherwise circumstances, and the Hartwick literature stated, we hereafter use definition I of provides a useful interpretation of sustainable growth (non-declining consumption intergenerational equality in terms of maintaining C) and definition II of sustainable development the capital stock intact (recently noted by Solow (non-declining utility U). How can these 1986, and discussed further in Section 3.11 definitions be related to the several conditions below). for sustainability quoted in Appendix 1 that specify some capital stock should be preserved? 3.9 Income distribution Two such sustainability conditions, although very different ones, are: We now turn to the question of intragenerational "...a society that invests in reproducible equity, better known as income distribution. It capital the competitive rents on its current is clear from Allen (quote 2), Goodland and extraction of exhaustible resources, will Ledec (quote 1), Porritt (1984, passim), Tolba enjoy a consumption stream constant in (quote 1), WCED (quotes 1-3) and the World time [and hence achieve intergenerational Bank (quotes 1-3), that many regard alleviating equity]. .. .this result can be interpreted as poverty, both within a nation and between saying that an appropriately defined stock nations, as an integral part of sustainable of capital-including the initial endowment development. No system of weights, which can of resources-is being maintained intact, theoretically compute an increase in social and that consumption can be interpreted as welfare U whilst the poor get poorer, is the interest on that patrimony." (Solow acceptable to this point of view; and definition II 1986, p141). in Table 2 reflects this. WCED (quote 4) goes "We summarize the necessary conditions further, and seeks curbs on any high living [for sustainable development] as standards which would be physically impossible 'constancy of the natural capital stock'. for everyone to have. This point of view, More strictly, the requirement is for non- including the notion of development as a state as negative changes in the stock of natural well as a process, is reflected in definition m of resources such as soil and soil quality, sustainable development. ground and surface water and their quality, land biomass, water biomass, and the 3.10 Definition of a future generation waste assimilation capacity of receiving environments." (Pearce, Barbier and Finally, it is important to reflect on what is Markandya 1988, p6) meant by a ufuture generation". Most of the Our discussion here centers on the current generation of decision makers expect to substitutability of various inputs for each other stay alive for several decades more, and may be in both the production function Q(K,R.,RJ,T,E1) willing to make sacrifices for a while in return and the utility function U(C,E2) . (We have for a better future. Therefore a more appropriate again conflated natural resources S and pollution measure of a generation's welfare would be P into two environmental aggregates E, and E2, discounted utility over an appropriate time which differ because different parts of the Part 1: Concepts 15 natural environment are important to amenity Maintaining a non-declining Q would tien and to productivity; we also ignore labor input obviously require that the natural capital stock L.) The perspective is neoclassical in the sense El (from which the flow R. is also derived) is that doubts in Section 2.4 about the moral preserved. However, a closer reading shows that validity of substitution are ignored, but crucial they do acknowledge that technology, non- doubts are raised about the normal neoclassical renewable resources and man-made capital assumptions about the technical feasibility of clearly have some role in production; but their substitution. We quote the above papers as implicit production function is not neoclassically recent and fairly representative examples of smooth. They stress that: broad 'neoclassical" and "ecological" schools "...natural capital differs from man-made of thought on sustainability. capital in a crucial respect. Man-made Solow's condition of a constant capital capital is virtually always capable of stock (including non-renewable resources) can symmetric variation-it can be increased or readily be derived from a constant utility decreased at will. Natural capital is subject criterion. His implicit utility function ignores to irreversibilities in that it can be environmental amenity: U = U(C) only, so decreased but not often increased if constant utility U requires constant consumption previous decrements led to extinction. C. His production function ignores renewable ... natural and man-made capital are resource flows R and the environmental stock substitutes only to a limited extent." E1: Q = Q(K,R.,T) = KaR.bemt, a Cobb- (Pearce, Barbier and Markandya 19,88, Douglas function. As non-renewable resources p15, emphasis in original) K. are depleted, man-made capital K must be Hence, if natural capital is already depleted built up to substitute for resources in the to the point where irreversible damage may be production process. The mathematically caused, a necessary (but perhaps not sufficient) 'smooth" nature of the Cobb-Douglas functions condition for sustainability is that natural capital assumed by Solow and other neoclassical writers is conserved; man-made capital, technology or such as Stiglitz (1974) ensures that such non-renewable resources are no substitutes in substitution is always technically feasible. Solow such a situation. then shows that maintaining a constant C The differences in the neoclassical and requires some aggregate stock of K and R. to be ecological approaches to sustainability-Solow preserved by choosing a certain level of assumes away any technical limits to investment I (= Q-C). substitutability and ignores the biosphere, The analysis of Pearce et al is not whereas Pearce et al emphasize "threshold presented as a mathematical model, so the effects" and 'critical minimum stocks" [of following formal interpretations are necessarily natural capital], and give the biosphere pride of more open to debate. Their definition of place-are to some extent explained by the sustainability is also to sustain output Q, and difference in context. Solow is explicitly they also make no mention of the environment concerned with developed nations using non- as a source of amenity. However, their renewable resources (he uses the depletion of necessary condition for sustainability is that the North Sea oil as his example), whereas Pearce et productive natural capital stock El be preserved, al focus on microeconomic decisions in so this should also ensure that utility U = developing countries. Renewable resources are U(C,E) is preserved, given the similarity of relatively unimportant in developed counrtries environmental measures El and E2. and relatively more important in develojping The real interest is in how Pearce et al countries, and this may well affect the degree to reach their sustainability condition. In places it which renewables can be substituted for by raan- might appear that they are assuming that made capital. But we doubt whether it is renewable inputs and the state of the sufficient to reconcile the stark differences in the environment are the only inputs that really two approaches, and feel the whole question of matter, so that Q = Q(Ra, El) and the economy- substitutability, which crops up again in Section environment model looks like Figure 4. 8.2, warrants a good deal of further research. 16 Part I: Concepts Our conclusion here is that most single- U. What is the optimal depletion rate of non- valued sustainability criteria, including our renewable resources? What are the optimal chosen definition of non-declining utility, can be stocks and flows of renewable resources? Does reduced to 'maintaining the capital stock optimal pollution increase or decrease over time? intact". However, this does not make choosing Above all, does optimal utility increase or between them any easier. Deciding what is the decrease, that is, is optimal development relevant capital stock, and how it should be sustainable or unsustainable? (Recall that measured, inevitably boils down to deciding how 'optimal" here means nothing more than the essential to and substitutable in production are maximization of discounted utility, as shown in the different components of capital: machines, Table 2.) technical know-how, renewable and non- Unfortunately a general mathematical renewable resources. solution of such a complex system, using The answers directly affect the operational dynamic optimization techniques such as optimal relevance of a sustainability criterion. In a world control theory, is quite impossible. We are then of perfect certainty, a sustainability criterion forced to use highly simplified models, with all which focuses on preserving just the natural the defects that were pointed out in Section 1.3. capital stock will not make sense if man-made The alternative is simulation modelling, i.e. capital can always substitute for natural capital computing a solution using real data and and is steadily being accumulated. Nor will it estimated functional forms. This approach has make sense to focus on a purely physical value-for example, it can shed light on measure of the "effective resource base" (man- Ucatastrophes" (Ayres and Sandilya 1987)-but made plus natural capital) if technological any results will lack generality and may give progress is steadily increasing the economic little theoretical insight. value that can be produced from one physical Good summaries of what optimal control unit of capital. The uncertainty of future theory can achieve in the field of economic technical progress may mean that a purely growth with natural resources and pollution are physical measure is prudent, but one must be in Clark (1976), Smith (1977) and Kamien and aware that the answers given will be over- Schwartz (1982). There are intrinsic cautious. Finally, if sustainability is to mean mathematical problems in analyzing systems anything for trading and manufacturing nations, with more than two endogenous 'state it will not make sense to focus solely on a variables", and even assuming away all the nation's own resource stocks; what will matter problems of aggregation, the model in Section 2 is maintaining balanced trade and the has five state variables: capital K, renewable productivity of its physical and human capital, resources S., non-renewable resources S,, possibly in the face of rising real prices for pollution P and population N. Kamien and resource inputs it needs to buy on world Schwartz review models that include just K, S. markets. and P as endogenous variables, and thus ignore both population and renewable resources, or at 4. Optimal control and least treat them as exogenously determined. sustainability They find that only one model (Maler 1974, Ch 3) attempts to cover all three variables, and then 4.1 The role of optimal control models says little useful about the solution. The models that do cover renewable resources are typically Returning now to the general model of Section partial equilibrium fishery or forestry models 2, one would like to be able to analyze the that take the price of the resource as given rather various features of an optimal development path, than endogenous. These cannot tell us much depending upon the initial stocks in the about the sustainability of poor economies that economy, the various functional relationships in depend on renewable resources, although they Section 2.3 for output Q, resource growth G, may be useful for looking at trade in cash crops. pollution assimilation A, population growth One important result for sustainability does dN/dt, labor productivity L and social welfare spring out of the models that just look at capital Part I: Concepts 17 Figure 4 A developing country model totally dependent on renewable resources Environmental Qualty E Natural Resource Stock 8 Pofluton Stock Renewable P Resourcesl 8^ Waste Flows D ,,~ U(C ) L { PRODUCTION Output Q(R4.,E) _ -Consumption| i-I- HOUSEHOLDS. _ _c 18 Part I: Concepts K and non-renewable resources Sn, namely that rates caused by income tax, imperfect the optimal solution often results in declining information, etcetera. Why then should the utility in the distant future, i.e. is not current generation seek to impose a sustainability sustainable. This occurs in Dasgupta and Heal constraint on its own decisions? If it is (1979, p299), Dixit (1976, pl60), Kamien and concerned that high discount rates will lead to Schwartz (1982, p61), and Lusky (1975, p325), profligate resource consumption now and to name but a few. The intuition is fairly hardship for future generations, why does it not obvious. If non-renewable resources are essential lower its discount rate? to output, consumption and utility, and if This is not an easy question. One answer discounting reduces the perceived value of future could simply be that the "sustainability lobby" utility, then in the absence of continuous has a lower discount rate than the rest of technical progress which allows output per unit society, and is seeking to impose its own world of resource input to rise without limit, declining view (in which the optimal future is also utility is eventually inevitable as the resource sustainable) on the rest of society (which doesn't runs out. If the resource input is also an care about sustainability). enviromnental amenity (or if the resulting A more appealing answer is that people do pollution is a disamenity), the decline of utility not have a single set of preferences that apply to will be even worse. This is not always obvious all decisions. Preferences for social goals may in the literature, as authors (e.g. Lusky 1975) be separate from preferences for private may not bother to compute the optimal utility behavior (Solow 1974a, p9, plO and Page:, 1983 path. Sections 6 and 7 give simple models of and 1988). It is possible to feel differently about non-renewable resource depletion which a course of action according to whether one is illustrate these points. One result there is that listening to one's individually selfish desires or high discount rates can cause the optimal utility one's sense of social responsibility about the path to be unsustainable: this is relevant in future; to behave one way in the market place discussing the relationship between optimality and yet to vote for a government which has and sustainability, to which we now turn. policy goals separate from just perfecting the market place by supplying of public goods and reducing public bads. This comes near what 4.2 Optimality and sustainability Marglin (1963a, p98) calls the 'schizophrenic" answer: 'The Economic Man and the Citizen We noted at the start of Section 3 that are for all intents and purposes two different sustainability criteria are constraining, not individuals." We would not go as far as, this, maximizing, criteria. Several different futures since Economic Man can still maximize self- may be sustainable, and a sustainability criterion interest (seek optimality) within the bounds will not say which sustainable future is the best (sustainability) that the Citizen lays down. to pick. The obvious answer would be to pick A strong word of caution is necessary to the optimal sustainable future: that is, the balance the above remarks on private selfishness sustainable development path which gives the and public responsibility. There is a strong greatest present discounted value. This notion, tendency in neoclassical welfare economics, that sustainability constrains optimality rather based in part upon analyses of savings externa- than completely replacing it, is clearly spelt out lities such as in Sen (1967), that collective in Goodland and Ledec (1987, quote 1), Pearce concern for the future, as expressed by govern- (quote 2), and Tietenberg (quote 2). ment policies, is greater than private concern, as Yet it raises important philosophical expressed by free market decisions. This is not questions about collective decision-making. A necessarily so, as the public choice literature fully optimal solution will fully reflect the quite clearly shows. Given that governments do interests of the current generation of decision not simply maximize social welfare, they may makers: it will have corrected for all market use higher discount rates than private failures such as pollution externalities (as we individuals, so that less rather than more shall see in Section 7), distortions in interest govermnent intervention will be what is needed Part I: Concepts 19 to make economic development more borne in mind throughout this paper, even if it sustainable. This countervailing theme is is not always explicitly stated. discussed in Section 7.2 below, and it should be Having completed our survey of concepts be of growth, development and sustainablity, we now apply them in Part II. Notes 1. The following classification of natural resources is implicit throughout this paper, with examples given in brackets: a. Non-renewable materials (metals) b. Non-renewable energy (fossil fuels) c. Renewable materials (plants) d. Renewable energy (solar) Some resources can fall into more than one category: fossil fuels are also used as chemical feedstocks, plants can be used for fuel or food energy. We avoid the term exhaustible resources, because of possible semantic confusion. Some writers use it to cover categories a, b and c (category d, solar energy, is clearly inexhaustible), but others restrict it to categories a and b, arguing that since renewables can provide a sustained yield through natural growth they are not exhaustible. 