قدرت بازار در بازار منفک شده عمده فروشی برق : مسائل مربوط به اندازه گیری و هزینه کاهش خطر

An analysis of three recently deregulated markets—California, PJM, and New York—finds that none of them can be regarded as highly competitive, contrary to what conventional measures of market power indicate. Auctions for generation are unlikely to be competitive and costly steps will be needed to mitigate market power, likely eroding any benefits from increased operating efficiency in deregulated markets. Thus, FERC and state legislators need to reexamine the desirability of deregulating the generation portion of the industry.

Deregulation of airlines, trucking, railroads, waterborne freight, petroleum, and natural gas provided large benefits to consumers and the economy.1 When providers had to satisfy consumers rather than regulators as their “customers,” they were forced to lower prices, attend to consumer demands, and speed the introduction of technological change. This deregulation movement began in earnest in 1978 and spread quickly in the next decade, producing large social benefits in transportation and energy. Although extremely successful during the first half of the century, by the mid-1960s electric utilities had to confront growing technical and business difficulties. In 1965, New York City and much of the East Coast experienced a power blackout. Despite the efforts to fix the system, partly through the formation of the North American Electric Reliability Council (NERC), the problems occurred again in the 1980s. The OPEC oil embargo of 1974 and subsequent price increases plunged the economy into recession, interrupting electricity demand growth and causing utilities to cancel orders for many generation units, including some under construction; the cancellation costs eventually raised consumer electricity bills. Consumer prices rose even more with the large cost overruns in constructing many nuclear plants. By 1990, the electricity industry had high prices, reliability problems, and seemed unable to handle the challenges it faced. The beneficial experience with deregulation in the transportation, oil, and natural gas markets led economists and policymakers to propose restructuring the electricity industry. Beginning with the passage of the 1978 Public Utilities Regulatory Policies Act (PURPA),2,3 less-regulated wholesale electricity markets were given a boost with the 1992 Electricity Policy Act (EPAct), which allowed non-load-serving entities to buy and sell bulk power. California and Pennsylvania competed to be the first states to move towards market-based ratemaking (rather than regulated ratemaking based on a “cost plus” approach). Many other states were not far behind. At first, only large industrial customers were allowed to sign contracts with non-utility generators (NUGs) for “direct access” service, but eventually both California and Pennsylvania allowed residential and commercial customers to select their electric service providers. The experience of the two leading states could not have been more different. After a transition period, in which utilities were allowed to pay off stranded costs, prices in Pennsylvania have fallen and consumers are benefiting.4 Despite some isolated incidents of individual generators attempting to exercise market power, the net benefits to consumers appear to have been positive.5 The first two years of California deregulation paralleled Pennsylvania’s experience before an extended wholesale price escalation. High prices prevailed for nearly a year before the Federal Energy Regulatory Commission (FERC) intervened to impose price caps, investigate generators who appeared to be withholding capacity, and threatened to disallow previous sales prices. By the end of the energy crisis, the State of California and the three California investor-owned utilities (IOUs) were in deficit by more than $30 billion. Several analysts saw California’s energy crisis as a “perfect storm” in which a confluence of forces converged: a long drought that cut the availability of inexpensive hydro power, a gas pipeline explosion that curtailed gas deliveries to turbine generators, and high prices for NOx, emissions allowances. Although a flawed market structure certainly received its share of the blame for California’s power woes, the conventional wisdom maintained that minor modifications to the market structure, together with a respite from the perfect storm, would produce a competitive electricity market that would serve consumers far better than the regulated system. In other words, this conventional view is that deregulation, even in California, is still in the public interest and that FERC should continue to press for deregulation and encourage or even force other states to establish competitive wholesale generation markets. California’s experiment with deregulation has produced a slew of analyses that demonstrate how prices in California have regularly exceeded the “competitive” level, even in times of surplus capacity.6 Recent studies accuse firms of manipulating the electricity and natural gas markets to raise prices. Oversight reviews of the California Power Exchange and Independent