British Academy: The UK's National Academy for the Humanities and Social Sciences
Partha Dasgupta: 'The Economics of the Environment'
Copyright © The British Academy, 1996
Printed in Proceedings of the British Academy, Volume 90, pp. 165-221
Acknowledgements
I am grateful to Kenneth Arrow, Scott Barrett, John Dixon, Paul Ehrlich, Carl Folke, Frank Hahn, Geoffrey Heal, C.S. Holling, Bengt-Owe Jansson, Simon Levin, Mohan Munasinghe, Charles Perrings, Jonathan Roughgarden, Ismail Serageldin, Robert Solow, David Starrett, Andrew Steer and, in particular, Karl-Göran Mäler, discussions with whom over the past many years have improved my understanding of the subject matter of this lecture.
Go to References . . . . Return to Contents
Notes
1. This etymology of produced goods and services does not yield a "resource theory of value". Like Marx's labour theory of value, any such theory would run into ground. One reason is that there are many natural resources, not one; and this alone would make the putative theory incoherent. Koopmans (1957) contains a simple proof of why.
2. See May (1972) and May and MacArthur (1972).
3. The contrast is illusory, as will become apparent below. That's why one can belong to both tribes with ease.
4. Ehrlich, Ehrlich and Holdren (1977) remains the outstanding treatise on both population and ecosystem ecology.
5. The natural capital-base includes in addition minerals and ores underground.
6. Minerals and fossil fuels aren't renewable. For reasons of space, I will ignore non- renewable resources in this lecture. For an account of what economics looks like when we include exhaustible resources in the production process, see Dasgupta and Heal (1979), Hartwick and Olewiler (1986), and Tietenberg (1988). For an account of both the theory and the historical role of the substitution of new energy resources for old, see Dasgupta (1989).
7. For a formal demonstration of this, see Dasgupta (1982).
8. By the same token, it has proved all too congenial for ecologists to regard the human presence as an inessential component of the ecological landscape. This has enabled them to ignore the character of human decisions and, so, of economics. Thus, ecologists in great part continue to think that environmental degradation resulting from increased human encroachment on ecosystems can be stemmed effectively by centralised command-and-control modes of operation (see below in the text). For further discussion of the interface of economic and ecological concerns, see Dasgupta and Ehrlich (1996).
9. Problems arising from an absence of forward markets for the distant future are no doubt ameliorated by the fact that we care about our children's well-being and know that they in turn will care for theirs, and so on, down the generations. This means, by recursion, that even if we don't care directly for the well-being of our distant descendants, we do care for them indirectly. Arrow et al. (1995a) contains a succinct account of these considerations.
10. This example is taken from Dasgupta (1990). Chichilnisky (1994) provides an extended discussion of it.
11. But see Hodgson and Dixon (1992) for an attempt at such an estimation for the Bacuit Bay and the El Nido watershed on Palawan, in the Phillipines. The cause of damages (to tourism and fisheries) was due to logging in the uplands. The authors' computations were incomplete, but such as they were, the analysis did point to the desirability of a reduction on logging.
12. And they are so left under the hundred-year old water laws in South Africa, where small groups of upstream farmers enjoy ownership rights over the water that flows through their lands. See Koch (1996).
13. Rulers had control over such resources in many early societies. But that was not the same as private property rights. Rulers were obliged to make them available to the ruled. Indeed, one of the assumed duties of rulers was to expand such resource bases.
14. It should be noted that a resource being common property doesn't mean that people have free access to it. Often, only those households with a historical right of access are permitted by the community to avail themselves of local common-property resources (see below in the text).
15. Not everyone writing on the subject has misread the literature. For illuminating accounts of the way communities have jointly controlled common-property resources, see Feeny et al. (1990), Oström (1990), and Stevenson (1991). Seabright (1993) contains a good theoretical discussion of the problems; as does Putnam (1993), who also provides empirical substance to the notion of "social capital" as an engine of local collective action (see below in the text).
16. The most complete account I have read of the centrality of local forest products in the lives of the rural poor is Falconer and Arnold (1989) and Falconer (1990) on Central and West Africa. The importance of common-property resources for women's well-being in historical times has been stressed by Humphries (1990) in her work on 18th century rural England. The parallels with modern-day poor societies are remarkable.
17. Provided people are sufficiently far-sighted, norms of behaviour that sustain co- operation can be shown to be self-enforcing in stationary environments. See Section 12 for further discussion.
18. This section is taken from Arrow et al. (1995b).
19. Whether the proportion of expenditure devoted to environmental amenities increases with rising income is an empirical matter, and little is known. The one study I have seen on this question, namely Kanninen and Kriström (1993), suggests otherwise: the proportion of expenditure devoted to amenities decreases with rising income!