2. It is normal to assume that social welfare U depends on per capita consumption c C/N rather than total consumption C. We make this distinction only in Section 9 where population N is assumed to vary; elsewhere population is assumed constant, and it is unnecessary to distinguish between C and c. 20 Part II: Applications 5. Economic growth and the by current flows of output and clean-up. The environment-balancing appropriate 'stripped-down" version of Figure consumption and clean-up 1 for this economy is Figure 5. Therefore the expenditure change in environmental quality as output eechanges is This section is a slight digression from the main dE = EdQ + ExdX < (EQ + Ex)dQ question of sustainability. It looks the related issue of when and why economic growth and The inequality is because we would not environmental improvement may be mutually expect the increase in clean-up expenditure, dX, consistent objectives over time, even if they are to be greater than the increase in total output, antagonistic at any point in time. Our simple dQ. So: comparative static analysis seems hardly new, but in fact there is remarkably little literature on -EQ) > Ex - dE/dQ <0 the way in which changes in economic growth affect the optimal of environmental quality. Most This says that if increased output dQ papers either ignore the fact that society can generates more pollution than can be cleaned up choose a level of environmental quality by by spending all of the increased output on clean- varying its spending on pollution control up, then economic growth will inevitably cause ("clean-up"), or look only at long run steady environmental degradation. It is important to states. The closest analysis to what follows state this to counteract the simplistic view still seems to be the optimal control model in Forster often expressed that 'we must grow in order to (1973), but he does not address the issue of clean up the effects of growth"; clearly this is whether optimal environmental quality improves not always true even physically, let alone or declines as optimal growth proceeds. economically when there is a choice between First of all, assume that investment I is spending on consumption and spending on clean- some fixed proportion of output Q, so that we up. Everything depends upon the type of growth ignore the problem of how to determine the and how much extra pollution it causes. optimal level of saving and capital accumulation. A more interesting question is how the The analysis is comparative static, with output Q optimal, rather than technically possible, level of regarded as exogenously determined. We may as environmental quality changes as output grows. well then treat output as simply divided between The choice between consumption C and clean-up consumption C, and clean-up expenditure X: expenditure X is determined by choosing X (and hence C = Q - X) to maximize social welfare Q=C+X + + U[C,E] = U[Q-X, E(Q,X)] The physical waste flows inevitably associ- ated with output Q (whether used for con- The first order condition for maximization is sumption or clean-up) cause pollution and lower overall environmental quality E, but clean-up -aU/aC + (8U/aE)(aE/8X) = 0 expenditure itself lowers pollution and raises environmental quality: From this equation (checking that it does maximize rather than minimize welfare), one can E = E(Q,X); EQ < 0, Ex > 0 in principle calculate the optimal clean-up expenditure and optimal environmgntal quality as Note that there is no stock pollution effect e here: environmental quality is purely determined Pau II: Applications 21 Figure 5 Economic growth and the environment: a static model with clean-up expenditure CtEnvironmenal Quaty EP Stocka Technology t 0 |~~~~~~~~~~~~~~~~P-E(Q X' K~ T Technology Services AL X T Waste Flows I Cap ~ a Dr ,aCX Services ~C+X K AL l PRODUCTIO | AiL Labor vs J ~~~~~~~~Services I( w_ . ~~~~~~~~~~~~L iWaste , OutS t a X Consumption L HOUSEHOLDS I Waste Flowsa C I ~ = -U (C,E), Environmental Amenity Clean-up Expenditure X 22 Part MI: Applications functions of the given level of output: hypothesis. This states that environmental policy is inevitably weak in the early stages of X* = X*(Q), E* = E*(Q) industrialization, because the environment has not been polluted before, so no 'modernm and the interesting question is then whether property rights over it exist.' Property rights dE*/dQ > 0 or < 0: that is, does environmen- take many years to establish through the political tal quality optimally improve or decline as the process, and in the meantime growth follows a economy grows? path of intensive resource use, causing excessive It is impossible to answer this in theory: pollution and leading the economy to point A. everything depends upon the functional forms Only when environmental property rights and an U(.) and E(.), and neither of these are easy to active environmental policy are established, will measure in practice! However, making heroic sufficient output be diverted to cleaning up assumptions about how aggregate environmental pollution so that environmental quality can quality should be measured, casual empiricism recover to its present level E2 at point B. suggests that most industrialized countries seem The true optimal path, if environmental to have grown along something like the path P1 property rights had been defined and enforced in Figure 6, certainly with regard to local air from the start, might be path P2 passing through and water pollution. Environmental quality starts B (or it might first rise and then fall-see Pearce off at a pre-industrial level EO; declines to a et al 1988, p 18). Along this path environmental minimum El at the height of resource-intensive quality steadily declines, and the rosy future industrialization; then recovers to E2 at point B, promised by path PI is simply unattainable: the representing the present position of a mature environment's assimilative capacity and the industrial country where output has grown to returns to further clean-up expenditure are *Q2. reaching some ultimate limits (see Section 8.2). Is such a typical path optimal? It is not In a dynamic sense it may then not be desirable hard to suggest reasons why it may be. In the to proceed down path P2, and economic growth early phases of growth, people are poor and are ideally comes to a halt and (if we are lucky and willing to trade off decreases in environmental avoid continual environmental deterioration from quality for significant improvements in material cumulative pollution) a steady state is reached. consumption. But then as they grow richer and Under this hypothesis, environmental the environment gets worse, their relative improvement and economic growth are only valuation of consumption and environmental consistent in the real world during the goods alters and they spend an ever greater Ucatching-up" phase from A to B when proportion of output on cleanup X rather than on environmental externalities are being consumption C, leading to the turn-around in internalized. environmental quality beyond point A as shown. Under this hypothesis, economic growth 6. Non-renewable resources I: and environmental improvement are indeed SustainabDlity and the discount compatible in the later stages of growth, and rate much of the new conventional wisdom on growth and the environment holds that they will This and the following three sections make use continue to be compatible in the future, as of optimal control models of resource depletion shown by the upward slope of path P1 beyond over time. Only the key formulae and results output Q2. The message for developing coun- are presented here, and discussed in an intuitive tries would then be that poor enviro.nental way; for further mathematical detail the reader quality is just a necessary phase to pass through is referred to Appendices 2-5, which correspond on the optimal road to mature development. to Sections 6-9 respectively. Sections 6-8 deal However, this does not mean that devel- oping~~~~~~~~~ conre a goetene o ni with non-renewable resources and thus are oping countries can ignore the need for envi- mainly of interest for industrial countries; ronmental policies. The fact that they often do Section 9 deals with renewable resources and is gives rise to an alternative, much gloomier on f a poor rarianrcoury. ' ~~~~~~of interest for a poor agrarian country. Part II: Aplications 23 Figure 6 Possible trade-ofts between output and environment quality (as output Q grows over time) E Environmental Quaity EO Path P1 El - - _ - X ~ 9 Path P2 I . I . ~~~I I Qi Q2 Q Economic Output 24 Part 11: Applications The model in Appendix 2 is one of pure number just like the series 1 + 0.1 + 0.01 + Ucake-eating". The economy simply processes 0.001 + . . .The condition for sustainabiity is some non-renewable resource stock s(t) into a that the rate of technical progress 7 exceeds the consumption flow c(t) as follows (s, c, u are all rate of utility discounting 5. So if a is high per capita quantities respectively equal to S/N, enough-that is, if the current generation's C/N, and U/N): valuation of the future is low enough-then a future that is steadily impoverished is optimal. c = -s elt; initial resource stock s(O) = so The higher a is, the higher the initial consumption c*(O) and the faster it declines. Any other inputs (capital, labor, renewable In such a circumstance the government resources) that are required for this process are intervention can create incentives for resource assumed not to be scarce, and are ignored; conservation that will achieve sustainability, and Figure 7 gives the reduced form of Figure 1 to such incentives are analyzed in Section 7 below. which this model corresponds. As time Conversely, it is possible that private actions proceeds, the cake-processing becomes steadily would result in sustainability, but the more efficient because of exogenous technical government may already be subsidizing resource progress at a constant exponential rate 'A. depletion. If so, removing the government Consumption yields utility, but the marginal intervention will restore sustainability. utility of consumption diminishes: The model can easily be interpreted in terms of the rate of return on investment. u(c) = c"> 0 < v < 1. Investment is simply abstention from resource depletion; the return to investment is the Clearly assumptions that allow the ratio of increased value of the resource over time thanks consumption output to cake input to increase to technical progress at rate r. A high T, without bound, and without any resources meaning a high interest rate, is beneficial to devoted to advancing technical progress (thus resource conservation. Thus high interest rates abandoning the assumption in Section 2 that do not necessarily harm conservation, a theme to technology is a produced input like physical which we return in Section 8. capital), are highly questionable, and are discussed further in Section 8.2. For the moment 7. Non-renewable resources II: we are interested in the results of the above sustainability and environmental assumptions. dependence combined Appendix 2 shows that the optimal solution of the simple cake-eating model is a steadily 7.1 The model-cake-eating with declining rate of resource depletion, which may The mel-cake-eativity however (depending on the rate of steady environmental amenity or productivity technical progress in resource processing) be converted into increasing consumption, and Appendix 3 sets out the details of an extended hence increasing utility, over time: cake-eating model, which differs from the model in Appendix 2 in two ways. Firstly, the c*(t) = Osoe('73t, u*(t) ci el,*'t; economy explicitly comprises N non-cooperating where 4 = (5-Tv)/(1-v) but economically identical people, with the total and hence (T-4) = (T4-)/(1-v) resource stock S = Ns where s is the per capita resource stock. Secondly, the total resource In this model Turner's remark (1988, stock is also assumed to be the 'environment", quote 4) about sustainable use of a non- and has either an environmental amenity effect renewable resource does not apply: resource onutility: depletion is always positive, but sums to a finite U = u(c,S) = ceSr;0o0 even if the government ignores the measures environmental productivity nonoptimality. Note that sustainability implies no criticism of a positive utility discount rate (5 > The idea of environmental amenity goes 0) per se. back at least to Vousden (1973); the idea of associating environmental quality with the level A2If 1(1+ euv) < 6 S r, the non-cooperative of unextracted resource is found in Kamien and path is both nonoptimal and has u < 0, i.e. is Schwartz (1982); and the two are combined in unsustainable, while the socially optimal path is Krautkraemer (1985). What is new here is the sustainable. Optimality again requires effect of non-cooperation: in both the amenity government policy intervention to conserve and productivity cases, people ignore the resources. Such optimal intervention will at the environmental value of the resource when same time make the economy sustainable. planning their privately optimal path. With these particular, multiplicative functional forms, there BIfja..z, both the non-cooperative and socially is no difference in the privately optimal paths of optimal paths have -a < 0, i.e. are resource depletion or utility between the amenity unsustainable. In this case resource conservation and productivity cases. The results are, with the policies which achieve the social optimum are cooperative (socially optimal) results give for not enough to achieve sustainability. comparison: Sustainability requires a stronger intervention policy, the strength of which can be justified Non-cooperation (private optimum) only by a moral commitment to intergenerational equity. s*(t) -se* Appendix 3 shows how the government can where # = (6iTv)/(l-u-e) > alter the rate of resource depletion in the economy by offering conservation incentives: u*/u* = V(T4-{ 1+ C/uV)(l-u-e) either proportional resource conservation Cooperation (social optimum) subsidies a, or declining depletion taxes. According to the strength of a, any desired s*(t) = se4 resource depletion rate (and hence utility growth where 6 = v(5-7u)/[(l-v)(v+e)] < * rate in Figure 10) can be achieved. Also all in Section 6 these results are shown to have a direct u*/u* = v(i-t5)I(1-v) interpretation in terms of 'maintaining the effective resource base" of the economy, an Thus resources are depleted faster and alternative approach to sustainability introduced sustainability (u*/u* 0) is harder to achieve in Section 3.6. than in purely materialistic model of Section 6, because of the environmental effect u. 7.2 Relevance to policy-can environmental There are three policy cases here, policy help sustainability? illustrated in Filgure 10, with Case A of Appendix 2 now split into two sub-cases. (We The above model gives a simple example of how expect this threefold classification to apply to environmental protection and economic welfare many 'tragedies of the commons" cases where can be compatible in the long term. In an externalities and non-cooperative behavior lead economy that is totally reliant on natural to a nonoptimal profile of resource depletion.) resources for economic output, and where the resource itself has envirownental value, 41 40~~~ - - - - -- - - - - - l I - -- - - -- - -- - - - 28 Part 11: Applications Figure 9 A cake-eating model with environmental productivity r Environmental Quality E-S Technology l Natural Stock Resource aTt l Stock S(t) .Non- I Renewable T I 'Cake' Enviromnental Sl Technology Productivity Servlaes l Resource Depletion _ _F_ _ _ _ _ R(t) 1'S' ,R1t---- UtIlIty U.cv Pan II: Applications 29 Figure 10 Sustainability, optimality and government intervention Optimal it(t): maximizes Ju(t)(l)dt Sustainable u(t): has u2 0 v t>O Private Social a Optilmum Optimum UNSUSTAINABLE . SUSTAINABLE CASE Al I . - d/u \ No need fr Optimal Intervenilon sepaate austahbity poicy hi Private Social /Lcsss At A2 Optimum 0 Optimum CASE A UNSUSTAINABLE I SUSTAINABLE L CASE A2 - - I / Optknal lntervention Private social Optimum Optimum - UNSUSTAINABLE . SUSTAINABLE CASE A3 |i d 2/ Optimal Morar Interventlon Needed interventlon to Attaln Sustanabilty Ih a cases. chaooln a Naher daco.mt ratae * *moves the opthasn towwro s umaabhsbizy 30 Part I: Applications Uenvironnental protection" (reducing the rate of However, one must also point out the natural resource depletion) is essential for differences between long-run optimality and "sustained economic growth", i.e. positive short-run output maximization in the model of growth of consumption and utility into the Appendix 3. In all cases, a higher steady growth indefinite future. Resource conservation rate of consumption and utility is achieved only incentives can make the economy more by lowering the initial levels of conswnption and sustainable (i.e. move us to the right on utility, as illustrated in Figure 11. In the real Figure 10). world this means negative economic growth in Conversely, resource depletion incentives the short term, which will impose heavy can make the economy more unsustainable. As transitional costs on an economy (costs which noted at the end of Section 4.2, many are not included in our model) and thus tough governments implicitly use high discount rates political choices, even though the outcome may and promote policies which amount to incentives be optimal in the long term. Thus our model for resource depletion. Page (1977) highlighted provides a crude explanation of the observation the role of depletion allowances for non- by the World Bank (1987, quote 4) that renewable resource extraction in the U.S. The 'Promoting growth, alleviating poverty, World Bank (1987) and Repetto (1988b) and protecting the environment are emphasize how resource extraction in developing mutually supportive objectives in the long countries is often heavily subsidized, although run. ...In the short run, however, the many of these are renewable resources to which objectives are not always compatible..." our model cannot directly apply. The World Resources Institute has done sterling work in 7.3 Property rights and environmental policy cataloguing examples of such subsidies in developing countries (Repetto 1985b, 1986b, Section 7.2 pointed out the importance of 1988a). So the emphasis for both environmental conventional environmental policy in improving and sustainability policy in such circumstances the sustainability of development paths. must be to reduce government intervention, not Environmental policy is all about internalizing increase it. externalities; -and internalizing externalities The suggestions of the above analysis for usually amounts to establishing some kind of sustainability policy-they can only be property rights over the environment. Instead of suggestions, because the model is so very air, water etc. being open access resources, they simple-are as follows. If natural resource use is have to be owned by someone. Here we briefly socially excessive (as judged by conventional review the problems that can arise in optimality criteria which take all environmental determining the distribution of own spillover effects into account), a separate environmental property rights. sustainability criterion may simply be redundant In the ideal case of symmetric congestion in many practical cases. Removing depletion externalities, such as analyzed in Section 7.1, no incentives, and replacing them where necessary distribution problem arises. Everyone both by conservation incentives, will usually improve contributes equally to, and suffers equally from, sustainability as an automatic side-effect (Case the social problem of excessive resource A2 in Figure 10). These conclusions also emerge degradation. Under the proposed solution strongly from the numerous empirical studies by (government conservation incentives), everyone Repetto cited above. Given that it will be much contributes equally to the solution and no equity easier to sell such policy changes by appealing problems arise. to the collective self-interest of the current Environmental problems are rarely generation, rather than to noble concepts of symmetric in the real world. Often one. can intergenerational justice, there is much to be said separate the polluter from the pollutee, and the for concentrating practical efforts on question then arises as to who should own the strengthening conventional environmental environmental property right. Should the factory policies. own the river, and charge local citizens for Part 11: Applications 31 Figure 11 Effect of a lower effective discount rate on initial utility Privatei* UtMlty Path ult) With Low (o-oa) u*(t) With Hgh (6 -a) - Tkle t In the cake-eating models of Section 7 the non-cooperative (privately optimal) path of utillty Is (t) (6 - V - TV) V+ xp{TV. (-O)(v+fe) Therefore a lowering of the effective discount rate (6-a) - whether by consumers choosing a lower utity discount rate 6 or by the goverrunent raising the conservatlon subsidy a - will inprove sustainabilty (raise C*/u*). but wW elso lower the IniUal level of utlity ui(o). 