System Operator have also used a price-cost margin metric. These analyses demonstrate that market power was exercised in California’s wholesale energy auctions, increasing prices and profits. Non-competitive behavior presumably can occur only within the confines of a non-competitive market structure. Antitrust regulators have used measures of market concentration to assess the competitiveness of market structure; the most popular of these has been the Herfindahl-Hirschman Index (HHI), computed as the sum of the market shares of each firm squared. A purely monopolistic market (one firm with a market share of 100 percent) would have an HHI of 10,000. To assess the competitiveness of market behavior, economists and regulators examine price-cost margins; the Lerner Index, one measure of the price-cost markup quite common in the antitrust literature, is computed as market price minus cost divided by price.7,8 More recently, however, regulators in particular have abandoned the conventional approaches in favor of methods based on comparisons between the generation capacity of individual firms and the surplus of the system as a whole (the “supply margin”). FERC, for example, has begun to use a “pivotal supplier” analysis in reviews of market-based ratemaking, which evaluates the ability of a single generating company to disrupt the grid by withholding supplies from the market. A recent study of the proposed Western regional transmission organization (RTO) used seasonal HHIs to determine the dominance of individual electric utilities in particular service areas.9 Regardless of the actual measure of market power employed, these studies and many others have focused on the performance of individual markets, demonstrating the exercise of market power in some areas at some times. Conversely, the analysis presented here is one of market structure, and in the spirit of the HHI, examines the potential for one firm (or a group of firms) to raise the price of power above the competitive level.

The rush to deregulate electricity markets did not recognize the unique properties of electricity. FERC and the state PUCs saw the generation markets as competitive because they were using the conventional economic measure of market concentration. If, instead, the pivotal supplier concept illustrated here is used, California, PJM, and even New York have structures that indicate the potential for market power during the times of peak demand. The unique nature of electricity, requiring that supply and demand are always in balance, gives monopoly power in markets that ordinarily be seen as highly competitive. We show that for much of the year, three, two, or even a single firm have monopoly power and can raise price. We see three ways to mitigate this market power. One option would be expanding generation and transmission capacity so that the largest few suppliers no longer have monopoly power in peak periods. We estimate that this capacity expansion would raise generation costs by 0.3 to 1 cents per kWh in California. Whether these costs can be outpaced by coincident savings resulting from increased efficiency remains to be seen, though deregulation’s performance thus far has not been inspiring.24 A second way to mitigate the exercise of market power is through a forced generation divestiture, which would ensure that no firm controlled more than a few percentage points of capacity. This route would decrease the efficiency of generation, raising costs. It is further limited by the fact that some generation units are large. Not permitting the construction of large, efficient generation units would raise costs even more. Furthermore, in small states, the indivisibility problem of generating units is exacerbated, where the largest generating unit can easily represent on the order of 20 percent of total demand.25 A third way to mitigate market power is for FERC or state regulators to set price caps, as is currently the case in California. FERC knows, or can estimate closely, the marginal cost of generation for every unit. Rather than inviting bids from generators, FERC could assign each unit this marginal generation cost. In this world, the owner of each generator would simply indicate that a unit is ready for service, allowing FERC to build its supply schedule from the marginal generation costs of each unit. This is entirely feasible, but it is certainly not a competitive market. An even larger problem with price caps as envisioned by the FERC is that it could easily be challenged in court on the grounds that the government is taking property without due compensation. FERC’s price caps could only be made consistent with the Fourth Amendment if generation owners were able to earn a return based on both their fixed and variable costs. However, dispatching generation on the basis of cost is largely equivalent to the merit-order dispatch that prevailed under state regulation, with one notable exception. Under regulation, each generator was paid its average cost, not the market clearing cost. Thus, this plan would prove much more expensive than the former state regulation. None of the states that we examined have market structures that