قدرت بازار در بازار انتشار قابل تجارت : بستر آزمایشی آزمایشگاهی برای انتشار تجارت در بندر خلیج فیلیپ ، ویکتوریا

In theory, competitive emission permit markets minimize total abatement costs. Permit markets are often imperfectly competitive, however, and may be thin and dominated by large firms. The dominant firm(s) could exercise market power and increase other firms’ costs of pollution control, while reducing their own emission control costs. This paper reports a testbed laboratory experiment to examine whether a dominant firm can exercise market power in a permit market organized using the double auction trading institution. Our parameters approximate the abatement costs of sources in a proposed tradable emissions market for the reduction of nitrogen in the Port Phillip Watershed in Victoria, Australia. We vary across treatments the initial allocation of permits to sources, so that in one treatment the seller of permits is a monopolist and in another treatment the market is duopolistic. We also vary the information that subjects have about the number and abatement costs of their competitors. We find that prices and seller profits are higher and efficiency is lower on average in the monopoly sessions compared with the duopoly sessions, but the differences are not substantial and are not statistically significant due to pronounced variation across sessions. Moreover, prices, profits and transaction volumes are usually much closer to the competitive equilibrium (CE) than the monopoly equilibrium.

For almost four decades economists have argued that tradable emissions permits can allocate emissions control responsibility, under conditions of asymmetric information and heterogeneous abatement costs, at least cost, for any given emissions ceiling (Dales, 1968, Montgomery, 1972, Tietenberg, 1985 and Baumol and Oates, 1988). Policy makers have slowly begun employing trading schemes for environmental management, predominantly for point source pollution. One of the first and most ambitious schemes was the United States Federal SO2 Trading Program, (Title IV 1990 Clean Air Act Amendments), which regulates sulfur dioxide emissions that cause acid rain from electricity utilities across the US. Another example is the Regional Clean Air Incentives Market, established by the South Coast Air Quality Management District in Southern California, which regulates nitrogen and sulfur oxide emissions from 400 industrial sources in the South Coast air basin. Examples of point source water quality trading schemes include the Cherry Creek Basin Trading Program for phosphorous in Colorado (USA), and the Hunter River Salinity Trading Scheme in Australia, which regulates saline water exports from point sources in the Hunter River New South Wales, Australia. These programs have reported considerable cost savings. For example, Schmalensee et al. (1998) estimate that the US Federal SO2 Trading Program reduced abatement cost by 25–34% as compared with a scheme with the same allocations of allowances but no ability to transfer allowances. Theoretical models of allowance trading often assume some level of competition (often perfect), such that no firm(s) can exercise power on the selling or buying side of the market. This may be appropriate for large markets like the US SO2 market, but other permit markets are often concentrated, and therefore, only imperfectly competitive. The dominant firm(s) could exercise market power and increase other firms’ cost of pollution control, while reducing their own emission control costs (Hahn and Noll, 1982, Hahn, 1984, Newberry, 1990 and Vickers and Yarrow, 1991). Some empirical studies, however, find that the extent of market power would need to be great (up to 90% of the potential trading volume) for a firm or group of colluding firms to manipulate control costs and even then, the savings from the permit market are not eroded to a significant extent (Maloney and Yandle, 1984). Port Phillip Bay in southern Victoria, Australia is one such case where an emissions permit market would be dominated by one large firm. In 1996 the Port Phillip Bay Environmental study was conducted to investigate sustainable use and management of the Bay (CSIRO, 1996). This study identified nitrogen as the nutrient that limits algal growth in the Bay. It recommended that nitrogen emissions to the Bay should be reduced, with a specific target set for 2006. The Department of Natural Resources and Environment Victoria undertook a feasibility study into nitrogen emissions trading in Port Phillip Bay to reduce and maintain emissions from point sources (DNRE, 2002). One of the issues that needed to be explored was the likelihood of the dominant firm exercising market power. The testbed laboratory permit market reported in this paper was conducted to specifically test if the dominant firm could exercise power in this type of market. In Port Phillip Bay there is one large emitter of nitrogen—a sewerage treatment plant managed by the Melbourne Water Corporation; three small sewerage treatment plants—City West Water Authority, Western Water Authority, Yarra Valley Water Authority; and two land based fish farms. The Melbourne Water Corporation's emissions account for approximately 94% of nitrogen emissions from point sources