32 Part II: Applications swimming and fishing in it? Or should the since they will certainly limit the pace at which community own the river, and charge the factory redistribution can proceed. for discharging its effluent into it? There are two Finally, one must question whether the main schools of thought on this. One school, property rights approach can be a universal started by Coase (1960), holds that efficient solution to environmental problems. Can one resource allocation may be achieved irrespective really extend the notion of ownership to global of whether the pollutee or the polluter has the resources such as the stratosphere and the right to use the environment. The sole role for oceans? The costs of excluding non-owners from government is in defining and enforcing property using these resources suggests that this may be rights. The conditions for this tCoase theorem' impossible, and alternative mechanisms may be to hold are very restrictive: all users of the needed. Now for thousands of years indigenous environment must have perfect information, peoples have managed many common property bargaining between them must be costless, and (as opposed to open access) resources on a small changing environmental property rights should scale in a sustainable way, using non-legal and cause no significant income effects. non-economic mechanisms such as consensus, Nevertheless, the Coase perspective is cooperation and tradition, as well as private important, particularly in contrast to the other rights (Southgate and Runge 1985, Runge 1986). school which assumes that internalizing The ultimate challenge for the human species externalities means that the 'polluter must pay", may therefore be to rediscover and reapply these for example through emission charges. This common property mechanisms on a global scale. view dates back to Pigou (1932), was boosted by the declaration of the Polluter Pays Principle 8. Non-renewable resources Em the (OECD 1972) and is still widespread (e.g World role of investment, and Bank 1987, p23). technological limits to growth An amalgam of the two views is that while government environmental policies (beyond merely defining property rights) are necessary to 8.1 The model-capital growth with overcome the problems of imperfect information environmental amenity or productivity and transaction costs, it will often be Sections 6 and 7 ignored the role of capital (K) counterproductive for such policies to make the polluter pay. De facto pollution rights often exist in economic growth and resource depletion. within the political system, and policies such as Clearly capital can substitute for resources in emission charges may radically change pollution many ways: using a clock thermostat to reduce rights and thus be politically unacceptable. More energy consumption for space heating is a progress may therefore be made towards simple example. Appendix 4 sets out the bare efficient and sustainable use of environmental details of a simple model that allows capital resources if other policies are pursued (such as investment or accumulation, using a Cobb- a charging/subsidy mix) which internalize Douglas production function for output. As in environmental costs without challenging the Section 7/Appendix 2 model we regard the pollution rights (Pezzey 1988). total resource stock S also having the properties This is not to say that environmental of a public environmental good. We introduce property rights should never be changed. The either multiplicative environmental amenity discussion in Section 10, about the income and (strength e) into the utility function, or hence allocative effects of redistributing multiplicative environmental productivity environmental property rights from rich to poor (strength w) into the production function, as people, shows that such redistribution might be follows (L = labor, R = resource flow, r = an effective way of simultaneously improving technical progress): the lot of the poor and improving the Environmental amenity environment. But the political difficulties of such redistribution should never be underestimated, Utility u = ceS' Output Q = AKLDRle!t Part 1: Concepts 33 Environmental productivity 8.2 Capital-resource substitution, interest rates and technological limits Utility u = cv Output Q = AK¶IRYS elt The reason why higher interest rates are The environmental amenity model is consistent with resource conservation in the illustrated in Figure 12. For such systems we above model is because capital and resources are cannot calculate the exact optimal growth path substitutes in the Cobb-Douglas production for resource depletion and capital accumulation function, so capital investment saves resources. starting from any given initial stocks of If the effective value to the investor of resources resources and capital. The most we can analyze saved is raised by implementing a conservation is the optimal steady state when all stocks and subsidy, then the return on investment, i.e. the flows are growing (or declining) exponentially. rate of interest, will also be raised. Appendix 4 gives the privately optimal resource However, it may not always be the case depletion rate, real interest rate (= return on that cost-minimizing production decisions result capital investment) and utility growth rate for in capital and resources being substitutes. It is each system, assuming again that the economy easy to think of examples where labor-saving consists of non-cooperating agents who ignore capital equipment (e.g. a bulldozer) also requires the environmental cost of private resource resources (i.e. diesel fuel) in order to be depletion. The important things to note about the productive, although it is harder to model such privately optimal solutions are that: capital-resource complementarity * the resource depletion rate (-S/S) rises mathematically. We call the two types of as the environmental parameter (e or investment resource-saving investment and 7w) rises; resource-using investment, and the distinction * the interest rate rises in the amenity proves to be important in our discussion of case as e rises, but falls in the discount rates in Section 11. productivity case as ir rises; The optimistic conclusion of the model in * the growth rate of utility falls as e or Section 8.1 is that, given high enough - rises, and a higher e or ir raises the technological progress (and suitable resource minimum technical progress r needed conservation policies if environmental effects are to ensure sustainability. important), sustainable development is possible Further analysis then shows that a with per capita output, consumption and social proportional conservation subsidy a (subject welfare growing without limits. Many dismiss again to certain restrictions on parameter values) such a future as physically impossible (e.g. Daly can move the economy onto an optimal growth 1987), and it is important to note briefly why path by countering all the depletion and utility this may be so. An unconvincing argument is effects: a higher o will slow resource depletion that ever-growing output must necessarily run and raise the growth of utility (i.e. improve out of material inputs and create ecologically sustainability). The case of resource depletion unsustainable pollution loads. This; is taxes has not been analyzed here. unconvincing because the throughput of material The suggestions (again, no general proof resources required to produce a unit of valued emerges from such a specialized model) for output might decline. Note that outpult is policy intervention to improve sustainability, are measured here in value, not physical units. The thus the same as in the 'cake-eating" model of reduction in the energy/GNP ratio for the U.S. Section 7. The effects of conservation subsidies economy in the 1970s is an example of how high on reducing resource depletion and improving resource prices can induce substitution away sustainability are in line with intuition, but the from resource inputs; further reductions in the result that conservation incentives lead to higher material intensity per dollar of output are clearly interest rates warrants some discussion, which possible with continued capital accumulation and now follows. technical progress. 34 Pan H: Appicatlow Figure 12 A capital accumulation model with environmental amenity Environmental Qualty E-8 1 _ Ca TWobe Natural Il saptockeholg Resource i tlck StS S( Reo urces TeOlogy R sourc o So' ~~~~~~~~~~~~De Isjon L~~~~~~~~~~~~ | rvlcosL< T ( PIRODUCTION ) Labor Service. L jtutw t _I ConaLumpton.I HOUSEHOLDS Capital khWestnnt IL Uu. C Sf -S--- Environmental Amenity Part I: Concepts 35 The important question concerns the 9. Renewable resources: poverty, ultimate limits of capital-resource substitution survival, and outside assistance and technical progress. The laws of thermodynamics suggest that there must 9.1 The model-corn-eating and subsistence ultimately be a minimum requirement for resource inputs per unit of valued output and consumption also per unit of man-made capital. Physical capital depreciates and requires material resource In this section we consider a simple 'cor- inputs to maintain it. Furthermore, natural eating" model which tackles the most elementary resources, particularly biological resources, are questions of sustainability in a poor economy, vitally different from man-made capital in a where population is growing, and where output number of ways (Pearce, Barbier and numaerdya 1988) wanyso mPeaybe, suBstubler a(essentially food supply) is entirely dependent on Markandya 1988) and so may be substitutable a sigl reealXeouc,'o- onl upto om liit ein prt f alivng a single renewable resource, "corn". only up to some limit. Being part of a living Consumption per capita c is close to some ecosystem, biological resources are inherently subsistence minimum c., reflected in a per multifunctional, are subject to irreversible and capita utility function u .(c - c)". The model, possibly catastrophic changes if stressed beyond analyzed in Appendix 5 and illustrated in certain thresholds, and they directly support life. Figure 13, excludes any role for inputs of However it is very hard to say what the limits of capital, labor, non-renewable resources or substitution might be, and whether they must technical progress in the production process, and ultimately reduce welfare levels (i.e. lead to aes that theres no enrownentlani ... . . . . ~~~~~assumes that there IS no environmental amenity unsustainability) in the distant future, or at least effect on utility. The rationale for these bring growth to an end (i.e. level off in a steady assumptions would be that per capita state). consumptLon and resource stock levels are so Also it is important to realize that technical low Iwthat concern for environmental quality per progress is not a free good but itself requires se is negligible, and people have neither the scarce resources to be produced and time, energy or education to bring about any communicated (Ayres and Miller 1980). It gives technical progress in the cor-growing and rise to external adjustment costs, borne by harvesting processes. So the model very crudely society as a whole. The role of education may illustrates some ke olic choices for turn out to be crucial, as longer and longer y in y period of edcto (cnumn materia subsistence farming in the developing world. periods of education (consuming material The model does not involve any common resources) become necessary, both to learn and property enviroanental problems leading to apply existing technologies, and to continually n a i c discover new ones. Technological limits to dnonoptbmalty, although It could be extended to growth form a hugely complex subject area do so by makae g the natural resource stock an which we cannot consider further here. The pr y limits to growth debate that was started by resource. Assuming that decision makers give no Meadows et al (1972) no longer catches the to futurepolion growth when public eye, but it is still actively pursued (Smith discou futiye model re ath per 1979, Lehman 1981, Gibbons 1984, Baumol capita scountumg utlocty, the model results are that per 1986 Norhaus1986 Perings1987and yres capita consumption c (and therefore utility u) 1986, Nordhaus 1986, Perrings 1987 and Ayres cno utialrvddta oh 1988a, are Just a few of the many contributions gro y p in the last decade). (1) the resource growth potential p exceeds the sum of the utility discount rate 6 and the population growth rate X: p > 6 + X. If not, then the 36 Part II: Applications Figure 13 A corn-eating model with subsistence consumption Natural Resource Stock S Renewable Cornb I -Sa Resource Extraction Ra - Natural Growth3 S - Stock Growth b -~~~~~~~~~~~ - Otput =Qu gS-Sl Conasunptlon - HUEHOLDS = : Ppltio 'u uctm v Pail R1: Applications 37 consumption level declines We can see here why the predominant exponentially to c, and the society concerns of survival here lead to a sustainability grinds along at subsistence levels for concept based on physical resource conservation ever. rather than improving social welfare (see (2) the minimum subsistence level of per Sections 3.4 and 3.5 above). In a more realistic capita consumption cm is less than the model with several different resources, resource per capita productivity of the initial accounting techniques would be important for resource stock sO, allowing for measuring different resource stocks and growth population growth: cm < so(p-X). rates, and for calculating meaningful aggregates If decision makers do give weight to future (Repetto and Magrath, 1988, Ahmad et al., population growth when discounting utility, 1989). which would seem a fairer criterion in this model where population growth is exogenous, 9.2 Possible extensions then the sustainability criterion (1) becomes p > 5. This is easier to achieve and the growth of The shortcomings of the simple corn-ea,ting per capita consumption is slowed, because model are many. There is no endogenous people are making conscious provision for extra determination of population growth. There is no mouths to feed in the future. However, the consideration of limitations that the carrying weighted utility criterion is questionable capacity of the environment would impose^ on philosophically, since population growth is resource growth, limitations that would often be rarely exogenous in practice, and using a modelled using a logistic growth function weighted criterion effectively treats future population growth as a good thing (Koopmans S = pS(Sc-S) where Sc is the carrying 1977). Note how a zero discount rate a = 0 is capacity of the environment. quite unrelated to sustainability here, as in the previous models with non-renewable resources Less renewable resources such as soil (see for example Section 7.1). quality are also vitally important. Morey (1985) The condition (2) is an initial condition to has an interesting model that assumes a constant enable "take-off" into sustainable growth. If the absolute (as opposed to the proportional increase harvest from the initial resource stock is not big above) in soil quality, and he studies the enough, people will be forced to eat what should conditions required for total depletion of soil be set aside as seedcorn simply in order to quality (i.e. desertification) to be optimal. survive the present, and this leads to inevitable However his is a partial equilibrium model that disaster. takes the value of output from the soil (i.e fiood) Crude implications for policy are that for as exogenously given; he suggests endogenising an unsustainable economy to be converted into food value by using a utility function, much in a sustainable one, one or all of the following the way we have done above. This is an must happen: interesting line for further work. * Increase the resource growth rate p Many other potentially interesting models ('improve the efficiency of farming"); suggest themselves, but so far apparently have * Decrease the population growth rate X not been investigated. Possible topics to cover ("promote family planning"); are: * Increase the initial resource stock SO * Economies with both a renewable ("seek development aid"); resource (agriculture) and a non- * Decrease the initial population No renewable resource (say copper ("famine and starvation"). mining). Governments of very poor countries may * Economies with a renewable resource not be able to implement the first three policies and capital accumulation. Most opitimal without outside assistance, hence the case for control models with renewable development aid from rich countries; otherwise resources ignore capital (Clark 1976, the fourth grim solution will impose itself. Smith 1977). Clark et al (1979) tive a 38 Part I: Applications partial equilibrium model here which One connection derives from the frequent could perhaps be turned into a general observation (crudely modelled in Section 9, and equilibrium model of interest. reflected in WCED quote 3 and World Bank * Economies with two renewable quote 5 in Appendix 1) that very poor people resources (forestry and agriculture) may be driven to destroy their environment; where excessive use of one resource desertification of grazing areas in the Sahel is imposes external costs on the other perhaps the best known example. So any policy (e.g. floods, siltation). to help these poor people must take * Modelling the effect of open access environmental dependencies into account. versus private ownership of renewable Moreover, environmental degradation frequently resources is also necessary (similar to affects other parts of society, as deserts encroach the modelling of non-cooperation upon cropland, or as deforestation high in a versus social optimality in the models watershed drives genetically valuable species to of Section 7.1 and 8.1). extinction, and causes floods and sedimentation The intergenerational equity literature has problems downstream. But by definition, people looked at some of these areas. For example, who are so poor that they are driven to destroy Okuguchi (1979) formulates a model using their environment will not be able to pay capital, labor and several renewable and non- anything for the external damage they are renewable resources as inputs to production. By causing elsewhere. The Coase theorem, that the assuming that all inputs are substitutable at the allocation of property rights will not affect the margin, he not surprisingly shows that allocation of environmental resources (see maintaining