20. The General Agreement on Tariffs and Trade and the North Atlantic Free Trade Agreement, respectively.
21. Unless it is accompanied by judicious environmental policy, expansion of international trade should be expected to result in an increased stress on the global commons. Copeland and Taylor (1995) provide a formal analysis of the pathway through which this occurs.
22. Low (1992) contains discussions of these matters.
23. The accounting price of a resource (whether or not it is an environmental resource) is the increase in the maximum value of social well-being if a unit more of the resource were made available costlessly. An accounting price of a commodity is, thereby, the difference between its market prices and the tax (or subsidy) that ought to be imposed on it. Dasgupta, Marglin and Sen (1972) and Little and Mirrlees (1974) offer procedures for estimating accounting prices. Neither book, however, has anything to say about environmental resources.
24. See, for example, Brown and McGuire (1967) for irrigation water; Clark (1976) and Cooper (1977) for fisheries; Magrath and Arens (1989) and Repetto et al. (1989) for soil fertility; Newcombe (1984) and Anderson (1987) for forestry; and Solorzano et al. (1991) for the latter three. Dixon and Hufschmidt (1986) and Dixon et al. (1988) are excellent sets of case-studies on these matters.
25. A second approach to the estimation of accounting prices of environmental resources is based on contingent valuation methods (or CVMs). They involve asking concerned individuals to reveal their valuation of hypothetical changes in the flow of environmental services. CVMs are useful in the case of amenities, and their applications have so far been confined to advanced industrial countries. As I'm not focusing on amenities in this lecture, there isn't any point in developing the idea underlying CVMs any further here. The most complete acccount to date on CVMs is Mitchell and Carson (1989). See also the report on the NOAA Panel on Contingent Valuation (co-chaired by K.J. Arrow and R.M. Solow) in the Federal Register, 58(10), 15 January 1993.
26. During the moratorium the whale population grows at the fastest possible rate. In his numerical computations, the commerically most-profitable duration of the moratorium was found to be some 10–15 years.
27. The pioneering works on quasi-option values are Arrow and Fisher (1974) and Henry (1974).
28. For a formal account of this, see Dasgupta and Heal (1979), Dasgupta and Mäler (1991, 1995), and Mäler (1991). Lutz (1993) contains a collection of articles that explore the practicality of moving to a system of national accounts that includes the environmental resource-base.
29. All values are assumed to be measured in terms of consumption. This involves no loss of generality, since all remaining objects that help realise social well-being (including distributional considerations) can in turn be valued in terms of consumption (Dasgupta, 1993). Note also that, in an open economy, the value of net exports ought to be deducted from the expression for NNP in the text (Sefton and Weale, 1993). Furthermore, the expression is correct only if labour is supplied inelastically (in this case it is a matter of indifference whether or not we include the wage bill). However, if the supply of labour is responsive to wages, the wage bill should be deducted from the expression (Nordhaus and Tobin, 1972).
By the value of net "investment" in the expression in the text, I mean the value of net changes in capital assets, not changes in the value of these assets. This means that anticipated capital gains (or losses) should not be included in NNP. As an example, the value of the net decrease in the stock of oil and natural gas (net of new discoveries, that is) ought to be deducted from GNP when NNP is estimated.
Finally, it has been argued by Putnam (1993) that, in addition to manufactured, environmental, and human capital, "social" capital (involving, among other things, trust and interpersonal networks) matters in the production of goods and services. Assuming that a suitable index of social capital were in hand, the expression for NNP in the text would include net investment in social capital. Answer to the question how we should estimate NNP should not be a matter of opinion today, it is a matter of fact; the problem isn't that we don't know what items NNP should ideally contain, rather, it is that we don't have adequate estimates of various accounting prices.
30. The existence of thresholds means that an ecosystem can flip to a quite different state in a short space of time when subjected to stress. Formally, and more generally, an exclusive reliance on accounting prices is justified only if production technologies are convex. Threshold effects are a prime example of non-convexities. Key articles on this matter are Baumol and Bradford (1972) and Starrett (1972).
31. See Mäler (1990) for a more detailed discussion of these issues.
32. How a national government allocates the nation's rights among agencies within the country is a different matter.
33. Barrett (1990, 1994) contain fine discussions of these problems. The latter, in particular, identifies reasons why agreement on the use of CFCs has been easier to reach than on carbon emissions.
34. See also Gadgil and Guha (1992) for a narrative on India's ecological history as seen from this perspective.
35. Formally, it is a Nash equilibrium.
36. See Kreps and Wilson (1982), Milgrom and Roberts (1982), Kreps et al. (1982), and Benoit and Krishna (1985) for demonstrations that co-operative behaviour is possible even when people know that the interactions will be for a (large) finite number of periods. For a non-technical discussion of the force of the assumption of common knowledge, see Binmore and Dasgupta (1986) and Aumann (1987).
37. Each of these qualifications can be relaxed. See Radner (1981) for weakening the first qualification, and Sabourian (1988) for relaxing the second.
Go to References
Return to Contents