LexisNexis(TM) = Academic

2001

22 Energy L. J. 335 =

LENGTH: 12773=20 words

ARTICLE: VENUES AND MARKETS: = REGULATING=20 COMPETITIVE ELECTRICITY IN THE WEST

Robert J. Michaels*=20

* Professor of Economics, California State = University,=20 Fullerton, and Affiliate Consultant, Tabors, Caramanis & Associates. = B.A.=20 University of Chicago, Ph.D University of California, Los Angeles. The = opinions=20 expressed in this article are not necessarily those of the author's = affiliations=20 or clients.

SUMMARY:
... It is the view = here that=20 the Federal Energy Regulatory Commission (FERC or Commission) can best = advance=20 competition in electricity if it replaces or supplements its comparisons = between=20 energy prices and expenses with analyses of relevant = markets=20 defined on the basis of opportunity costs. ... The institution and = continuation=20 of these caps was rationalized on grounds that certain owners of = California=20 generation had market power and by their individual = decisions=20 were able to affect prices at the Cal-ISO and PX. ... Throughout the = chaos that=20 has overtaken California since mid-2000, large volumes of short-term = wholesale=20 energy continue to be traded over the Western Interconnection. ... = Because the=20 relevant market cannot simultaneously be all three of = these,=20 its case for generation market power is incomplete at = best. ...=20 Even if California is a relevant market, a 9% share of = capacity=20 in it would not normally cause market power concerns = under=20 federal Merger Guidelines. ... Short-term energy has dominated = California as a=20 consequence of legislated exchange institutions and requirements. ... = The Order=20 also generalized the requirement that unscheduled California generation = be=20 offered to the Cal-ISO with a requirement that all public and non-public = generators in the West (hydro excepted) offer their energy in "the spot=20 market of their choosing. ... Even narrow studies of = the=20 California PX and ISO often do not produce substantial structural = measures of=20 market power. ... Relevant markets = defined on=20 opportunity cost considerations form a superior foundation for the = analysis of=20 market power. ...

TEXT: =
[*335]

I. Introduction

Regulatory = agencies seldom actually practice cost-of-service regulation. Much of = their=20 activity is better described as expense-of-service regulation. When = regulated=20 industries are monopolies the distinction is immaterial, but as the = scope of=20 potential competition expands, the distinction becomes critical. = Regulatory=20 methods that worked acceptably in controlling the = market=20 distortions of unregulated monopolies are likely to create distortions = of their=20 own if indiscriminately applied to competitive markets. =

It is the view here that the Federal Energy Regulatory = Commission (FERC=20 or Commission) can best advance competition in electricity if it = replaces or=20 supplements its comparisons between energy prices and expenses with = analyses of=20 relevant markets defined on the basis of opportunity = costs. The=20 Commission's recent responses to California's electrical chaos provided = the=20 ideal opportunity for change. The FERC's November 1, 2000, Proposed=20 Market Order moved away from sole reliance on expenses = and=20 included opportunity costs as potential justification for supplier bids = above a=20 "soft" price cap. n1 The final Market Order issued on = December=20 15, 2000, rejected opportunity costs, using rationales that cast light = on=20 shortcomings in the Commission's basic approach to the analysis of = competition.=20 n2 The FERC's continuing adherence to expense-based regulation becomes = an=20 ever-riskier guide to policy as markets become more = complex and=20 competition becomes possible in a broader range of activities. =

In the=20 regulation of monopolies, it is often appropriate to equate foregone = economic=20 opportunities with recorded expenses. In competitive situations, = equality=20 between expenses and opportunity costs will be the exception rather than = the=20 rule. Any method of market analysis risks two types of = error.=20 First, analysis can conclude that market power exists = or has=20 been exercised when in reality competition prevails, producing policy = rec-=20 [*336] ommendations that foreclose desirable behavior. = Second,=20 analysis can conclude that competition exists when, in reality,=20 market power is excessive, producing recommendations = that allow=20 undesirable behavior to continue. If the FERC continues to base policy = on=20 comparisons of prices and expenses, it runs an increasing risk of = producing the=20 first type of error. It will be worse than ironic if an agency that has = so=20 consistently attempted to further competition in the industries it = regulates=20 continues to base its policies on standards that competition has = rendered=20 obsolete.