the stock of just renewable resources Section 7.3) does not hold here. If property is not necessary for sustainability. However, rights are effectively given to the poor., their there is no analysis of optimal (as opposed to wealth will be greatly increased and the sustainable) paths for the economy, or of open environrnent will be affected. The rest of society access effects. (presumably richer) will have an interest in paying the poor not to destroy their 10. Income distribution and environment, which will simultaneously improve sustainable development the environment and redistribute income. The second connection, noted by the The applications in Sections 5-9 have all ignored World Bank (1987, p6) is that poor people are the distribution of income or welfare within the often the greatest victims of pollution. The society being considered; all that mattered was property rights perspective is again interesting aggregate consumption and aggregate here. If the wealthy industrialist effectively owns environmental quality, etc. Yet as Section 3.9 the rights to the environment, pollution will pointed out, the etcYical concen for continue at a high level because the poor who intergenerational equity underlying sustainability suffer the pollution can pay very little (say X) to notions is naturally associated with an ethical have the pollution reduced. However, if concern for intragenerational equity. We give environmental property rights are transferred to the poor-clearly an act that redistributes wealth, this some thought here, although this is one of sic suhrgtaevlab-hnte the less satisfactory sections of this draft no sindusa wihat ar theporeno to consistent models have yet been worked out. idustrialist will have to pay the poor enough to An immediate question is: what is the make them willing to accept a given level of policy connection between intergenerational and pollution. This payment may be much higher intagnertina ethan X, and the industrialist will find it .instra ue in ro equity? D poice to ta worthwhile to reduce pollution significantly. So redistributetincomenwithineaesocietyy(ifathatais In both cases, policies that effectively give what society wants) necessarily have any eniomtaroetihstoheorhul connection with sustainability policies? On a p pr y g small scale-say at the level of a small both improve the environment and alleviate developing country-two connections are poverty. Whether they will also contribute to possible. sustainable economic growth is another matter, Part R.: Applications 39 not yet considered here. sustainability. Our approach here reviews At a global level the redistribution question standard analyses (both positive and normative) is inextricably connected to the environment, of discount rate issues, and adds in the second because the question is (Ayres 1988b): how do half of Section 11.2 a novel, purely we permit the LDC's to industrialize without environmental reason why interest rates may be destroying the environrnent? In other words, too high. Much of the discussion springs from how do we tackle the problem of ideas in Markandya and Pearce's recent papers intragenerational equity without making it (1988a, 1988b), which will here be referred to impossible for future generations to enjoy our as MPa and MPb, and from Page (1983), who standard of living? If we believe that it is commented (p57): sHow 'the discount rate', and hence all ecologically impossible for the whole of the interest rates are to be manipulated is human race to enjoy anything like the current usually left unclear. Presumably, standard of living of Western industrialized adjustments are to be done through the tax nations-and this raises empirical questions structure, or perhaps through monetary about the limits of capital-resource substitution policy." raised in Section 8.2 above-then "equitable" We focus here on manipulation of the tax sustainable development will require a reduction structure and ignore monetary policy. in the living standards of rich nations, as WCED (1987, quote 4) implies. 11.2 Changing the demand for investment One interesting, if politically fanciful funds mechanism for such an international transfer of wealth might be a compensation fund for The interest rate is the market price of cumulative global pollution problems, such as investment funds, and its level is therefore the greenhouse effect and ozone depletion. determined, like any other price, by the Payments into the fund would be made in interaction of supply and demand. Let us first proportion to nations' cumulative contributions concentrate on the demand for investment funds, to the problem (so that rich nations pay most) and let us suppose (as is the case in most and payments out from the fund would be made developed countries) that investment income is in proportion to damages cause by the problem taxed. This drives a wedge between the rates of return earned by borrowers and savers, (so that all nations receive a fair share.) reu. eane by bofwrsadsaes (so taalntnreiea.illustrated in Figure 14. The supply of finds is Sl(r), the pretax demand for funds (the gross 11. Are discount rates too high? return to investors) is DT2(r), and the posttax * demand (the net return that investors can pay to 11.1 Discount rates and sustainability savers) is DTl(r), where r is the interest rate (all measured in real terms). Equilibrium is at A, The question addressed here is the perennial one with total investment equal to 11T1, an of whether the discount rates used for cost- opportunity cost of capital rl (we are most benefit analyses of both public and private interested in this because this is the main iinterest investment projects, are in some sense "too rate used to discount costs and benefits in project high" when the projects involve long term appraisal, especially in the private sector) and a environmental costs or benefits. We are talking consumption rate of interest rl'. here about the discount rate for goods and Whether or not rl, rl', or some services, and we will henceforth call it the combination of the two should be used for interest rate, to avoid confusion with the utility various types of public sector project appraisal discount rate 5 used so far in this paper; the has long been a subject of debate, and revolves relationship between the two is shown in Section around how much public investment displaces 11.3. The arithmetic is well known: a 10% real private investment, and how much the returns interest rate reduces a $100 sum 50 years hence are reinvested and how much they are consumed to a present value of less than $1 now, etc. This (Marglin 1963b). MPb suggest using some mean sort of discounting may lead us to choose of the two, but Kolb and Scheraga (1988) projects which do long term environmental damage and harm prospects for future 40 Part 11: Applications Figure 14 Removing investment income tax lowers the interest rate r Interest Rate r _Investment ; t / ~~~~~~~~~~~~~~FuJnds rz t P ~~ ~ ~ ~ ~ ~ i/ DT2(r) r V Port-tax p 1 Demand IT 1 I-~ PT2 I Investment _ _ _~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~Dmn Part I: Concepts 41 suggest an innovative two-stage approach Figure 15 this will shift the resource-using whereby capital costs are first annualized at rl, investment demand curve inward to DU3, the then all annual costs and benefits are discounted resource-saving demand curve outward to IDS3, at rl'. This is not our concern here, and suppose and (because of our assumption that resource- now that investment income tax is abolished. using investment is dominant) on balanceb the (This reduces a market distortion and so would total demand curve shifts inward to DT3. Given generally be regarded as welfare-enhancing, an unchanged supply curve SI, market although it has effects on income distribution equilibrium moves from B to C, investments and government revenues.) The net demand for move to IU3 (decrease), IS3 (increase) and MT3 funds now rises to DT2, and the new (a net decrease), and the market-clearing interest equilibrium is at B, with higher total investment rate drops to r3. MT2 and a lower opportunity cost of capital As shown in Section 7, a tougher (interest rate) r2. environmental policy is likely to make Next, suppose that the new total demand development more sustainable (the shift from for funds DT2 can be divided in the manner resource-using to resource-saving investment suggested in Section 8.2, that is into the demand will itself improve sustainability), and this is DU2 derived from investments that are consistent with the lower market interest: rate resource-using (i.e. cases where capital and that it causes, since lower interest rates give natural resources are complements) and the relatively more weight to the distant future in demand DS2 derived from investments that are present value calculations. Another policy resource-saving (cases where capital substitutes observation is that if environmental policy takes for resources). Total investment is 1T2, divided the form of revenue-generating market into resource-using investment 1U2 and resource- mechanisms such as emission charges or saving investment IS2, as shown in Figure 15. auctioned marketable emission permits, the In drawing these curves we assume that revenue raised could balance out the above loss resource-using investment is dominant. There is of revenue from abolishing investment income little evidence either way on this crucial tax. A similar idea has been explored empirical assumption, but some other authors empirically for the U.S. by Terkla (1984). seem to agree with it. Page (1983, p54) comments: "As an empirical matter, it appears 11.3 Changing the supply of investment funds that with the present accumulation of man-made capital, dependence on the physical resource Let us finally look at the supply of investment base is growing, not shrinking." MPb (p30) funds. The interest rate r that a lender requires simply state that uSince natural resources are to divert his income from consumption to required for investment..". This implies that all investment (the 'consumption rate of interest") investment is resource-using in net terms, which can be divided into two parts as follows: seems to go too far, but it supports our assumption. r = a - ucl/u Suppose also that environmental policy is (MPa, p3, with different notation) currently far from complete, so that the material resource flow connected with every investment where a = the utility discount rate used in the causes many external costs and benefits. For optimal growth models above, and uc is the resource-using investments the external costs are (expected) marginal utility of per capita assumed to significantly outweigh the benefits, consumption. Assuming for simplicity that u = and the converse is assumed for resource-saving (c-cm)u, 0 < v < 1, c. = subsistence investments (more crucial empirical consumption, then assumptions). Now let a tougher environmental policy be introduced, which internalizes many r = 5 + (1-v)c/(c-cm) (*) more externalities. The accounting cost of resources used or saved in both types of When aggregated over several different investment projects will then be driven up. On consumers this gives the supply curve Sl1(r). If 42 Part I: Applications Figure 15 Tightening environmental policy may lower the interest rate r Interest Rate 12 Ia EITbhten Environmental PolIcyt OS , Demand forfuns frResOUwe-saving Demand Expands: Dmd f x fd f rsI Urce-usmng Demand Contracts Greatly, 0? - Tta fo tmtfdTotal Demand Contracts r2.:- - -_ _ _ r3 - - -?k- ~-~ ! N i k >> \D~~~~~~~~T2 { , *\ X . / | ~~~~~DU2 1DT3 IS2 1S3 Wt.3 1IU2 IT3 IT2 I Investment DS - Oenmand for funds lor 'resource-savW investments DU - Dwnnd for funds for 'resource-uhg' investments OT - Tota d mand fW ihvestment funds Part 1: Concepts 43 savers care about future generations in general rate: he may care a lot about his future welfare as well as their own heirs, this process of per se, but any change in his consumption level aggregation leads a supply of investment funds will have such a big effect on his welfare that which may be less than is socially optimal at any his consumption discount rate is very high. given interest rate. This is the well-known Another, related effect on supply may be that the 'isolation paradox" (Marglin 1963a, Sen 1967). probability of dying in the near future depends We can thus see three ways in which the supply on consumption (Zarembka 1972, pp73-81). curve can be shifted to the right, leading to a On the demand side, the observation in curve . . Section 5 that environmental property rights tend still lower equilibrium interest rate:.., .. (1) The utility discount rate o is lowered, to be very weak in 'frontier" developing ..ighuiidionrate is countries is relevant here. Suppose for example Lowering the utility discount Is that traditional communal management of likely to reduce resource depletion tropical forests has been disrupted, but modern rates and increase sustainability, as we private ownership has not yet been established. saw in Sections 6 and 7, but the case Then both the marginal user costs (depleting the for such lowering is a purely moral or forest reserve left for future generations) and the ethical one (Parfit 1983). marginal external costs (soil erosion and climate (2) The expected growth of consumption change) of deforestation will be ignored (Pearce cl(c-cm) in the future is lower. Suppose and Markandya 1987). The situation may be a lender expects the past growth rate made even worse by explicit subsidies to of his consumption to continue into the deforestation (Repetto 1988a). Therefore the future, but there is some reason, logging company perceives a rate of return well unbeknown to the lender, why it will above the social rate of return and hence logs at not-perhaps the thermodynamic a greatly excessive rate. limitations mentioned in Section 8.2. Therefore if natural resource exploitation Then if the lender is informed of this is especially dominant and property rights very reason, his expected 6/(c-cmj will drop weak in a country's economy, policy changes to and his supply curve will shift to the establish proper ownership of resources, and to right. eliminate unwarranted subsidies for their (3) exploitation, will not only alter the perceived ( cSaving is subsidized in some way, to costs and benefits of exploitation at any point in ;orrect for the isolation paradox. time. They may also lower the interest razte in Figure 16 gives a diagrammatic analysis of the whole economy. This ten encourages these changes. Any or all three of them will investors to take a l then vewncurge invstos t tae alonger term view and treat increase the supply of investment funds from S1 forests, etc, as sustainable rather than depletable to S2, moving the equilibrium point from C to resources. D, increasing total investment from IT3 to 1T4, and lowering the interest rate yet again from r3 12. Information and uncertainty to r4. So far we have assumed perfect information in 11.4 Interest rates in developing countries all our sustainability models. In reality information is never perfect, and ignorance A good test of the above analysis will be if it abounds. Maybe many poor farmers, burning can help explain the common observation that down a patch of rainforest in Amazonia to make real interest rates are much higher in developing a smallholding, don't know that the soil becomes countries than in developed countries. We can depleted and worthless after a few years. Maybe suggest some reasons, although these are only governments don't realize how faulty their very tentative. resource policies are. Much unsustainable On the supply side, one reason for high exploitation of natural resources could be discount rates is poverty. Someone close to explained by ignorance, and much of the current subsistence is likely to have a very high research effort on sustainable development is consumption discount rate, as shown in equation directed at overcoming such ignorance. (*) above if c is only just above cm. This is not A rather different form of imperfect the same as saying he has a high utility discount information is the inherent uncertainty of the 44 Part 11: Applications Figure 16 Increased saving lowers the interest rate r interest Rate Increase In Saving Because of * Greater Concern for the Future, and/or * Reduced Expectation of Future Growth, and/or * Government Subsidles to Ellminate 'Isolatlon Paradox' r4 - - - - - - - - - - - - 1T3 1T4 I Investment Part 1: Concepts 45 future. Previous sections contain only one 13. Operationality: putting the id-eas recognition of this, in Section 11.3 where it is into practice pointed out that expectations about future rates of consumption growth may be ill-informed. Yet . .i risk and uncertainty (we make no distinction accepted as socially desirable, how can it be pu here) are pervasive on the timescales to which into practice? In exploring this question, this sustainable development concepts apply, and section pulls together some of the diverse cannot be ignored. Unfortunately it is beyond threads of this paper. The dangers of simply the scope of this paper to present a full analysis talking about 'sustainabiity" are obvious from of risk-further work is clearly needed the numerous different definitions given in here-and we can only refer to some key results Intuitively, it is easy to see that the increased Setion 3, bu t theyall haefnrom on variability of possible future environmental notion of concern thiat thie aggregate welfare of damagei(aboutyof anounchangedt mean)rus ntisal future generations should be protected in some damage (abouts envirnmen policyjevenfor a way. We will try to be more specific where this morisk-utrals policy-maker. Roisk , aver n mer e is necessary, although as in other places in this risk-neutral policy-maker. Risk aversion merely pprteiseo nrgnrtoa qiyi strengthens the conclusion. Yet further caution parte issoeon is justified if there are thresholds, beyond which lge red. environmental damage may be catastrophic and There seem to be three separate questions irreversible, meaning that the worst that can concerning operational ity: > . . ^ ~~~~~(1) To what system should the happen to the environment is much further sustainability criterion apply? below the mean than the best that can happen. (2) Is a separate sustainability criterion These are broadly the conclusions reached by necessary in practice? Siebert (1987, Chapter 14) using a formal model necessaryim cter of optimal (not sustainable) decision-making ()Cnatainabl under environmental risk from cumulative but operational? assimilable pollution. He finds that: These are now discussed in turn, the first assimilable pol.antincrHease uncen th e briefly and the latter two in more depth. '-...an increased uncertainty in the damage function implies a lower level 13.1 To what system should the sustainability of pollution." * if risk aversion is increased, the criterion apply? steady state [requires] a higher penalty This is essentially the question raised in Section on emissions." 