Equity and Efficiency in Environmental Markets: Global Trade in Co2 Emissions

We study the connection between property rights and Pareto e¢ciency in a competitive market for emission permits when the level of emissions a¤ects utility. Of the in...nitely many ways of distributing a given total of permits between a ...xed number of parties, only a ...nite number can lead to Pareto e¢ciency. We can therefore evaluate alternative distributions of property rights on the use of the planet’s atmosphere not only in terms of their equity, but also in terms of their e¢ciency. The traditional orthogonality of equity and e¢ciency, valid for markets with private goods, does not hold here. This has important implications for arguments about the initial international distribution of entitlements to produce carbon dioxide, and for equity in environmental markets. In a North-South world we show how a redistribution of property rights in favor of the South can make both regions better o¤. We explain the connection between our results and the properties of Lindahl equilibria. 1This paper is forthcoming as Chapter 3 of Environmental Markets, edited by G. Chichilnisky and G.M. Heal, Columbia University Press, 1999. 2UNESCO Chair of Mathematics and Economics, Program on Information and Resources, Columbia University. Support from NSF Grant SE 92 16028 is acknowledged. 3Garrett Professor of Public Policy and Corporate Responsibility, Program on Information and Resources and Graduate School of Business, Columbia University. 4Professor of Economics and Chair, Department of Economics, Stanford University. The authors acknowledge valuable comments from Partha Dasgupta, Duncan Foley, Yun Lin, Karl Gn”{o}ran Mn”{a}ler, Paul Milgrom. 2 Chichilnisky et al 3.1 Equity, e¢ciency and CO2 abatement This paper addresses a topical issue: the creation of a global market for CO2 emission permits.5 The recent adoption of an ambitious target for CO2 emission has focussed attention on policy instruments for achieving this goal. In addition, increasing awareness of the economic burden of environmental protection has produced an interest in market-based policy instruments which can minimize detailed government intervention. As a result, markets for emission rights are today the approach of choice of the US administration.6 We show that a market for emission permits has an important characteristic not previously noted, a characteristic which has signi...cant economic and political implications. When the level of emissions a¤ects utilities, there is an unexpected link between equity and e¢ciency: the initial distribution of property rights or emission permits determines whether or not a competitive global CO2 permit market will operate e¢ciently.7 Prior to now, it has been generally assumed that the manner in which emission permits are initially distributed will not a¤ect the e¢ciency of the market.8 We show below that of all the many possible ways of distributing a given total of emission rights, very few are compatible with e¢cient markets. In this case equity and e¢ciency are not orthogonal, as in the ...rst and second theorems of welfare economics for standard competitive markets. How does this happen? The key to this result is the fact that the atmospheric concentration of CO2 is a privately produced public good, a¤ecting the utility levels of 5The atmospheric concentration of carbon dioxide has become a matter of international concern. It is generally recognized that it has the capacity to change the global climate in ways which are potentially harmful and irreversible. For a review see Chichilnisky and Heal [3] and Chichilnisky et all. [9]. Consequently countries at the 1992 “Earth Summit” in Rio de Janeiro agreed to cut back CO2 emissions to their 1990 levels by the end of the century. This policy could easily cost several percent of GNP (see Weyant [28]). In conformity with the conclusions of the “Earth Summit”, the US administration has recently made a tentative move in the direction of capping CO2 emissions in industrial countries. 6According to a statement by Tim Wirth, US Assistant Secretary of State for Global A¤airs, at the 1996 Berlin Conference of the Parties of the Framework Convention on Climate Change. 7The term “e¢ciently” here is used in the standard economic sense of “so as to attain Pareto e¢ciency”. 8 It will of course a¤ect the distribution of income resulting from the operation of the market. This is the original Coase position [10]: that whatever the initial distribution of permits, trading rights can bring about a Pareto e¢cient allocation of resources. In fact a stronger claim is sometimes made: that the equilibrium allocation of resources is not a¤ected by the initial distribution of permits. Clearly the conditions for this stronger claim to be true are very restrictive indeed—a total absence of income e¤ects, see Milgrom and Roberts, chapter 2 [23]. 3. Equity and E¢ciency in Environmental Markets: 3 all people. The reason is that carbon dioxide mixes thoroughly in the atmosphere, leading to a uniform concentration over the globe. So we have a global public good. People or regions cannot choose their concentration levels independently. The concentration is, however, determined by every individual who runs a car or a heating furnace, and by every ...rm operating transportation or burning fuel in any other way.9 Therefore we have a privately produced public good. The fact that CO2 concentration is a privately produced public good a¤ecting the welfare levels of individuals leads to the equity-e¢ciency interaction. As noted, everyone has de facto to consume the same CO2 concentration: for e¢ciency, this common level must be what they demand given prices and their incomes. In summary: for agents’ to demand freely the same amounts of CO2 at an equilibrium requires a particular choice of the distribution of income. These points were made in Chichilnisky [2] and Chichilnisky and Heal [4],10 where this simple observation was shown to have other far-reaching consequences. In particular, these papers establish that the equalization of marginal abatement costs across countries is neither su¢cient nor necessary for Pareto e¢ciency: Pareto e¢cient allocations may have di¤erent marginal costs. Here we show that this line of argument, when developed further, implies that e¢ciency and distribution cannot be separated in environmental markets. E¢ciency requires an appropriate distribution of property rights. The fact that many distributions of property rights lead to ine¢cient outcomes allows us to construct an example of a two-region world in which a transfer of property rights from the North to the South, accompanied by a decrease in the total of emission permits, leaves both regions better o¤. Finally we investigate the extent to which an equilibrium concept related to that of Lindahl is the appropriate concept in permit markets. There is a simple reason why this might be so: a Lindahl equilibrium is the only market equilibrium known to lead to Pareto e¢ciency with public goods.11 As a permit market is a market which determines the production of public goods, we might therefore expect that e¢ciency would require the key feature of a Lindahl equilibrium, namely a multiplicity of prices, in fact one price per pair of traders. In a Lindahl equilibrium, each producer of a public good is paid for her production by each consumer, and the per unit payment typically varies from consumer to consumer. So relative to the framework of a Lindahl equilibrium, a permit market as formalized here is an “incomplete market” because everyone pays the same price for the permits. This can be interpreted as assuming that the “individualized” 9Carbon dioxide, a public bad, is a by-product of the consumption and production of private goods. 10There is also an early discussion of closely-related issues in La¤ont [20] and in Eyckmans et all [14]. 11See Foley [15]. 4 Chichilnisky et al markets between buyers and sellers are missing. Our main result shows that in a certain sense, it is possible to compensate for the absence of markets by reallocating property rights in tradeable permits.12 3.2 E¢ciency and international emissions Following the model set out in Chichilnisky [2] and developed further in Chichilnisky and Heal [4], we consider a world economy with I regions, I ̧ 2; indexed by i = 1; ::; I: Each region has a utility function ui which depends on its consumption of a vector of private goods ci = (ci;1; ci;2; :::; ci;M ) where M is the number of private goods (indexed by m), and also on the quality of the world’s atmosphere, a; which is a public good.13 The quality of the atmosphere a can be thought of as a measure of abatement. It could be measured by for example the reciprocal or the negative of the concentration of CO2: the more abatement there is, the lower is this concentration. The concentration of CO2 is “produced” by emissions of carbon, which are positively associated with the levels of production of private goods. Let yi be a vector in R giving the production levels of the M private goods in country i. Then