would be regarded as highly competitive. They range from highly non-competitive in California to reasonably competitive in New York State. We urge FERC and state legislatures to pause in their deregulation efforts. In our judgment, most regional markets will not be sufficiently competitive so that consumers will see net benefits. Trying to force competition in these states will most likely raise electricity prices, harming consumers. If FERC and individual states are intent on deregulating their electric power markets, they must realize that a pure supply-side solution will not achieve the desired result. In particular, if ISOs and RTOs insist on operating as if they faced a purely vertical demand curve, individual generators (or small groups) will find it easy to extract high prices. Some attempt must be made, through real-time pricing or other means, to introduce some elasticity into the demand curve. The doctrine of “obligation to serve” need not disappear entirely, but in a deregulated marketplace, prices faced by consumers need to reflect the choices they have made. 1 Robert Crandalland Jerry Ellig, Economic Deregulationand Customer Choice: Lessonsforthe Electricity Industry (Arlington, VA: Mercatus Center, 1997). 2 Deregulation in the wholesale energy market actually has its roots in the construction of the Pacific Interties in the 1940s and 1950s. The Interties allowed utilities in California and the Pacific Northwest to exchange surplus energy on a seasonal basis. These informal exchange agreements blossomed into North America’s first bulk power market, in which utilities throughout the Western Interconnect participated in exchanges or outright purchases and sales of wholesale electricity. The Western experiment was officially sanctioned by FERC in 1991 with the formation of the Western Systems Power Pool (WSPP). 3 The passage of PURPA opened the doors to non-utility generators (NUGs), which had previously been disallowed under the Public Utility Holding Companies Act (PUHCA). Most of the initial NUGs represented alternative or renewable energy, and the generation was largely under long-term contract with a load-serving utility. 4 The “stranded cost” allowances permitted utilities in California and Pennsylvania to pay off unprofitable investments (made under the era of regulated electricity) through surcharges imposed on ratepayers. 5 Data from the Energy Information Administration’s Electric Power Monthly suggests that retail rates in Pennsylvania are now below the national average. 6 The most noted of these analyses have been Severin Borenstein, James Bushnell, and Frank Wolak, Diagnosing Market Power in California’s Deregulated Wholesale Energy Market, University of California Energy Institute, POWER Working Paper PWP-064, 2000, and Paul Joskow and Edward Kahn, A Quantitative Analysis of Pricing Behavior in California’s Wholesale Electricity Market during Summer 2000, National Bureau of Economic Research, Working Paper 8157, 200l. A similar approach has also been taken by Robert McCullough, Price Spike Tsunami, Pub. Util. Fortnightly, Jan. 2001, at 22, in which the slope of California’s aggregate supply curve is estimated at a particular point, both before and after the onset of the power crisis. 7 Some have argued that price-cost margin tests for competitiveness are inappropriate for the electric power industry, because the use of marginal cost rather than variable cost (including capital cost) implies that, in equilibrium, no generator will enter the market. See Timothy Brennan, Checking for Market Power in Electricity: The Perils of Price-Cost Margins, available at http://www.ipa.org.au/pubs/Moranwebpapers/brennan.pdf, last updated April 23, 2002. Performance analyses using price-cost margins have also been questioned by William Hogan and Scott Harvey, On the Exercise of Market Power Through Strategic Withholding in California, Harvard Electricity Policy Group, Working Paper, 2001, available at http://www.whogan.com, last updated April 24, 200l. They argue that the true variable cost of generation has not been accurately measured. 8 Early analyses of price-cost markups in wholesale electric markets focused on duopoly pricing behavior in the British wholesale power pool. See, for example, Richard Green and David Newberry, Competition in the British Electricity Spot Market, J. Pol. Econ., Oct. 1992, 100 (5), at 929–953, as well as Frank Wolak and Robert Patrick, The Impact of Market Rules and Market Structure on the Price Determination Process in the England and Wales Electricity Market, University of California Energy Institute, POWER Working Paper PWP-047, 1997. 9 Tabors, Caramanis, and Associates, RTO West Filing Cost Study: Final Report Presented to RTO West Filing Utilities, 2002, available at http://www.tca-us.com/publications/pub2.html, last updated March 11, 2002. 10 The relative merits of uniform price auctions versus pay-as-bid auctions are beyond the scope of this article. However, the interested reader should see Giulio Federico and David Rahman, Bidding in an Electricity Pay-as-Bid Auction, Working Paper, available at http://www.stoft.com/e/lib/html/Lau_Raham.htm, last updated April 2001. They model the U.K. spot electricity market under perfectly competitive uniform price and pay-as-bid scenarios, and find that while output (supply) is higher under a uniform price auction, prices are lower in a pay-as-bid auction. 