to the Bay. City West Water accounts for approximately 2%, Western Water 0.95%, Yarra Valley Water 1.9% and the two fish farms 0.4% each (Argent and Mitchell, 1998). In order to testbed the permit market, abatement cost and scope for each firm in the Bay was estimated in consultation with the Melbourne Water Corporation, the Environmental Protection Authority Victoria, fish farmers and the Marine and Freshwater Research Institute Victoria (see Table 2 below). The firms differ in their cost and potential quantity of abatement. The Water Authorities have high fixed costs for the first unit of nitrogen abated. The Melbourne Water Corporation then faces low constant marginal costs and can abate a greater quantity of nitrogen than the other water authorities. The fish farms do not have up front abatement costs; however, the quantity of nitrogen they can abate is less than the water authorities. A primary focus of this laboratory study is how the potential for market power could differ depending on the initial endowment of permits. Hahn (1984) finds that the initial allocation could have an impact on both the post trade allocation of permits and the permit price in the presence of market power. This finding is contrary to the results for competitive markets, where both the price and the ultimate allocation of permits would be independent of the initial allocation. The intuition for this is the following: Whenever a single, price-setting firm receives an initial allocation that is either higher than or lower than its cost effective allocation, an incentive for trading would be created. When a price-setting firm receives an initial allocation containing fewer permits than its cost-effective allocation, it would exercise market power on the buyer's side. Similarly, if it received more permits than needed given its relative abatement costs, it could exercise power on the seller's side of the market. The further the initial allocation diverges from the cost-effective allocation, the greater is the potential for the price setter to exercise power over the market. In the current application, if permits are endowed symmetrically so that all of the small sources receive an allocation that corresponds to the same fraction of their current estimated emissions, then under some conditions Melbourne Water would be the monopoly seller of permits in equilibrium. But an asymmetric allocation of permits could provide more permits to some smaller firms, and in some conditions this could lead to an additional source of permits in the market. The asymmetric allocation examined in the experiment would result in a second seller in equilibrium, and therefore, a duopoly market structure. We find that prices and seller profits are higher and efficiency is lower on average in the monopoly sessions compared with the duopoly sessions, but the differences are not substantial and are not statistically significant due to pronounced variation across sessions. Moreover, prices, profits and transaction volumes are usually much closer to the competitive equilibrium (CE) than the monopoly equilibrium. This suggests that at least for the double auction trading institution considered here, market power could have a smaller impact on the market performance than indicated by standard models of imperfect competition.

The testbed laboratory permit market reported in this study was conducted to inform policy makers about how effectively a dominant firm in an emissions permit market could exercise market power, when trading was organized using the continuous double auction trading institution. The laboratory market approximates the economic and environmental conditions present in the Port Phillip Watershed, Victoria Australia. A primary focus of the experiment was to determine how the potential for market power could differ depending on the initial allocation of permits. In the Monopoly/Asymmetric Endowment treatment, all small firms were allocated permits equal to 50% of their current emissions. In the Duopoly/Symmetric Endowment treatment one small firm with low abatement costs was allocated permits greater than 50% of its current emissions, so that this small low cost firm could become an active seller and introduce some competition into the selling side of the market. Another small firm was allocated permits less than 50% of its current emissions and received a monetary subsidy instead of permits to compensate. This is akin, for example, to an environmental agency subsidizing the first few units of abatement for the disadvantaged firm in order to create more competition in the market. It is no different from allocating permits free of charge for those units of emissions. A secondary focus of this laboratory study was to examine how information can influence the ability of the dominant firm to exercise market power. In the ‘Less Information’ treatment all firms know only their own abatement costs and do not have any information about the abatement costs of other firms. In the ‘More Information’ treatment the dominant firm (the Melbourne Water Corporation) receives information on the distribution of abatement costs of other firms. When the dominant firm holds information about the abatement costs of other firms it may strengthen this