طراحی بهینه یک برنامه انتشار تجارت
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|18948||2000||19 صفحه PDF||سفارش دهید||8094 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Public Economics, Volume 75, Issue 2, February 2000, Pages 273–291
This paper studies a phase-in emissions trading program with voluntary opt-in possibilities for non-affected firms and derives optimal permits allocations to affected and opt-in firms when the environmental regulator has incomplete information on individual unrestricted emissions and control costs. The regulator faces a trade-off between production efficiency (minimization of control costs) and information rent extraction (reduction of excess permits allocated to opt-in firms). The first-best equilibrium can be attained if the regulator can freely allocate permits to affected and opt-in firms; otherwise a second-best equilibrium is implemented. The latter is sensitive to uncertainty in control costs and benefits.
In recent years we have witnessed a significant increase in the attention given by environmental policy makers to market-based instruments, particularly tradeable emissions permits.1 The sulfur dioxide (SO2) emissions trading program under Title IV of the 1990 Clean Air Act Amendments is the largest experience with the use of tradeable permits ever implemented. Furthermore, it is the first emissions trading program to include a voluntary participation provision — the Substitution provision — in its first phase of implementation.2 Electric utility units not affected by the emissions limits of Phase 1 can voluntarily opt-in and receive tradeable permits. 3 Although the Substitution provision was primarily designed to allow those non-affected electric units with low control cost to opt in, Montero (1999) explains that a large number of non-affected units opted in because their unrestricted or counterfactual emissions (i.e. emissions that would have been observed in the absence of regulation) were below their permit allocation. In other words, they had received excess permits. While shifting reduction from high-cost affected units to low-cost non-affected units reduces aggregate compliance costs, excess permits may lead to social losses from higher emissions than had the voluntary provision not been implemented. As with any other regulatory practice, the optimal design of a phase-in emissions trading program with opt-in possibilities for non-affected firms is subject to an asymmetric information problem in that the regulator has imperfect information on individual unrestricted emissions and control costs. Furthermore, if we believe that either for practical or political reasons, phase-in or less than fully comprehensive trading systems are likely to be the rule rather than the exception in future environmental policy, the same sort of issues observed in the SO2 emissions trading program are likely to arise in attempts to implement tradeable permit schemes in practice. In fact, a salient example is provided by current emissions trading proposals in dealing with global warming that call for early carbon dioxide restrictions on OECD and few other countries with voluntarily opt-in possibilities with the rest of the world (see Tietenberg and Victor (1994), and The Kyoto Protocol to the Convention on Climate Change).4 In this paper I study the welfare implications and implications for instrument design of this particular asymmetric information problem. As shown below, in a world with perfect information and no transaction costs, a regulator would issue permits to opt-in firms equal to their unrestricted emissions in each period. In practice, however, the environmental regulator cannot anticipate the level of unrestricted emissions. Yet, s/he must establish a permit allocation rule in advance that cannot be changed easily even if new information would suggest so.5 Thus, we must recognize that a voluntary program is subject to an asymmetric information or adverse selection problem in that firms reducing emissions below their permit allocation independent of the environmental legislation will receive excess permits.6 In deciding how to set the allocation rules for affected and opt-in firms, the regulator faces the classical trade-off in regulatory economics between production efficiency (minimization of aggregate control costs) and information rent extraction (reduction of excess permits). For instance, a too restrictive allocation rule for opt-in firms may inefficiently leave too many low-cost firms outside the program. The regulator’s problem reduces to that of finding the permit allocations for affected and opt-in firms that maximizes social welfare under conditions of imperfect information, distributional concerns and cost and benefit uncertainty. To my knowledge there is no paper addressing the issue of adverse selection and optimal design of a phase-in program. Montero (1999) identifies and tests the issue empirically but does not derive optimal permit allocations. In this paper, I study the optimal design problem following literature on the economics of regulation (Laffont and Tirole, 1993), instrument choice under uncertainty (Weitzman, 1974; Stavins, 1996), and optimal environmental regulation under imperfect information (Kwerel, 1977 and Dasgupta et al., 1980; Spulber, 1988). One of the results of the paper is that if the regulator has two instruments — the permit allocation to originally affected firms and to opt-in firms — in the absence of income effects and distributional concerns, s/he can achieve the first-best outcome. This result is similar to those of Kwerel, 1977 and Dasgupta et al., 1980, and Spulber (1988) in that information asymmetries may not prevent the environmental regulator from achieving the social optimum. If the regulator, however, cannot make ‘permit transfers’ from affected to opt-in firms, so that s/he has only one instrument — the permit allocation to opt-in firms — s/he achieves a second-best outcome in which the opt-in allocation is lower than the first-best opt-in allocation to the point where gains from information rent extraction are just offset by the productive efficiency losses of leaving low-cost non-affected firms outside the program. I also find that the second-best result is sensitive to uncertainty in aggregate control costs and benefits. If benefit and cost uncertainties are correlated negatively or not at all, the regulator always benefits from setting the opt-in rule slightly above the ‘certain’ second-best allocation. The remainder of the paper is organized as follows. In Section 2, I present the model and explain the trade-off between production efficiency and information rent extraction. In Section 3, I derive the social optimum or first-best permit allocation rules for affected and opt-in firms when the regulator lives in a world of complete information and certainty. In Section 4, I derive the optimal design when the regulator has incomplete information on individual unrestricted emissions and marginal control costs, but still she has two instruments and there are no income effects from permit allocation transfers between affected and opt-in firms. Although the results in 3 and 4 can be demonstrated graphically (and are), I also develop analytical proofs to structure subsequent sections. In Section 5, I include distributional concerns and restrict the analysis of Section 4 to the optimal design when the regulator has only one instrument — the permit allocation to opt-in firms. In Section 6, I study the effect of benefit and cost uncertainty on instrument design. Concluding remarks are in Section 7.
نتیجه گیری انگلیسی
For either political or practical reasons, phase-in emissions trading programs that include opt-in provisions for non-affected firms are receiving significant attention. I have presented a theoretical analysis of the welfare implications and implications for instrument design when the phase-in trading program is implemented under conditions of imperfect information regarding individual unrestricted emissions (i.e. emissions that would have been observed in the absence of the regulation) and control costs, distributional concerns, and cost and benefit uncertainty. I have shown that the regulator faces a trade-off between production efficiency (control cost minimization) and information rent extraction (reduction of excess permits for opt-in firms). A regulator having two instruments — the permit allocation to originally affected firms and to voluntary opt-in firms — can, in the absence of income effects and distributional concerns, implement the first-best outcome. If the regulator cannot make permit transfers from affected to opt-in firms, so that she has only one instrument — the permit allocation to opt-in firms — she implements the second-best allocation to opt-in firms, which is lower than the first-best allocation to the point where gains from information rent extraction are just offset by the productive efficiency losses of leaving low-cost non-affected firms outside the program. Finally, I showed that the second-best result is sensitive to uncertainty in aggregate control costs and benefits. If benefit and cost uncertainties are correlated negatively or not at all, it always pays off for the regulator to set the new optimum allocation slightly above the ‘certain’ second-best allocation. The results of this paper can have important policy implications for the design of an emissions trading scheme to prevent global warming. Current proposals call for early carbon dioxide restrictions on OECD and few other countries with voluntarily participation possibilities from less developed countries. In deciding about the total number of permits (aOP) to be credited to an ‘opt-in project’, the results of this paper indicate that the imperfectly informed regulator would face the same trade-off between production efficiency and information rent extraction. I leave for future research the development of a dynamic setting to study this particular application in more detail.