3.5: does every resource need to be conserved, '...an increased uncertainty in the or are tradeoffs acceptable? Can resource assimilative capacity of the accounting help us to make aggregate judgments enviromnent implies a lower level of about such tradeoffs? A similar, although more pollution." l* uWit. t radical question is: does every country have to poltionhe benin e a higher experience sustainable development? Or can poluironmeintl quncertain, a ighe some rise and some fall? In any case it is clear environmental quality is optimal in the that we need to define the system to which a steady state. .Higher environmental quealiy ctane .beHinterpreted tas sustainability policy is to apply, before we can quaityuranc beainsteethed isk of answer questions (2) and (3). Policies for insurance against the risk. sustaining a narrowly defined ecosystem, or environmental degradation or as a risk . a premium." ~~~~~~even a single species, will look very different premium." from policies for global sustainability. .A solution of handling Exogenous factors, such as resource prices and irreversibilities is to explicitly environmental effects from outside the system, introduce an option value being will be very different at different levels of defined as the value.. that arises from system. Only at the global level will all such retaining an option to a good or factors be endogenously determined within the service for which the demand is system. uncertain." 46 Pan 11: Applications 13.2 Is a separate sustainability criterion rhetoric if this proves politically necessary in practice? useful, but leaving sustainabilitypolicy per se to academic discussions (such as The case against a separate sustainability policy this?). is that sustainability, while clearly desirable as a The opposite point of view is that social goal, will be achieved in the course of sustainability is a real problem, particularly at pursuing the more operational goals of a proper the global level (see for example Daly 1986, environmental policy. This view was discussed 1987). There is indeed no guarantee that the in Section 7.2. The models of Sections 7 and 8 fully optimal state achieved by thorough suggested how the inescapable physical environmental policies will be sustainable: the connection between resource depletion and world (or whatever system we are concerned environmental externalities, meant thrat with) may be like Case B in Figure 10. The conventional environmental policies to discussion on ultimate physical limits to growth internalize these externalities (policies that are in Section 8.2 is relevant here. Particular stressed throughout the World Bank 1987 paper) importance is attached to rising global levels of are inherently likely to reduce resource depletion cumulative pollution: the greenhouse effect and and promote sustainability. Recall from Section ozone depletion will probably affect most 7.3 that internalizing externalities does not countries, and may cause serious harm both to necessarily mean govermnents intervening with amenity and to productivity (the distinction regulatory controls or economic incentives to between the two was defined in Section 2 and "make polluters pay". In some instances the explored in Sections 7 and 8). We cannot definition and enforcement of property rights resolve this complex empirical debate here, so over the environment will be enough; although we turn to the main question that is relevant who gets these property rights may have a big when sustainability is a problem: how could a impact on both income distribution and sustainability criterion be applied in practice? environmental quality, if polluters and pollutees have very different income levels, as noted in 13.3 Can a sustainability criterion be made Section 10. operational? It is of course very hard to know in practice whether a full environmental policy will This question has to be addressed at two levels, automatically achieve sustainability (C(ase A2 in the system level and the project level; hence the Figure 10) or not (Case B). The practical view importance of first defining what system we are here is that since: concerned with (Section 13.1). In general, if the (a) there are so many problems to system is small and homogeneous, it may be overcome in developing a coherent and possible to measure sustainability (assuming that rigorously enforced environmental influences from outside the system do not policy, particularly in developing change) and to devise sustainability policies; but countries where subsidies which it will be harder to justify both making actually encourage depletion and sustainability of this particular subsystem into an pollution are common; important objective, and ignoring possible (b) environmental policy will probably changes in outside influences. help sustainability automatically; As already noted, the difficulties of (c) it is very hard to measure aggregate measuring sustainability of a large, sustainability anyway; heterogeneous system are obviously great. (d) it is even harder to apply sustainability Liverman et al (1988) find serious weaknesses in criteria to individual projects (see all readily available measures of sustainability, Section 13.3); and these can only be put right with great efforts (e) ethical principles of intergenerational of resource accounting and simulation equity, which have to be invoked to modelling. But let us assume that somehow they justify sustainability, are not necessary are overcome, and it has been determined that to justify environmental policy; policy the system is unsustainable, or nearly efforts should therefore be confined to unsustainable, even with all environmental promoting conventional environmental externalities internalized. What policies follow? policies, perhaps using a sustainability Part II: Applications 47 At the system level, we will find that the for CFC and C02 use. depletion of some resources, or some resource But how could a sustainability criterion aggregate, needs to be controlled. This may be work at a project level, where a project is only an absolute limit to depletion, if there is a a small part of the overall system that is to be minimum stock of a critical resource (say the sustained? How indeed: there is a profound land area of a game reserve, if we want to conceptual problem here, deriving from the sustain a particular species) below which the mathematical definition of sustainability as a system is unsustainable, or just a slowing down constraint rather than a maximization rule like of depletion to allow natural growth, capital optimality. The integral in Table 2 that the accumulation or technical change enough time to optimality criterion seeks to maximize is a sum replenish the "effective resource base". Either of discounted utility costs and benefits for the way, there must be some aggregate constraint, whole system. It is thus possible to work out imposed from outside, on the total rate of whether or not an individual project helps or resource depletion in the system (Daly 1986). hinders optimality, by summing its own This constraint could be regulatory (legal bans discounted monetary costs and benefits. Several or limits on particular resource uses), or big assumptions have to be made in this process economic (conservation incentives to slow down (crucially that the marginal utility of income is depletion, as modelled in Sections 7 and 8), or constant and that the distribution of costs and a combination of both. What will happen in benefits is irrelevant) but at least the Final either case is that the effective price of the procedure (cost-benefit analysis) is conceptLally resource will be driven up (perhaps to infinity) obvious, and more or less operational at the throughout the system, inducing conservation in project level. countless separate project decisions. But how can we say whether or not an Obvious illustrations of this philosophy are individual project contributes to system constraints on global emissions of CFCs or sustainability? If sustainability means that C02. These constraints must be set at a global aggregate welfare shall not decline, is any (or near-global) level, whether through project that decreases welfare at any future time regulation (such as the Montreal Protocol for considered harmful to sustainability, even if it CFCs) or market mechanisms (a mooted global greatly increases welfare at other future times? carbon tax). Prices of CFCs or fossil fuels to This would be an absurd conclusion, and so the final users will rise, and conservation efforts notion arises that project costs and benefits imust will be induced. Together with other system- somehow be smoothed out over time before we level sustainability policies, this may cause can judge its contribution to system sufficient changes in resource prices for the sustainability. This in turn leads to the notion of market rate of interest to be lowered so that all intergenerational compensation projects, a project appraisals automatically give greater concept that clearly attracts a lot of support: see weight to future generations (see Section 11). for example Pearce (1983, p75), Tietenberg The political and scientific problems of making (1984), World Bank (1987, p8), Markandaya a sustainability criterion operational at a system and Pearce (1988a, pplO-11), Pearce, Barbier level are daunting, but at least the concept is and Markandya (1988, Section 11), and d'Arge fairly obvious. (1989, p328). The following quotes illustrate the Another approach is that of approach: intergenerational compensation. This idea is at UIf a particular project being considered the heart of Hartwick's rule (Hartwick 1977): maximizes the present value, but confers the current generation should compensate the some unacceptably low or negative net future for depleting non-renewable resource benefits on future generations, then some stocks by investing in enough capital that the of the current gains could be set aside as a productivity of the resources-capital aggregate is trust fund to compensate for the negative preserved. The idea is extended by Spash and net benefits ...... Whatever its form, the d'Arge (1989) to include compensation in the compensation mechanism provides a way form of increased technical knowhow or specific of sharing maximum net benefits among bequests of goods. Policies to achieve such generations without resorting to a policy compensation might not look very different from that wastes net benefits in a misguided the sort of constraints that were suggested above search for intergenerational fairness." 48 Part 11: Applications (Tietenberg p432) may then lay down rules that a project depleting 'Sustainability can be introduced into cost the resource must be compensated by an benefit analysis by setting a constraint on "environmental improvement" project the depletion and degradation of the stock regenerating that resource (Markandya and of natural capital. Essentially, the Pearce 1988, pp1O-11). But how can this be economic efficiency objective is modified applied to non-renewable resources, or several to mean that all projects yielding net resources which are substitutable for each other? benefits should be undertaken subject to Returning to the example of global the requirement that environmental damage pollutants illustrates some of these points. If (i.e. natural capital depreciation) should be global carbon dioxide emissions are posing a zero or negative. However, applied at the threat to sustainability, it is hard to see how level of each project such a requirement specific compensation mechanisms for individual would be stultifying. Few projects would projects can help. For countless, private, daily be feasible. At the programme level, decisions on fossil fuel burning-how high to set however, the interpretation is more the room heating thermostat, whether or not to interesting. It amounts to saying that, drive to work, how many trees to cut down and netted out across a set of projects burn today-affect carbon dioxide emissions and (programme), the sum of individual long term climate change. What would be damages should be zero or negative." suitable compensating investments anyway? The (Pearce, Barbier and Markandya, Section more appropriate policy seems to be to set a 11, authors' emphasis) system-level constraint on carbon use. Then the This seems a fine concept that can be made resultant higher prices for carboniferous fuels operational. It is already being applied, with for will work their way through normal market example a recent proposal to replant forests in mechanisms and encourage the appropriate Central America as compensation for the carbon intergenerational compensation (more dioxide that will be produced by a new power investments in energy conservation) in these station in New England. However, many millions of daily decisions. questions are still unanswered. Logically, compensation projects seem to be neither 14. Conclusions and suggestions for sufficient nor necessary to achieve system further work sustainability. How are we to judge what is an unacceptably low net benefit for future A few broad conclusions of the paper seem generations, and for which generations? What if worth restating here. Firstly, almost all investments in the trust fund themselves affect approaches to sustainable growth or sustainable sustainability? How is a 'program" of projects development contain the same core ethic of to be defined? If there are many programs, intergenerational equity, that future generations should not the criterion of zero or negative are entitled to at least as good a quality of life as aggregate environmental damage be applied to we have now. Quality of life is a broad concept the collection of all programs? entailing much more than per capita Above all, how can the compensation idea consumption of marketed goods and services. A work in theprivate sector? How can one define neoclassical formalization of the core ethic is a program if there are countless small private that utility (equivalent to quality of life) should investors instead of one big agency? Even if one not decline, although this may allow tradeoffs can, who will carry out the uneconomic between various aspects of life that some "compensating project" designed to balance out consider should be non-tradeable. One important the environmental damage of the other, part of sustainability not covered by the core economic projects? pr fssanblt o oee ytecr It may be concluded from considering ethic is that of intragenerational equity. l system sustainability-particularly in the Secondly, the way in which the core ethic overall system sustainability-c an is translated into a set of conditions for case of poor countries where sustainability can sustainability is highly dependent on the context. be reduced to sustainable resource use (see Sustainability conditions for a small developing Section 3.5)-that the stock of some particular country over the next decade, for the U.S.A. resource must be absolutely protected, or other over the next century, and for the entire planet resource targets and rules of thumb set. One Part 1: Concepts 49 over the next millennium, will all look very * The idea that stricter environmenital different. Deriving sustainability conditions policy can lower the economy-wide inevitably requires judgments on which natural interest rate (Section 11) warrants and anthropogenic resources are essential to further theoretical and empirical production and to welfare, and on the extent to scrutiny. Included in this needs to which these resources are substitutable for each bemore analysis of why real interest other. Many conditions can be seen as rates are so high in developing 'Amaintaining the capital stock intact", but this countries. does not avoid the need for these judgments. * The roles of both common property The existence of natural thresholds, beyond and open access regimes of renewable which environmental damage is irreversible and resource management need to be possibly catastrophic, may represent a significant related to sustainability. This is limit to the substitutability of capital and particularly to explain why some technological knowledge for natural resources. natural resource systems that have long Thirdly, although sustainability has ethical been stable in some developing foundations that lie outside the mainstream of countries suddenly become neoclassical welfare economics, neoclassical unsustainable, and perhaps then to analysis can be illuminating and should not be suggest how open-access global rejected. In particular, it can show how resources like the stratosphere and conventionally justified enviromental policies oceans can be sustainably managed as may make the economy more sustainable as an global commons. automatic side-effect. * The public choice approach to Suggestions for further work include: government decision-making needs * More analysis of general equilibrium much greater attention. It has to be growth models using renewable (as explained why so many governments opposed to non-renewable) resources. encourage unsustainable resource Various models need to allow for practices, and how they are going to capital accumulation, technological be persuaded to change their policies. development and possibly a mix of * Last and perhaps most importantly, the renewable and non-renewable importance of uncertainty about the resources, as suggested in Section 9. future, in making potentially irreversible decisions about the management of natural resources, needs much greater exploration than Note: has been provided in Section 12. Threshold effects and uncertainty 1. A more careful analysis might show that property rights over mighombinecto a centinal the environment did exist before industrialization did exist, but in might combme to give a conventional the traditional, common property form practiced by peasants or economic justification of presenring indigenous tribespeople. The advent of Western concepts of physical stocks of natural capital in capitalist development and 'modernm (that is, private and order to guarantee sustainability, a legalized) property rights may break down such communal systems j t i n p b of restraint. This paradoxically leads to no property rights at all, justification that IS not provided by the and hence severe environmental degradation, in 'frontier perfect information, marginalist economies". See for example Southgate and Runge (1985). analysis of this paper. so References Ahmad, Yusuf J., El Serafy, Salah, and Lutz, Boulding, Kenneth E. (1988). "Can there be Ernst (eds.) (1989). Environmental models for sustainable development?" In AccountingforSustainableDevelopment. Davidson, Alex and Dence, Michael A World Bank Symposium Paper, (eds.), The Brundtland Challenge and Washington, D.C. the Cost of Inaction. The Institute for Allen, Robert (1980). How to Save the World. Research on Public Policy, Halifax, Barnes and Noble Books, Totowa, New Nova Scotia. Jersey. [Based on IUCN (1980)1 Brown, Lester R. (1981). Building a Sustainable Ayres, Robert U. (1978) Resources, Society. Norton & Co / Worldwatch Environment and Economics Institute, New York. -Applications of the Materials/Energy Brown, Becky J. et al (1987). "Global Balance Principle. Wiley-Interscience, sustainability: toward definition." New York. Environmental Management 11(6), 713- (1988a). "Optimal investment policies 719. with exhaustible resowces: an Burmeister, Edwin and Dobell, A. Rodney information-based model." Journal of (1970). Mathematical theories of Environmental Economics and economic growth. Macmillan, London. Management 15(4). Burness, H.S. and Cummings, R.G. (1986). (1988b). Personal communication. "Thermodynamic and economic concepts Carnegie-Mellon University, Pittsburgh. as related to resource-use policies: Ayres, Robert U. and Miller, Steven M. (1980). reply." Land Economics 62(3), 323-324. 