11 A theoretical discussion of our alternative market concentration measure is given in Dmitri Perekhodtsev, Lester Lave, and Seth Blumsack, An Application of the Theory of Pivotal Oligopoly to Electricity Markets, Working Paper, Carnegie Mellon University, 2002. 12 See also Severin Borenstein, James Bushnell, and Christopher Kittel, Market Power in Electricity Markets: Beyond Concentration Measures, University of California Energy Institute POWER, Working Paper PWP-059, 1998. They argue that the HHI is also not an appropriate market power metric for industries in which the price elasticity of demand is very small in magnitude. The doctrine of “obligation to serve” in the electricity industry has likely imposed very small short-term price elasticities on the market. 13 The Nash equilibrium (the equilibrium strategy in which no player has an incentive to deviate) in a one-shot prisoner’s dilemma game is the non-cooperative equilibrium in which each generator bids a low price. However, in a repeated prisoner’s dilemma (in which the game is played more than once, with the same players and same payoffs), the equilibrium strategy may be for the generators to cooperate. See Drew Fudenbergand Jean Tirole, Game Theory (Cambridge, Mass.: MIT Press, 2002), at 110. 14 See FERC Order in Docket Nos. ER96-2495-015 and 97 FERC 61,219 (2001). 15 The calculations behind Figures 1 through 3 are made using load, generation, and import data from the California ISO, PJM, and New York ISO (where appropriate). Sufficiently detailed import data was not available for PJM or the New York ISO. Capacity factors are estimated using generation data from EIA Forms 759 and 906. 16 Note that the calculations presented here do not explicitly account for outages. To the extent that certain generating units were not available in some hours (and not being withheld), the figures presented here will understate the degree of market power. 17 The IOUs were allowed to contract forward using the Power Exchange “block forward market,” but were restricted as to the volumes of allowable contracting. Consequently, few parties other than the IOUs bid into the block forward market. 18 The data underlying Figure 4 were provided by the Energy Market Report, published by Economic Insight, Inc., of Portland, Ore. The data represent bilateral day-ahead transactions in California and PJM. 19 This calculation is made as follows: If a gas turbine costs $500,000/MW to build, with an interest rate of 10 percent, its annual cost is $50,000/MW—assuming it is kept on standby. If total capacity is 54,000MW and a 25 percent capacity expansion is needed to prevent the exercise of market power, additional generation needs amount to 13,500MW. The annual cost of these 13,500MW is 13,500×50,000=$675 million. Data from the EIA indicate that generation in California amounted to 221 million MWh in 2000. Dividing $675 million by 221 million yields 3.05$/MWh or 0.305cents/kWh. Adding access to large quantities of gas, property acquisition, and connection to the grid could easily triple these costs, raising the electricity price by 0.915cents/kWh. 20 This calculation uses figures from Eric Hirst, Key Transmission Planning Issues, Elec. J., Oct. 2001, at 59–70. The capital cost of a 500-kV line is assumed to be $1.2 million per mile. Again assuming an interest rate of 10 percent, the annual cost would thus be $0.12 million per mile. Necessary capacity additions again amount to 0.25×54,View the MathML source,500MW, meaning that 6.75 lines would need to be built if each line has a capacity of 2,000MW (real transmission investments would, of course, be lumpier than in this stylized example). The total annual cost for the transmission lines (again, assuming that they are never actually used) would be View the MathML source million. The implied cost increase is thus 0.367cents/kWh. 21 See Casey et al. Potential Economic Benefits to California Load from Expanding Path 15—Year 2005 Prospect, 2001, California ISO (CAISO). The study uses the Residual Supplier Index as an indicator of the market competitiveness, and concludes that an expansion of the transmission network connecting the northern and the southern parts of California may result in significant mitigation of market power. 22 Ownership shares in generating units can, of course, be divided, but operating control, as a practical matter, would be difficult to divide. 23 This calculation includes nuclear units. The largest thermal unit in the CAISO system represents 1.6 percent of capacity. 24 In particular, if retail prices rise by the full 0.3 to 1 cent per kWh needed to mitigate market power, this would represent (at the very least) a 3 to 10 percent increase in retail prices. Based on data from the EIA Electric Power Monthly, prices in Pennsylvania (where deregulation appears to have been most successful) have fallen by only 5 or 6 percent. 25 This is the case, for example, in Arkansas.