Section II of this article provides background = information on=20 power trading in the West, discussing its geography and the patterns of = product=20 flow within the region. With this background and data on prices, it = becomes=20 clear that the Western Interconnection is a relevant = market for=20 short-term power. Following that background, the operations of the = California=20 Power Exchange (PX) and the California Independent System Operator = (Cal-ISO or=20 ISO) are examined. Because casual references to the PX and ISO as=20 markets have produced important misunderstandings, they = will=20 subsequently be referred to as venues. A = venue=20 is an alternative location or a mode of transacting available to buyers = and=20 sellers who wish to participate in a relevant market. = Over a=20 typical day the decisions of a generator operating in the PX and ISO can = number=20 in the thousands, and that number has decreased only slightly with the = closing=20 of the PX in early 2001. Many of those decisions require commitments = that=20 foreclose other choices, and all must be made under conditions of = incomplete=20 information.

Section III of this article summarizes the relevant = events=20 and policy responses prior to spring 2001, beginning with the lowering = of caps=20 on real-time power at the Cal-ISO in Summer 2000. The institution and=20 continuation of these caps was rationalized on grounds that certain = owners of=20 California generation had market power and by their = individual=20 decisions were able to affect prices at the Cal-ISO and PX. Bids = substantially=20 in excess of operating expenses at these venues have = been=20 viewed as evidence of market power, while relatively = little=20 notice has been given to the convergence of prices for similar power = deliveries=20 at venues around the West. In November 2000, the FERC = proposed=20 a "soft cap" on bids into the PX and the Cal-ISO. Bids above the cap = were to be=20 judged acceptable if justified by cost, which could either be operating = expenses=20 or foregone opportunities. The Commission's December Final Order = disallowed=20 opportunity cost as a potential justification. After rebutting = ill-founded=20 objections to an opportunity cost standard, the FERC chose to reject = that=20 standard on questionable grounds.

Section IV attempts to make = the case=20 for opportunity cost in two ways. First, it examines anti-opportunity = cost=20 arguments and finds them generally lacking in logic or applicability. = Second, it=20 examines the FERC's series of "mitigation" orders issued during the = first half=20 of 2001. These orders set maximum prices on the basis of measured = expenses.=20 Lacking a basis in opportunity costs, these orders have understated the = size of=20 the relevant markets, leading to modifications and = expansions=20 of scope as their actual range has become clear. =

[*337] =20 Section V concludes that as electricity becomes more competitive, the = likelihood=20 increases that the FERC's analytical methods will fall into error. The=20 superficially similar methods used by the antitrust agencies and courts = are less=20 prone to these shortcomings. As the regional transmission organizations = (RTOs)=20 that will facilitate market growth form in response to = the=20 FERC's initiatives, the time is becoming ideal for the Commission to = rethink its=20 basic approach to competition.

II. Restructuring in California=20

A. Transactions in California and the West

1. Western=20 Markets

Since the construction in the = 1960s and=20 1970s of high-voltage lines linking California with the Pacific = Northwest and=20 Desert Southwest, power transactions among the regions have increased, = along=20 with the diversity of contractual arrangements underlying those = transactions.=20 Between 1990 and 1999 imports accounted for 23% of California's retail = power=20 consumption. n3 Both the volume of power trading under idiosyncratic = contracts=20 and standardized transactions for short-term energy (to flow within the = next=20 twenty-four hours, also sometimes called spot transactions) have = increased with=20 the growth of independent power production and marketing. Transactions = across=20 the state's borders benefited California's utilities and their customers = through=20 the 1990s. Inexpensive imports eased the burden of uneconomic plants = built=20 during the 1970s and 1980s when state policy favored micro-management of = resource choices. The transactions allowed postponement of politically = difficult=20 decisions on new plants in California, and reduced the rate impact of=20 above-market utility contracts with independent power = producers=20 negotiated during the 1980s. California was, and continues to be, both a = demander and a supplier. Hydroelectric power from the Northwest lowered=20 Californians' summer bills while exports from the state helped the = Northwest=20 meet its winter peak. In exchanges with the Southwest, temperature = differences=20 were major determinants of whether California imported power from the = Southwest=20 or exported it there. n4