'The role of technological change. " Clark, Colin W. (1976). Mathematical Journal of Environmental Economics bioeconomics: the optimal management and Management 7(4), 353-371. of renewable resources. Wiley- Ayres, Robert U. and Sandilya, Manalur S. Interscience, New York. (1987). "Utility maximizatlon and Clark, C.W., Clarke, F.H. and Munro, G.R. catastrophe aversion: a simulation test." (1979). "The optimal exploitation of Journal of Environmental Economics renewable resource stocks: problems of and Management 14(4), 337-370. irreversible investment.' Econometrica Barbier, Edward B. (1987). "The concept of 47, 25-49. sustainable economic development." Clark, C.W. and Munn, R.E. (1986). Environmental Conservation 14(2), 101- Sustainable Development of the 110. Biosphere. Cambridge University Press. iBartelmus, Peter (1986). Environment and Cleveland, Harlan (ed.) (1979). The Development. Allen and Unwin, Management of Sustainable Growth. London. Pergamon Press, New York. Baumrol, William J. (1986). "On the possibility Coase, R.H. (1960). "The problem of social of continuing expansion of finite cost." Journal of Law and Economics 3, resources." Kyldos 39, 167-179. 1-44. Becker, Robert A. (1982). "Intergenerational Cole, H.S.D. et al (1973). Models of Doom: A equity: the capital-environment trade- Critique of The Limits To Growth. off." Journal of Environmental Universe Books, New York. Economics and Management 9(2), 165- Coomer, James C. (1979). "The nature of the 185. quest for a sustainable society." In Coomer, J.C. (ed.), Quest for a Sustainable Society. Pergamon Press, New York. References 51 Daly, Herman E. (1986). "Thermodynamic and Goodland, R. and Ledec, G. (1987). economic concepts as related to "Neoclassical economics and principles resource-use policies: comment." Land of sustainable development." Ecologi'cal Economics 62(3), 319-322. Modelling 38, 19-46. (1987). "The economic growth debate: Hartwick, John M. (1977). "Intergenerational what some economists have learned but equity and the investing of rents from many have not." Journal of exhaustible resources." American Environmental Economics and Economic Review 67(5), 972-974. Management 14(4), 323-336. - (1978). "Substitution among exhaustiible (1988). "On sustainable development and resources and intergenerational equity." national accounts." In Collard, D., Review of Economic Studies 45(2), 347- Pearce, D. and Ulph, D. (eds.), 354. Economics, Growth and Sustainable Howe, Charles W. (1979). Natural Resource Environments. St Martin's Press, New Economics-Issues, Analysis and Policy. York. John Wiley & Sons, New York. d'Arge, Ralph C. (1989). "Ethical and economic IUCN (International Union for the Conservation systems for managing the global of Nature and Natural Resources) commons." In Botkin, D. (ed.), (1980). World Conservation Strategy: Changing the Global Environment: Living Resource Conservation for Perspective on Hwnan Involvement. SustainableDevelopment. IUCN-UNEP- Academic Press, Boston. WWF, Gland, Switzerland. d'Arge, Ralph C. and Kogiku, K.C. (1973). Kamien, Morton I. and Schwartz, Nancy L. "Economic growth and the (1982). "The role of common property environment." Review of Economic resources in optimal planning models Studies 40, 61-77. with exhaustible resources." In Smith, Dasgupta, Partha and Heal, Geoffrey M. V.K. and Krutilla, J.V. (edls.), (1979). Economic Theory and Explorations in Natural Resource Exhaustible Resources. Nisbet/CUP. Economics. Johns Hopkins University Dixit, A.K. (1976). 7he Iheory of Equilibrium Press, Baltimore. Growth. Oxford University Press, Kemp, Murray C., Long, Ngo Van and London. Shimomura, Kazuo (1984). "The Easterlin, Richard A. (1974). "Does economic problem of survival: a closed economy." growth improve the human lot? Some In Kemp, M.C. and Long, N.V. (eds.), empirical evidence." In David, P.A. and Essays in the Economics of Exhaustible Reder, Melvin W. (eds.), Nations and Resources. North-Holland, Amsterdam. Households in Economic Growth. Kolb, Jeffrey A. and Scheraga, Joel D. (1988). Academic Press, New York. "A suggested approach for discounting Forster, Bruce A. (1973). "Optimal capital the benefits and costs of environmental accumulation in a polluted regulations." Mimeo, Environmental environment." Southern Economic Protection Agency, Washington, D.C. Journal 39(4), 544-547. Koopmans, Tjalling C. (1977). "Concepts of Georgescu-Roegen, Nicholas (1988). "About optimality and their uses." American economic growth-a variation on a Economic Review 67(3), 261-274. theme by David Hilbert." Economic Kneese, Allen and Kopp, Ray (1988). Personal Development and Cultural Change 36(3) communication. Resources for the Supplement, S291-S307. Future, Washington D.C. Gibbons, Joel (1984). "Capital-energy Krautkraemer, Jeffrey A. (1985). "Optimal substitution in the long run." Energy growth, resource amenities and the Journal 5, 109-118. preservation of natural environments." Review of Economic Studies 52, 15' 3-70. 52 References (1986). "Optimal depletion with resource - . (1985). "Environmental economics: an amenities and a backstop technology." evolutionary critique and a plea for Resources and Energy 8, 133-149. pluralism." Journal of Environmental Krutila, John V. (1967). "Conservation Economics and Management 12(4), 382- reconsidered." American Economic 394. Review 57, 777-786. OECD (1972). Annex to Recommendation on Lehman, Dale E. (1981). "Doomsday" Guiding Principles Concerning reconsidered." Resources and Energy 3, International Economic Aspects of 337-357. Environmental Policies. OECD, Paris. Liverman, Diana M. et al (1988). "Global Okuguchi, Koji (1979). "Technical progress, sustainability: toward measurement." population growth and intergenerational Environmental Management 12(2), 133- equity in a model with many exhaustible 143. and renewable resources." Economics Lusky, Rafael (1975). "Optimal taxation policies Letters 3, 57-60. for conservation and recycling." Journal Olson, Mancur and Bailey, Martin J. (1981). of Economic Theory 11, 315-328. "Positive time preference." Journal of Maler, Karl-Goran (1974). Environmental Political Economy 89(1), 1-25. Economics: a Theoretical Enquiry. Johns O'Riordan, Timothy (1988). "The politics of Hopkins University Press, Baltimore. sustainability." In Turner R.K. (ed.), Marglin, Stephen A. (1963a). "The social rate SustainableEnvironmentalManagement: of discount and the optimal rate of Principles and Practice. Belhaven Press. investment." Quarterly Journal of London. Economics 77(February), 95-111. Page, Talbot (1977). Conservation and (1963b). "The opportunity costs of public Economic Efficiency. Johns Hopkins investment." Quarterly Journal of University Press, Baltimore. Economics 77(May), 274-289. - . (1983). "Intergenerational justice as Markandya, Anil and Pearce, David W. opportunity." In Maclean, D. and (1988a). "Natural environments and the Brown, P.G., Energy and the Future. social rate of discount. Project Appraisal Rowman and Littlefield, Totowa, New 3(1), 2-12. Jersey. -.(1988b). "Environmental considerations and --- .(1988). "Intergenerational equity and the the choice of the discount rate in social rate of discount." In Smith, V.K. developing countries." Environment (ed.), Environmental Resources and Working Paper 3. World Bank, Applied Welfare Economics. Resources Environment Department. Washington for the Future, Washington D.C. D.C. Parfit, Derek (1983). "Energy policy and the Meadows, Donella H. et al (1972). The Limits further future: the social discount rate." to Growth. Universe Books, New York. In Maclean, D. and Brown, P.G., Morey, Edward R. (1985). "Desertification Energy and the Future. Rowman and from an economic perspective." Littlefield, Totowa, New Jersey. Ricerche Economiche 39(4), 550-560. Pearce, David W. (1976). "The limits of cost- Nordhaus, William (1986). "Resources, benefit analysis as a guide to technology and development: will the environmental policy." Kyldos 29(1), table be bare when poor countries get 97-112. there?" Indian Economic Review 21, 81- - . (1983). "Ethics, irreversibility, future 94. generations and the social rate of Norgaard, Richard B. (1984). "Coevolutionary discount." International Journal of development potential."LandEconomics Environmental Studies 21, 67-86. 60(2), 160-173. - . (1987). "Foundations of an ecological economics." Ecological Modelling 38, 9-18. References 53 (1988a). "Optimal prices for sustainable . (1985b). "Paying The Price-Pesticide development." In Collard, D., Pearce, Subsidies in Developing Countries." D. and Ulph, D. (eds.), Economics, Research Report #2, World Resources Growth and Sustainable Environments. Institute, Washington, D.C. St Martin's Press, New York. . (1986a). World Enough and -- . (1988b). "The sustainable use of natural Time-Successfil StrategiesforResoturce resources in developing countries." In Management [condensed version of 'The Turner, R.K. (ed.), Sustainable Global Possiblel. Yale University Press, Environmental Management: Principles New Haven. and Practice. Belhaven Press, London. . (1986b). "Skimming The. Water-Rent- Pearce, David W., Barbier, Edward and seeking and the Performance of Public Markandya, Anil (1988). "Sustainable Irrigation Systems." Research Report development and cost benefit analysis." #4, World Resources Institute, LEEC Paper 88-03. London Washington, D.C. Environmental Economics Centre, . (1988a). "The Forest For The University College London'. Trees-Government Policies and the Pearce, David W., and Markandya, Anil Misuse of Forest Resources." World (1987). "Marginal opportunity cost as a Resources Institute, Washington, D.C. planning concept in natural resource . (1988b). "Economic policy reform for management." Annals of Regional natural resource conservation." Science 21(3), 18-32. Environment Working Paper 4. World Perrings, Charles (1987). Economy and Bank, Environment Department. Environment. Cambridge University Washington D.C. Press. Repetto, Robert and Magrath, William B. Pezzey, John C.V. (1988). Market mechanisms (1988). "Natural Resources of pollution control: "polluter pays", Accounting." World Resources Institute, economic and practical aspects." In Washington, D.C. Turner, R.K. (ed.), Sustainable Rescher, Nicholas (1980). "Technological Environmental Management: Principles progress and human happiness." In and Practice. Belhaven Press, London. Rescher, N., Unpopular Essays on Pigou, A.C. (1932). Ihe Economics of Welfare. Technological Progress. University of Macmillan, London. Pittsburgh Press. Pirages, Dennis C. (1977). "A social design for Riley, John G. (1980). "The just rate of sustainable growth." In Pirages, D.C. depletion of a natural resource." Joutrnal ( e d . ), The Sustainable of Environmental Economics and Society-Implications for Limited Management 7(4), 291-307. Growth. Praeger, New York. Runge, Carlisle F. (1986). "Common prolperty Porritt, Jonathon (1984). Seeing Green-The and collective action in economic Politics of Ecology Explained. Basil development." World Development Blackwell, Oxford. 14(5), 623-635. Rawls, John (1971). A Theory of Justice. Sen, Amartya K. (1967). "Isolation, assurance, Harvard University Press. and the social rate of discount." Redclift, Michael (1987). Sustainable Quarterly Journal of Economics 81(1), Development-Exploring the 112-124. Contradictions. Methuen, London. Siebert, Horst (1987). Economics of the Repetto, Robert (1985a). The Global Environment (2nd edition). Springer- Possible-Resources, Development and Verlag, Berlin. the New Century. Yale University Press, New Haven. 54 References Smith, Vernon L. (1977). "Control theory Tietenberg, T.H. (1984). Environmental and applied to natural and environmental Natural Resource Economics. Scott, resources-an exposition." Journal of Foresman, & Co, Glenview, Illinois. Environmental Economics and Tolba, Mostafa K. (1987). Sustainable Management 4, 1-24. Development-Constraints and Smith, V. Kerry (ed.) (1979). Scarcity and Opportunities. Butterworths, London. Growth Reconsidered. Johns Hopkins Tonn, Bruce E. (1988). "Philosophical aspects University Press, Baltimore. of 500-year planning. " Environment and Solow, Robert M. (1974a). "The economics of Planning A 20, 1507-1522. resources or the resources of Turner, R. Kerry (1987). "Sustainable global economics." American EconomicReview futu res-common interest, 64(2), 1-14. interdependency, complexity and global (1974b). "Intergenerational equity and possibilities." Futures 19(5), 574-582. exhaustible resources." In Symposium on --- . (1988). "Sustainability, resource the Economics ofExhaustible Resources. conservation and pollution control: an Review of Economic Studies. overview." In Turner, R.K. (ed.), -. (1986). "On the intergenerational allocation SustainableEnvironmentalManagement: of natural resources." Scandinavian Principles and Practice. Belhaven Press. Journal of Economics 88(1), 141-149. London. Southgate, Douglas and Runge, Carlisle (1985). Vousden, Neil (1973). "Basic theoretical issues "Toward an economic model of of resource depletion." Journal of deforestation and social change in Economic Theory 6, 126-143. Amazonia." Ricerche Economiche39(4), WCED-The World Commission on 561-567. Environment and Development (1987). Spash, Clive L, and d'Arge, Ralph C. (1989). "Our Common Future. Oxford "The greenhouse effect and University Press." ["The Brundtland intergenerational transfers." Energy Report"] Policy 17(2), 88-96. World Bank (1987). Environment, Growth, and Stiglitz, Joseph (1974). "Growth with Development. Development Committee exhaustible natural resources: efficient Pamphlet 14, World Bank, Washington and optimal growth paths." In D.C. Symposium on the Economics of - . (1988). Environment and Development: Exhaustible Resources. Review of Implementing the World Bank's New Economic Studies Policies. Development Committee Strotz, R.H. (1955/56). "Myopia and Pamphlet 17, World Bank, Washington inconsistency in dynamic utility D.C. maximization." Review of Economic Zarembka, Paul (1972). Toward a theory of Studies 23(3), 165-180. economic development. Holden-Day, Talbot, Lee M. (1984). "The World San Francisco. Conservation Strategy." In Thibodeau, Francis R. and Field, Hermann H., Sustaining Tomorrow-A Strategy for World Conservation and Development. University Press of New England. I A later version of the main technical parts of this paper is Terkla, David (1984). "The efficiency value of Barbier, E. BB., Markandya, A. and Pearce, D. W. (1990). Enviromnental sustainability and cost-benefit analysis. Environnent effluent tax revenues." Journal of and Planning A 22, 1259-1266. Environmental Economics and Management 11(2), 107-123. 55 Appendix 1 Definitions of sustainability in the literature This Appendix is not exhaustive, but it gives a good idea of the variety of definitions of sustainability concepts that have appeared in the last decade, and of the people who use such concepts.. (Bold type is added where appropriate to emphasize which concepts are being defined.) Allen (1980)-summarizing IUCN (1980) 1. 'Sustainable utilization is a simple idea: we should utilize species and ecosystems at levels and in ways that allow them to go on renewing themselves for all practical purposes indefinitely." (p18) 2. "The importance of ensuring that utilization of an ecosystem or species is sustainable varies with a society's dependence on the resource in, question. For a subsistence society, sustainable utilization of most, if not all its living resources is essential... .The greater the diversity and flexibility of the economy, the less the need to utilize certain resources sustainably, but by the same itoken the less the excuse not to." (p18) 3. "...it is essential... .to ensure that... .people protect those parts of the biosphere that need protecting and modify the rest only in ways that it can sustain." (p20) 4. "sustainable development-development that is likely to achieve lasting satisfaction of human needs and improvement of the quality of human life" (p23) Barbier (1987)-academic economist 1. "...the concept of sustainable economic development as applied to the Third World... is therefore directly concerned with increasing the material standard of living of the poor at the 'grassroots' level, which can be quantitatively measured in terms of increased food, real income, educational services, health-care, sanitation and water supply, emergency stocks of food and cash, etc., and only indirectly concerned with economic growth at the aggregate, commonly national, level. In general terms, the primary objective is reducing the absolute poverty of the world's poor through providing lasting and secure livelihoods that minimize resource depletion, environmental degradation, cultural disruption and social instability." (plO3) Brown et al (1987)-environmental scientists 1. UIn the narrowest sense, global sustainability means the indefinite survival of the human species across all the regions of the world. A broader sense of the meaning specifies that virtually all humans, once born, live to adulthood and that their lives have quality beyond mere biological survival. Finally the broadest sense of global sustainability includes the persistence of all components of the biosphere, even those with no apparent benefit to humanity." (p717) Burness and Cummings (1986)-academic economists 1. 'Professor Daly's notion of "sustainability' [in Daly 19861 is extraordinarily vague and ill-defined.. .in a pedagogical sense sustainability requires that all 56 Appendix 1: Definitions of sustainability in the literature processes operate only at their steady state, renewable level, which might then suggest a return to a regulated caveman culture." (p323) Clark (1986)-environmental scientist and policy analyst, IIASA 1. 'A major challenge of the coming decades is to learn how long-term, large-scale interactions between environment and development can be better managed to increase the prospects for ecologically sustainable improvements in human well-being." (p5) Coomer (1979) 1. '[The] sustainable society is one that lives within the self-perpetuating limits of its environment. That society.. .is not a 'no-growth' society... .It is, rather a society that recognizes the limits of growth.. .