firm's resolve to withstand the tacit collusion (withholding of permit demand) by the smaller firms. The ‘large’ firm(s) were able to raise transaction prices more in the Monopoly/Symmetric Endowment treatment than in the Duopoly/Asymmetric Endowment treatment. The difference, however, was not statistically significant except for opening prices, due to substantial price variance across experimental sessions, indicating that the market did not adjust to a unique equilibrium price. It is not uncommon to observe variance in prices in market power experiments as the ability to influence outcomes depends on one or two subjects in each experiment. Transaction prices did decline towards the CE prediction range and were significantly different from the predicted monopoly and collusive price in both treatments. The IMTE, a measure of firms’ ability to extract supercompetitive profits, also declined across periods, and was significantly less in both treatments than the predicted monopoly and collusive price level. Transaction volume starts below the CE level but increases towards the CE in later trading periods. Trading efficiency in both treatments quickly rises higher than the monopoly and collusive price equilibrium level, but after trading period 4 it increases very little. Nonparametric tests indicate that mean efficiency is not significantly different from the monopoly equilibrium prediction in the Monopoly/Asymmetric Endowment treatment, but the same tests indicate that mean trading efficiency significantly exceeds the monopoly and collusive equilibrium in many periods of the Duopoly/Asymmetric Endowment treatment. Overall, the results suggest that a monopoly supplier of permits may not be able to dominate a potential tradable emissions permit market in Port Phillip Bay, at least when trade is organized with double auction trading rules. Prices, however, could vary considerably—perhaps due to the large fixed capital abatement costs that the water authorities face. The non-convex feature of the cost function perhaps leads to a wider dispersion of prices and leads to prices higher than the competitive level in some sessions. These non-convexities are included in the experiment to reflect the cost conditions in the field. Although in most sessions the prices adjusted towards CE over time, this finding might be sensitive to the sophistication of the dominant firm traders. Nevertheless, at least in this laboratory testbed the initial allocation of permits does not significantly impact market performance, so policy makers may not need to vary initial allocation to introduce some competition into the market. It is important, of course, to assess the robustness of these conclusions. For example, the results may be influenced by the strong competitive tendencies of the double auction institution. It would be useful to compare a posted offer auction or bilateral bargaining trading institution with these double auction results to evaluate whether dominant sellers can better exercise market power in these alternative institutions. If so, this would suggest that policy should focus on encouraging (or sponsoring) competitive trading institutions rather than endowing permits to enhance competition. In further research it may be useful to investigate whether these conclusions continue to hold for more advanced levels of subject experience.8 The observed outcomes of market power experiments depend on one or two subjects in the experiment, therefore, the experience of subjects representing the dominant firm(s) could impact upon market outcomes. Harrison et al. (1989) in an experimental study of the effectiveness of monopoly regulation mechanisms, observed that experienced subjects, playing the role of monopsonists with simulated buyers, were more successful at earning monopoly profits than inexperienced subjects, and price and quantity variance was lower. Carlén (2003) argues that it is important that subjects who represent the dominant firm be confident with the role. Bohm and Carlén (1999) use Ph.D., students in Economics with no prior experimental experience except for one, the subject playing the role of the monopolist. In some asset market experiments, the variance of asset trading activity is greater with inexperienced subjects. Convergence to equilibrium is more rapid with experienced subjects, it is not, however, significantly different from inexperienced subject treatments (Forsythe et al., 1982 and Friedman et al., 1984). A number of studies do indicate that performance can vary with proxies for the aptitude of participants (Davis and Holt, 1993, page 17). In addition to experienced or sophisticated student subjects, it would also be useful to involve actual decision makers (for example, the managers of the firms participating in the permit market) in these permit trading experiments. The behavior of decision makers from naturally occurring markets has been examined in the experimental literature and typically their behavior has not differed from that exhibited by more standard and less costly student subject pools (Davis and Holt, 1993). However, before these testbed experiment results are applied directly in the field, further research should examine the impact of using more sophisticated participants.