[and] looks for alternative ways of growing." (p1) Daly-academic economist 1. "The market does not distinguish an ecologically sustainable scale of matter- energy throughput from an unsustainable scale, just as it does not distinguish between ethically just and unjust distributions of income. Sustainability, like justice, is a value not achievable by purely individualistic market processes." (1986, p320). 2. 'By 'growth' I mean quantitative increase in the scale of the physical dimensions of the economy; ... By 'development' I mean the qualitative improvement in the structure, design and composition of physical stocks and flows, that result from greater knowledge, both of technique and of purpose." (1987, p323) Georgescu-Roegen (1988)-academic economist 1. '...'growth' is if you get just an increasing number of the same type of mail coaches. And if you pass from traveling in mail coaches to traveling by railway, that is 'development'. (pS294) Goodland and Ledec (1987)-institutional environmental scientists 1. 'Sustainable development is here defined as a pattern of social and structural economic transformations (i.e. 'development') which optimizes the economic and societal benefits available in the present, without jeopardizing the likely potential for similar benefits in the future. A primary goal of sustainable development is to achieve a reasonable (however defined) and equitably distributed level of economic well-being that can be perpetuated continually for many human generations." (p36) 2. ..-sustainable development implies using renewable natural resources in a manner which does not eliminate or degrade them, or otherwise diminish their usefulness for future generations....Sustainable development further implies using non-renewable (exhaustible) mineral resources in a manner which does not unnecessarily preclude easy access to them by future generations... .Sustainable Appendix 1: Definitions of sustainability in the literature 57 development also implies depleting non-renewable energy resources at a slow enough rate so as to ensure the high probability of an orderly societal transition to renewable energy sources...." (p37) Howe (1979)-academic economist 1. 'Guidelines for a responsible natural resources policy (6) ...activities should be considered that would be aimed at maintaining over time a constant effective natural resource base. This concept was proposed by Page (1977) and implies not an unchanging resource base but a set of resource reserves, technologies, and policy controls that maintain or expand the production possibilities of future generations." (p337) Markandya and Pearce (1988a)-academic economists 1. 'The basic idea [of sustainable development] is simple in the context of natural resources (excluding exhaustibles) and environments: the use made of these inputs to the development process should be sustainable through time... .If we now apply the idea to resources, sustainability ought to mean that a given stock of resources-trees, soil quality, water and so on-should not decline." (pp9-10). 2. ". .sustainability might be redefined in terms of a requirement that the use of resources today should not reduce real incomes in the future...". (p1) Morey (1985)-academic economist 1. "...much of the desertification literature also suggests that desertification is nonoptimal from both the producer's and society's perspective. Sustainable use is generally put forward as the optimal strategy." [Morey then shows how sustainable land use may or may not be optimal] (p551) O'Riordan (1988)-academic environmental scientist 1. UIt may only be a matter of time before the metaphor of sustainability becomes so abused as to be meaningless, certainly as a device to straddle the ideological conflicts that pervade contemporary environmentalism." (p29) 2. uSustainability is a much broader phenomenon [than sustainable development], embracing ethical norms pertaining to the survival of living matter, to the rights of future generations and to institutions responsible for ensuring that such rights are fully taken into account in policies and actions." (p30) Pearce-academic economist 1. 'The sustainability criterion requires that the conditions necessary for equal access to the resource base be met for each generation." (1987, p13). 2. 'In simple terms [sustainable development] argues for (a) development subject to a set of constraints which set resource harvest rates at levels no higher than managed or natural regeneration rates; and (b) use of the environmenat as a 'waste sink' on the basis that waste disposal rates should not exceed rates of (natural or managed) assimilation by the counterpart ecosystems....There are self- evident problems in advocating sustainable rates for exhaustible resources, so that 'sustainabilists' tend to think in terms of a resource set encompassing 58 Appendix 1: Definitions of sustainability in the literaure substitution between renewables and exhaustibles. Equally self-evident is the implicit assumption that sustainability is a 'good thing'-that is optimizing within sustainable use rates is a desirable objective. On these terms, sustainability could imply use of envirommental services over very long time periods and, in theory, indefinitely." (1988a, p58) 3. 'The key concept [regarding natural resource degradation in developing countries] is 'sustainability'. Changes in resource management practice toward sustainable resource use could at least contribute to the preservation of the renewable resource base, and hence to the direct well-being of the population and to the future of the macroeconomy." (1988b, p102) Pearce, Barbier and Markandya (1988)-academic economists 1. 'We take development to be a vector of desirable social objectives, and elements might include: * increases in real income per capita * improvements in health and nutritional status * educational achievement * access to resources 3 a 'fairer' distribution of income * increases in basic freedoms. ... Sustainable development is then a situation in which the development vector increases monotonically over time." (p4) 2. 'We summarize the necessary conditions [for sustainable development] as 'constancy of the natural capital stock'. More strictly, the requirement as for non-negative changes in the stock of natural resources such as soil and soil quality, ground and surface water and their quality, land biomass, water biomass, and the waste assimilation capacity of receiving environments." (p6) Pirages (1977)-from conference funded by the Institute for World Order I1. '[Sustainable growth] means economic growth that can be supported by physical and social environments in the foreseeable future. An ideal sustainable society would be one in which all energy would be derived from current solar income and all non-renewable resources would be recycled." (pplO-11) Porritt (1984)-Director, U.K. Friends of the Earth 1. "All economic growth in the future must be sustainable: that is to say, iit must operate within and not beyond the finite limits of the planet." (p120) Repetto (1985a)-economist, World Resources Institute. Also in Repetto (1986a), ppl16-17 1. "The core of the idea of sustainability, then, is the concept that current decisions should not impair the prospects for maintaining or improving future living standards... .This implies that our economic systems should be managed so that we live off the dividend of our resources, maintaining and improving the asset base. This principle also has much in common with the ideal concept of income that accountants seek to determine: the greatest amount that can be consumed in the current period without reducing prospects for consumption in the future." (pl0) Appendix 1: Definitions of sustainability in the literature 59 2. "This does not mean that sustainable development demands the preservation of the current stock of natural resources or any particular mix of human, physical and natural assets. As development proceeds, the composition of the underlying asset base changes." (plO) 3. "There is broad agreement that pursuing policies that imperil the welfare of future generations, who are unrepresented in any political or economic forum, is unfair." (p11) Redelift (1987)-academic economist 1. "...to what extent is economic growth an adequate measure of development?" (p15) Solow (1986)-Nobel Prize academic economist 1. '.. . a society that invests in reproducible capital the competitive rents on its current extraction of exhaustible resources, will enjoy a consumption stream constant in time... .This result can be interpreted as saying that an appropriiately defined stock of capital-including the initial endowment of resources-is lbeing maintained intact, and that consumption can be interpreted as the interest on that patrimony." (pl41). Talbot (1984)-former Director-General, IUCN 1. "Objectives of the world conservation strategy Conservation has three basic objectives: (1) To maintain essential ecological processes and life support systems. (2) To preserve genetic diversity. (3) To ensure that the utilization of living resources, and the ecosystems in which they are found, is sustainable.' (p4) Tietenberg (1984)-academic economist 1. "The sustainability criterion suggests that, at a minimum, future generations should be left no worse off than current generations." (p33) 2. "Rather than eliminating the [positive] discount rate, the present-value criterion should be complemented by other criteria, such as sustain-ability... .For example, we might choose to maximize present value subject to the constraint that future generations are not made worse off". (p432) Tolba (1987)-Executive Director, U.N. Environmental Programme. 1. "[Sustainable development] has become an article of faith, a shibboleth: often used but little explained. Does it amount to a strategy? Does it apply only to renewable resources? What does the term actually mean? In broad terms the concept of sustainable development encompasses: (1) help for the very poor because they are left with no option other than to destroy their environment; 60 Appendix 1: Definitions of sustainability in the literature (2) the idea of self-reliant development, within natural resource constraints; (3) the idea of cost-effective development using different economic criteria to the traditional approach; that is to say development should not degrade environmental quality, nor should it reduce productivity in the long run; (4) the great issues of health control, appropriate technologies, food self-reliance, clean water and shelter for all; (5) the notion that people-centered initiatives are needed; human beings, in other words, are the resources in the concept." (p98) Tonn (1988) 1. 'Two principles of 500-year planning: * Principle 1: Future generations should not inherit, from present generations, unacceptable risks of death owing to environmental or other preventable catastrophes. * Principle 2: Future, as well as present, generations may inherit constraints on their primary freedoms as sacrifices for enjoying the conditions of Principle 1." (6th page of article) Turner-academic economist 1. "The World Conservation Strategy.. .gave considerable prominence to the sustainability concept, although its precise meaning and practical applications were not presented in a detailed and operational form." (1987, p576) 2. "The precise meaning of terms such as 'sustainable resource usage', 'sustainable growth' and 'sustainable development' has so far proved elusive." (1988, p5). 3. "In principle, such an optimal [sustainable growth] policy would seek to maintain an 'acceptable' rate of growth in per-capita real incomes without depleting the national capital asset stock or the natural environmental asset stock." (1988, p12) 4. "It makes no sense to talk about the sustainable use of a non-renewable resource (even with substantial recycling effort and reuse rates). Any positive rate of exploitation will eventually lead to exhaustion of the finite stock." (1988, p13) 5. "...in this [sustainable development] mode.. .conservation becomes the sole basis for defining a criterion with which to judge the desirability of alternative allocations of natural resources." (1988, p21). WCED (1987) [Brundtland Report] 1. 'We came to see that a new development path was required, one that sustained human progress not just in a few places for a few years, but for the entire planet into the distant future. Thus 'sustainable development' becomes a goal not just for the 'developing' nations, but for industrial ones as well." (p4) 2. "Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs. It contains within it two key concepts: * the concept of 'needs', in particular the essential needs of the world's poor, to which overriding priority should be given; and * the idea of limitations imposed by the state of technology and social organization on the environment's ability to meet present and future needs." (p43) 3. "Even the narrow notion of physical sustainability implies a concern for social equity between generations, a concern that must logically be extended to equity within each generation." (p43) Appendix 1: Definitions of sustainability in the literature 61 4. 'Living standards that go beyond the basic minimum are sustainable only if consumption standards everywhere have regard for long-term sustainablity. Yet many of us live beyond the world's ecological means, for instance in our patterns of energy use. Perceived needs are socially and culturally determined, and sustainable development requires the promotion of values that encourage consumption standards that are within the bounds of the ecological possible and to which all can reasonably aspire." (p44) 5. 'Economic growth and development obviously involve changes in the physical ecosystem. Every ecosystem everywhere cannot be preserved intact." (p45) 6. 'The loss [i.e. extinction] of plant and animal species can greatly limit the options of fiuture generations; so sustainable development requires the conservation of plant and animal species." (p46) World Bank 1. "...satisfy the multiple criteria of sustainable growth, poverty alleviation, and sound environmental management." (1987, plO) 2. "To a large degree, environmental management should be seen as a means of attaining the wider objectives of sustained economic growth and poverty alleviation." (1987, pl8) 3. "...elevating concern about environmental matters...and developing the capacity to implement sound practices for environmental management.. .are [both] needed to reconicile, and, where appropriate, make tradeoffs among the objectives of growth, poverty alleviation, and sound environmental management." (1987, p28) Assertions of economy-environment interactions Allen (1980)-summarizing IUCN (1980) 3. 'development.. .depends upon conservation, and that conservation depends equally upon development." (p9) 4. "conservation of the biosphere is a prerequisite for human survival and well-being; ..interdependence is an inescapable fact of life." (pl6) Bartenwus (1986) 1. "...tie overall goals of environment and development are not in conflict but are indeed the same, namely the improvement of the human quality of life or welfare for present and future generations." (ppl3-14) Clark (1986)-environmental scientist and policy analyst, IIASA 2. "Throughout most of history, the interactions between human development and the environment have been relatively simple and local affairs. But the complexity and scale of these interactions are in-creasing... .What were once straightforward questions of ecological preservation versus economic growth now reflect complex linkages-witness the feedlbacks among energy and crop production, deforestation and climatic change that are evident in studies of the atmospheric 'greenhouse' effect." (p5) Tolba (1987)-Executive Director, UNEP 2. '...economic development and environmental quality are interdependent and, in thte long term, mutually reinforcing. The rational management of the world's threatened natural 62 Appendix 1: Definitions of sustainability in the literature resource base forestalls a loss in environmental quality and enhances sustainable economic growth." (pl50) WCED (1987) 7. "...it is impossible to separate economic development issues from environment issues; many forms of development erode the environmental resources upon which they must be based, and environmental degradation can undermine economic development. Poverty is a major cause and effect of global environmental problems." (p3) World Bank 4. "Promoting growth, alleviating poverty, and protecting the environment are mutually supportive objectives in the long run... .In the short run, however, the objectives are not always compatible..." (1987, pS) 5. 'Poverty-of people and of countries-is thus a major cause of environmental degradation. That makes it essential, if environmental degradation is not to become completely unmanageable, to devise policies oriented toward economic growth, with special emphasis on improving the incomes of the poor... .Nevertheless economic growth may also destroy the environment and further jeopardize the already tenuous lives of the poor... .Thus, although growth is imperative for alleviating poverty, it may also adversely affect the poor and the environment if inadequate attention is paid to the poor and their needs." (1987, pp6-7) 6. "...economic growth, the alleviation of poverty, and sound environmental management are in many cases mutually consistent objectives." (1988, pl) 63 Appendix 2 Cake-eating model with no environmental effects The following example analyses a cake-eating economy, with exogenous technological progress in consuming ("eating") a non-renewable natural resource ("cake") and a large number N of economically identical agents who have explicit, purely materialistic preferences. This economy is derived from a more general model in Krautkraemer (1985). There are no environmental" effects on agents' utility functions or on the cake- eating process, although the dependence of consumption on eating into a finite, non- renewable cake is clearly a natural resource constraint. Intertemporal preferences are assumed to be dynamically consistent, so that the utility discount rate a is a constant, as shown by Strotz (1956). See Figure 7 for illustration. Each agent is assumed to eat into his personally-owned stock of cake s(t) (dropping the subscript n for non-renewable) so as to maximize the present discounted value of utility over an infinite time horizon, i.e. to choose s*(t) to maximize I :u[c]e4dt, where s = per capita stock of non-renewable natural resource u = per capita utility c = per capita consumption 6 = utility discount rate, > 0 t = time u[cl = c', 0 < v < 1 (this ensures the diminishing marginal utility of consumption) c = -ge!l (a rudimentary production function, assuming exogenous technological progress in 'cake-eating" at a constant rate T > 0) c 2 0 and s 2 0, v t > 0 (non-negativity constraints) s(O) = so > 0 (initial condition). Using Euler's equation, the differential equation for the individually optimal time path of the resource stock s*(t) is then: S*S* + [(6-TV)/(1-V)] *S* = 0 which has the following solution: s*(t)= soe-', c*(t) = *soe"(3', u*(t) a et°>^ where + = (-urv)/(1-v) and hence (r-b) = (T-)I(1-v) Hence there are two cases here (assuming that 6 > TV for convergence) A aj j_. In this case the optimal path has a 2 0, i.e. is 'sustainable". B a >_T. In this case the optimal path has i < 0, and so is unsustainable. 64 Appendix 3 Cake-eating with environmental amenity or productivity This model shows a very simple example of how environment/economy interactions can alter the simple cake-eating model of Appendix 2 and make it less sustainable. We use specific functional forms for the interactions, so the results are not general, but nevertheless quite suggestive. Suppose: * either that the utility function is u = cvS' instead of u = ce, where S = Ns is the total stock of the natural resource, which now has both direct amenity value as well as productive uses, and S' (e > 0) is the new environrental amenity term, as shown in Figure 8; * or that production function is c = -Sc'/u§e" instead of c = -te, with SI" as the environmental productivity term, as in Figure 9. Either way-this particular example does not actually distinguish between amenity and productivity effects-the optimization (present value maximization) problem for any individual becomes: find s*(t) to maximize I . (-s)"SleMdt (notation as in Appendix 2) (Assume that 6 > Tv for convergence of this integral; also require v + c < 1.) We assume that the economy comprises a large number N of identical non- cooperating individuals, who act as if as/as = 0: this non-cooperation is crucial in generating nonoptimal depletion of the environmental resource. Using Euler's equation, the differential equation for the optimal time path of the resource stock s*(t) can then be shown to be: § - [C/(l_u)](§*)2 + [(6_TV)/(1_v)]*S* = 0 which has the following solution: s*(t)= s0e-", c*(t) = s0e-3% u*(t) a e( *; hence this non-cooperative solution is sustainable if and only if T > 6(1+ Z/v) > 6. We can also show that the requirement for convergence of the utility integral, fi*/u* < a is equivalent to 6 > Tv, which then ensures that a(1l/u*)/8e < 0. That is, utility growth is reduced by a stronger environmental effect c, and a higher minimum rate of technical progress is required to just achieve sustainability in this case than in the purely materialist model in Appendix 2 above. On the other hand the social optimum, which Appendix 3: Cake-eating with environmental amenity or productivity 65 would be achieved if individuals behaved cooperatively as if 8S/Os = N, has a differential equation §*S* + (C1v)(§*)2 + [(6_ru)/(l-v)]s*s* = 0 which has the following solution: s*(t)= soe4, c*(t) = 6soe(r't, u*(t) at e@(,'e'*, ui*/u* = vr-(v+e)o = v(QT-)I(1-v) where 0 = v(5-'ru)/[(l-u)(v+e)] < *; hence this social optimum solution is sustainable if and only if T > 6, as in Appendix 2. There are thus three policy cases here, illustrated in Figure 10, with Case A of Appendix 2 now split into two sub-cases. (We expect this threefold classification to apply to many 'tragedies of the commons' cases where externalities and non-cooperative behavior lead to a nonoptimal profile of resource depletion.) Al If 6 5 T(1 + el/), the non-cooperative path is of course nonoptimal but has u 2 0, i.e. is sustainable". A2 If T/( + CIV) < 6 9 T, the non-cooperative path is both nonoptimal and has i < 0, i.e. is unsustainable, while the socially optimal path is sustainable. B If 6 .j, both the non-cooperative and socially optimal paths have u < 0, i.e. are unsustainable. What sort of policy options are available to achieve sustainability in our simple cake- eating economy? Solow (1974a, p12) suggests a policy solution of resource conservation subsidies, or of severance (resource depletion) taxes that fall through time. Still assuming economically identical, non-cooperative agents, consider therefore the effect of government conservation subsidies first, applied in a zero-revenue form. Assuming that agents do not cooperate (i.e. make a zero conjectural variation), this means that the individual perceives an increased incentive to conserve resources while the government's budget remains balanced. A zero-revenue resource conservation subsidy applied at rate % thus changes an agent's perception of the cake-eating relationship from § = -ce- to s = -ce- + a(s-SlN) Here the total resource stock S = Ns in aggregate, but an agent acts as if 8S/Os = 0 (we are assuming N is so large that an agent even ignores the effect that his own stock depletion actually has on the total stock S). This then can be shown to transform the differential equation for the optimal stock path s*(t) to s - [e/(I-v)](g*)2 + [(6a_or_)/(l_V)Ig*s* = 0 which has the solution 66 Appendix 3: Cake-eating with environmental amenity or productivity s*(t) = s0e-; u*(t) ea (O')"e[(")"-'13 where ' = (6-u-rv)I(1-v-e) and hence ui*/u* = (T-fl')v-¶'e = [m-(&a)(u+e)]I(I-u-e). Therefore, to just achieve sustainability (ui = 0) we need a subsidy rate or. = 6 - rI(l+ Cv) > a - r; whereas to achieve optimnality s* = se- = sOe7-)t1(1-0(u+9n we need O' = 0, i.e. a.,, = (8-rv)/[(1+v/e)(1-v)], > O if and only if 6 > TV. Note that when 6 = r we have or.. = °opt, as expected since then the optimal solution is just sustainable. It can be shown that Solow's alternative suggestion of a zero-revenue resource depletion tax, at a proportional rate X(t), can be related to the privately optimal path s*(t) by: ,$.(t)I[R(t)-1I = (l-v)§*/s* - (rTV5). So to get s* = soe-(+e), i.e. just sustainable we need i(t)/[X(t)-1] = -(l-v)Tvl(v+e) - (rv-5) = 6- r(l +e)/(l +elv) i.e. X,..(t) = 1 - g&it(1+i)(1+ilu)it, ko some constant and to get s* = s0e^, i.e. optimal we need X(t)/[X(t)-1] = (6-rv)I(l+v/C) i.e. X,(t) = 1 - koe("3)v(1+u-), k0 some constant Note how the depletion tax rate falls over time and may eventually become a depletion subsidy; the time profile of the tax rate is what matters. Interpretation in terms of sustaining the resource base We can interpret the above results in terms of maintaining the ueffective resource base". The remaining value of the cake resource at time t here is V(t) = I l(Se1)1'(xS)-dx in the enviromnental amenity case = I t(Se'[xS]"'1)'dx in the environmental productivity case Both integrals are of course the same. Sustaining the effective resource base requires that V 2 0, i.e. Sl+sue?l does not decline, i.e. S/S 2 -T/(l + c/v) which is same as the condition above for sustainabiity. 67 Appendix 4 Capital accumulation with environmental amenity or productivity The two models here are adaptations from Stiglitz (1974). They differ from Appendix 3 by introducing a production function requiring inputs of capital and labor, as well as the non-renewable resource flow R (= -dS/dt), in order to produce output Q. (For simplicity we continue to assume that population L is constant.) The first model has multiplicative environmental amenity and a Cobb-Douglas production function with exogenous technological progress: Utility u = c"St O < v, e < 1 Output Q = AKaLRYe' O< ca, ,S, y < 1; c + , + -y = 1; T > O (see Figure 12 for illustration of this model). The second has no amenity but multiplicative environmental productivity, so that output is affected not only by the resource flow R but also by the resource stock S: Utility u = c' O << 1 Output Q=AKLPRYSwe' O < a, ,, y, ir < 1; a + ,+ f y = 1; r > O Note that the second production function is assumed to have increasing returns to scale. This is to avoid interdependencies between parameters that would arise if we were to require that a + , + y + 7r = 1. The methodology is as in Stiglitz. It is not possible to give a general analytical solution, and so we look at possible steady state solutions which assume that all stocks and flows have constant (positive or negative) growth rates. We can then work in simple linear equations of growth rates defined as g or g, = x/x Thus for the environmental amenity model we have: g < u, > = -(1-v)g0 + eg8 where of course gs = -R/S gQ =atgK + YgR + r gK = Q/K - C/K (output not consumed leads to capital growth) QK - a = -g (Ramsey rule: excess of interest rate over utility discount rate compensates for declining marginal utility) g < QR> = QK (Hotelling rule: resource price rises at the rate of interest) For the environmental productivity model the first two equations are changed: g < ur> = -(-g 68 Appendix 4: Capital accumulation with environmental amenity or productivity gQ = gK + YgR + Trgs + r The fact that agents are non-cooperative, and therefore ignore the public environmental value of the total resource stock S, is reflected in the mathematics by using the standard Hotelling rule to determine the resource depletion rate. On the socially optimal path, the Hotelling rule should be modified to account for the social environmental value of the resource. Solving these linear equations gives the following results for the privately optimal (non-cooperative) paths: Environmental amenity model (environmental parameter is e) 6(1-a) - TV Optimal resource depletion rate - gs = [j + (l-)y - e(1-a)] T(l-v-e) + so6 Optimal real interest rate aQ/K = a[t,+ (l-u),y-c(1-a)] v(Tr-y6) - £(l-o)6 Optimal growth rate of utility gu = It is interesting to determine the effect of varying the environmental parameter e in these paths. First note that, for the above solutions to be meaningful, the integral of discounted utility (which our optimal control procedure is maximizing) must converge. This requires that the discount rate 6 must be greater than the above optimal growth rate of utility gu, which after some manipulation means that 6(j3+Y) > UT Note that (provided always that T > 0!) this justifies a positive discount rate, since a zero discount rate would lead to perpetual saving and current impoverishment, as Olson and Bailey (1981) have pointed out. Differentiating the three optimal growth rates with respect to e then gives a(-gs)/Oe > 0 a(ceQIK)lte ca 6(6+7) - vT > 0 agja/e oa Vr - #+-Y) < 0 Environmental productivity model (environmental parameter is 7r) 6(1-a) - Tv Optimal resource depletion rate -gs = [f + (l-0)y - 7rvl Appendix 4: Capital accumulation with environmental amenity or productivity 69 T(l-V) + (6-2r)b Optimal real interest rate aQ/K = aL8 + (l-v)-y - 7rv] v[r - 6( + )] Optimal growth rate of utility gu = [L + (l-v)Y - WrV] Convergence of discounted utility - 65(+ y) > vr again Response of optimal paths to change in environmental parameter 7r: a(-g,)/a7r > 0 8(aQ/K)/a-x a Tr - 6S(j+y) < 0 agu/ar a UT - 6B(J+'y) < 0 With proportional conservation subsidy a but no environmental effects: The Hotelling rule is modified to g < QR> = QK/(1 + a) with results: 5(1-a) - r(v+a) Optimal resource depletion rate - gs = [8(1+a) + (l-U)y] i(1-V) + $8 Optimal interest rate aQlK = a4 + (l-u)y/(l+o)] Ulr(l + 0)-7] Optimal growth rate of utility gu = 3(1 +au)+(l-u) y] Assume for utility convergence that 6[L+,y(l + a)] > Ur; note that a lower discount rate is required for convergence thanks to a > 0. Response of optimal paths to change in subsidy rate a: a(-g")aa la - T(1-u) - j36 < 0 0(aQ/K)!aa > 0 ag/laa a 6(1-u) + X5 > 0 Results with both environmental effects and conservation incentives still have to be computed. 70 Appendix 5 Corn-eating with a minimum subsistence level In this model population N grows exponentially at a constant, exogenous rate X, and is supported by the harvest from a single, renewable resource S C(corn"). Capital, non- renewable resources and labor inputs are ignored. There is a minimum subsistence level of consumption cm below which life is not worth living, and the utility function is purely materialistic: u a (C-C,)U, Cm > 0, 0 < V < 1; The resource has no public amenity value. See Figure 13 for illustration. The resource stock S grows naturally at exponential rate p, but is also eaten by people at rate C, with no technical progress in the consumption process: S = pS - C, so that S/S = p - C/S Since the resource clearly has a productivity value, we are assuming private ownership of the resource to avoid any nonoptimality caused by open access. Converting to per capita resource stock s = S/N and consumption c = C/N: S/S = s/s + N/N = V/s + X, and C/S = c/s (initial stock s(O) = so) 9/s + X = p-c/s - = (P-X)s - c, p, X > 0. Maximizing discounted unweighted utility Maximizing discounted utility I I u[c(t)]etdt can be shown to give the following differential equation for the optimal corn stock s(t): (1-v)s + [t-y(2-v)]9 + y(,y-b)s = (-y-b)cm wherey = p - X This is a linear equation with solutions for per capita variables: Resource stock s*(t) = cm/dy + (so - cm/y)ecM where i7 = (PA-6)/(1-v) Consumption c*(t) = cm + (-y-n)(so -cm/7)e Utility u*(t) = [(y-n)(so - cm/,y)evt]" (It can be shown that we must have -y > v for convergence of the utility integral, so we are assured that y - 'I > 0). For consumption and utility to grow sustainably we need both that Appendix 5: Corn-eating with a minimum subsistence level 71 7>0 - p>X+5 and s5 > CmI,Y. If X < 0, c approaches cm over time, in other words society is grinding along at subsistence level because it is too impatient (too high a 6) to allow sustainable growth in the resource base. In contrast, if v > 0 but sO < cmI7y, consumption crashes to zero in a finite time. This is because the initial resource stock is too small for the population too be fed from resource growth, so that people are forced to eat resource capital ("seedcorn") with inevitably disastrous consequences. Maximizing discounted utility weighted by future population levels We may instead weight future per capita utility levels by the number of people alive then, we maximize f z u[c(t)]e-(&Xtdt and simply replace a by 6-X in the above results: #'= (o-5)/(1-v) Sustainable growth now requires 7' > 0 - p > 6, a less stringent condition than before; and so > c./V, the same condition as before. Distributors of World Bank Publications ARGENTINA FINLAND MALAYSIA Forsubscr4ptw ordenr Carlos Hirsch, SRL Akateeminen Klrjakauppa University of Malaya Cooperative International Subscription Service Gaieria Guemes P.O. Box 128 Bookshop, Limited P.O. Box 41095 Florida 165,4th Floor-Ofc. 453/465 StM00101 Helsinki 10 P.O. Box 1127, Jalan Pantai Baru Craighall 1333 Buenos Aires 59700 Kuala Lumpur Johannesburg 2024 FRANCE AUSTRALIA, PAPUA NEW GUINEA, World Bank Publications MEX3CO SPAIN FIJI, SOLOMON ISLANDS, 66, avenue d'lena INFOTEC Mundi-Prensa Libros, SA- VANUATU, AND WESTERN SAMOA 75116 Paris Apartado Postal 22-860 Castello 37 D.A Books & journals 14060 Tlalpan, Mexico D.F. 28001 Madrid 648 Whitehorse Road GERMANY Mitcham 3132 UNO-Verlag NETHERLANDS Libreria Intemarconal AEDOS Victoria Poppelsdorfer Allee 55 De lindeboom/InOr-Publikaties Consell de Cent, 391 D>5300 Bonn 1 P.O. Box 202 08009 Barcelona AUSTRIA 7480 AE Haaksbergen Gerold and Co. HONG KONG, MACAO SRI LANKA AND THE MALDIVES Graben 31 Asia 2000 Ltd. NEW ZEALAND Lake House Bookshop A-1011 Wien 46-48 Wyndham Street EBSCO NZ Ltd. P.O. Box 244 Winning Centre Private Mail Bag 99914 100, Sir Chittainpalam A BANGLADESH 2nd Floor New Market GardinerMawatha Micro Industries Development Central Hong Kong Auckland Colombo 2 Assistance Society (MIDAS) HouseS, Road 16 INDIA NIGERIA SWEDEN Dhanmondi R/Area Allied Publishers Private Ltd. University Press Limited Forsisn titles. Dhaka 1209 751 Mount Road Three CrowrLs Building Jericho Fritzes Fackboksforget Madras - 600 002 Private Mail Bag 5095 Regeringsgatan 12, Box 16356 Bmnch ozis: Ibadan S-103 27 Stockholm 156, Nur Ahmed Sarak Branch orius: Chittagong 4000 15 J.N. Heredia Marg NORWAY For sucr'te ordeos: Ballard Estate Narvesen Information Center Wennergren-Williams AB 76, K.DA Avenue Bombay - 400 038 Book Department Box 30004 Kulna 9100 P.O. Box 6125 Etterstad S-104 25 StDckholm 13/14 Asaf Ali Road N0602 Oslo 6 BELGIUM New Delhi - 110 002 SWIERLAND Jean De Lannoy PAKISTAN For single tilies: Av. du Roi 202 . 17 Chittaranjan Avenue Mirza Book Agency Librairie Payot 1060 Brussels Calcutta - 700 072 65, Shahrah-e-Quaid-e-Azam 1, rue de Bourg P.O. Box No. 729 CH 1002 Lausanne CANADA Jayadeva Hostel Building Lahore 54000 Le Diffuseur 5th Main Road Gandhinagar For subscritiow orders: CP. 85,1501 Brue Ampere Bangalore - 560 009 PERU Librairie Payot Boucherville, Qu6bec Editorial Desarrollo SA Service des Abonnements J4B5E6 3-5-1129 Kachiguda Cross Road Apartado 3824 Cas postale3312 Hyderabad - 500 027 Lima 1 CH 1002 Lausanne CHINA China Financial & Economic Prarthana Flats, 2nd Floor PHIUPPINES TANZANIA Publishing House Near Thakore Baug, Navrangpura International Book Center Oxford University Press 8, Da Fo Si Dong Jie Ahmedabad - 380 009 Fifth Floor, Filipinas Ufe Building P.O. Box 5299 Beijing Ayala Avenue, Makati Maktaba Road Patiala House Metro Manila Dar es Salaam COLOMBIA 16-A Ashok Marg Infoernace Ltda. Lucknow -226001 POLAND THAILAND Apartado Aereo 34270 International Publishing Service Central Departnent Store Bogota D.E Central Bazaar Road Ul. Piekna 31/37 306 Silom Road 60 Baj1 Nagar 00-677Warzawa Bangkok COTE D'IVOIRE ' Nagpur440010 Centre d'Edition et de Difftsion For subscrtpion orders: TRINIDAD & TOBAGO, ANIIGUA Africaines (CEDA) INDONESIA IFS Journals BARBUDA, BARBADOS, 04 B.P. 541 Pt. Indira Limited Ul. Okrezna 3 DOMINICA, GRENADA, GUYANA, Abidjan 04 Plateau JI. Sam Ratulangi 37 02-916 Warmzawa JAMAICA, MONTSERRAT, ST. P.O. Box 181 KITTS & NEVIS, ST. LUCIA, CYPRUS Jakarta Pusat PORTUGAL ST. VINCENT & GRENADINES Cyprus College Bookstore Livrria Portugal Systematics Studies Unit 6, Diogenes Street, Engomi ISRAEL Rua Do Carmo 70-74 #9 Watts Street P.O. Box 2006 Yozmot Uterature Ltd. 1200 Lisbon Curepe Nicosia P.O. Box 56055 Trinidad, West Indies Tel Aviv 61560 SAUDI ARABIA, QATAR DENMARK Israel Jarir Book Store TURKEY SamfundsLitteratur P.O. Box 3196 Infotel Rosenoerns A11 11 ITALY Riyadh 11471 Narlabah$e Sok No. 15 DK-1970 Frederiksberg C Licosa Commissionaria Sansoni SPA Cagaloglu Via Duca Di Calabria, 1/1 SINGAPORE, TAIWAN, bstanbul DOMINICAN REPUBUC Casella Postale 552 MYANMAR,BRUNEI Editors Taller, C por A 50125 Firenze Information Publications UNITED KINGDOM Restauracidn e Isabel la Catelica 309 Private, Ltd. Microinfo Ltd. Apartado de Correos 2190 Z-1 JAPAN Golden Wheel Building P.O. Box 3 Santo Domingo Eastern Book Service 41, Kallang Pudding, #04-03 Alton, Hampshire GU34 2PG Hongo 3-Chome, Bunkyo-ku 113 Singapore 1334 England EGYPT, ARAB REPUBL1C OF Tokyo Al Ahram SOUTH AFRICA, BOTSWANA VENEZUELA Al Galaa Street KENYA Foraingletilles: Libreriadel Este Cairo Africa Book Service aLA.) Ltd. Oxford University Pres Aptdo. 60.337 Quaran House, Mfangano Steet Southern Africa Caracas 1060-A The Middle East Observer P.O. Box 45245 P.O. Box 1141 41, Sherif Street Nairobi Cape Town 8000 Cairo KOREA, REPUBLIC OF Pan Korea Book Corporation P.O. Box 101, Kwangwhamun Seoul The World Bank Headquarters European Office Tokyo Office U 1818 H Street, N.W. 66, avenue d'1ena Kokusai Building Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (202) 477-1234 Telephone: (1) 40.69.30.00 Facsimile: (202) 477-6391 Facsirnile: (1) 40.69.30.66 Telephone: (3) 3214-5001 Telex: wui 64145 woRLDBANK Telex: 640651 Facsimile: (3) 3214-3657 RCA 248423 WORLDBK Telex: 26838 Cable Address: wTBAFRAD WASHINGTIONDC Cover design by Walton Rosenquist ISBN 0-8213-2278-8