Throughout the chaos that has overtaken = California since mid-2000, large volumes of short-term wholesale energy = continue=20 to be traded over the Western Interconnection. n5 Amounts traded by = utilities,=20 power market- [*338] ers (some representing = retail=20 customers), and independent power producers have continued to rise until = recently. Short-term and longer-term bilateral transactions continue to = coexist,=20 but confidentiality, particularly for the latter, makes information on = volume=20 and pricing difficult to obtain or estimate. High correlations of = short-term=20 power prices between various delivery points in the west indicate that = the=20 region has become a unified, competitive market. The = FERC's=20 Staff Report on the events of Summer 2000 found correlations of 0.99 = between=20 daily on-peak prices at strongly interconnected delivery points in the=20 Northwest. Of other correlations calculated by Staff, the lowest was = 0.86,=20 between prices for delivery in northern California (NP15) and two points = in the=20 Southwest. n6 Other researchers examining different time periods and = price=20 samples have found similarly high correlations. n7 Changes in localized = supplies=20 and demands produce transactions that move power in directions expected = in a=20 unified market. The correlations appear to arise in = response to=20 immediate market forces rather than to external factors = that=20 similarly affect otherwise unrelated transactions. n8 Outside = California,=20 short-term transactions provide elements of diversified supply packages = that=20 utilities will resell at retail. The portfolios consist of contracts = with=20 diverse provisions and terms, as well as generation that utilities = continue to=20 own. Outside of California, utilities generally depend on short-term = power for=20 no more than ten percent of their sales to final customers. Most = utilities that=20 purchase short-term power also appear frequently as sellers of it. = California's=20 extreme dislocations stemmed in part from provisions in its = restructuring=20 legislation that left it almost completely dependent on unhedged = short-term=20 prices.

B. California's Restructured Institutions
=
In=20 1994, the California Public Utilities Commission instituted a rulemaking = and=20 investigation into allowing retail customers and their agents to = transact with=20 power producers, to be known as Direct Access. n9 After two years of = political=20 conflict over market designs and the recovery=20 [*339] years of political conflict over = market=20 designs and the recovery of of transition costs by utilities, Assembly = Bill (AB)=20 1890 became law in September 1996. n10 AB 1890 authorized, inter alia, = formation=20 of the California PX, a bid-based venue whose primary = function=20 was to match bids and demands for power to be delivered over the next = day. The=20 law also authorized the Cal-ISO, a nonprofit organization separate from = the PX,=20 to operate the transmission systems of the three major corporate = utilities as a=20 unit. The Cal-ISO also administers day-ahead and hour-ahead bidding=20 venues for ancillary services (reserves of varying = readiness=20 and load-following services) and a "real-time" or "imbalance" = market=20 (venue) for power that will flow within the next hour. = Alongside the PX=20 and Cal-ISO venues, AB 1890 allowed bilateral = transactions for=20 both retail and wholesale consumers and required the Cal-ISO to schedule = them in=20 a nondiscriminatory manner.

AB 1890 froze utility rates for all = retail=20 users at pre-enactment levels, with an additional bond-financed 10% cut = for=20 small customers. The state's three large corporate utilities were = required to=20 purchase all of their power through the PX and Cal-ISO. n11 They would = recover=20 their stranded costs in the residual between frozen retail rates and = fluctuating=20 wholesale prices, over a transition that would largely end in March = 2002. After=20 that date they could no longer recover transition costs in rates. AB = 1890=20 required the two largest utilities to divest approximately half of their = in-state gas-fired generators, but ultimately all three divested all of = those=20 plants. Most sold for more than book value, with the premium credited = against=20 transition costs. n12 Their new owners could bid into the various state=20 venues or arrange bilateral transactions as desired. = The PX and=20 Cal-ISO began operating in April 1998 as FERC-jurisdictional entities = allowed to=20 charge market-based rates conditional on the taking of = certain=20 actions to mitigate market power. n13=20

[*340]

C. A Generator's Bids in the PX and = Cal-ISO=20

Between April 1998 and January 2001, a strict daily time = line=20 governed the bidding of any generator that intended to participate in = the=20 venues operated by the PX and Cal-ISO. n14 The PX = ceased=20 operations in January 2001 due to utility insolvencies, but the = Cal-ISO's=20 ancillary services and real-time trading continue to operate largely as = before.=20 The number and range of potential bids into the Cal-ISO = venues=20 is so great that the absence of a PX still leaves a generator with a = highly=20 complex set of decisions. Simultaneous consideration of day-ahead, = hour-ahead,=20 and real-time bids in auctions for several types of service forces a = seller to=20 base its choices on the opportunity costs of revenue it does not receive = in=20 alternative venues rather than operating expenses. A = generator=20 must make its decisions on the basis of available information and = forecasts.=20 With minor exceptions, its owner cannot reverse previously made = commitments in=20 light of new information. A typical day when both the PX and Cal-ISO = operated=20 illustrates the complexity of the choices. n15 By 7:00 a.m. on the day = before=20 power is to flow, a generator or marketer submits "portfolio" bids for = each hour=20 of the PX day-ahead energy market. These bids do not = identify=20 specific generation resources. The Cal-ISO stacks them in ascending = order and=20 matches them against day-ahead demands submitted by utilities and other=20 purchasers. Within fifteen minutes, this process produces an = "unconstrained=20 market-clearing price" and aggregate quantity for each = hour of=20 the day ahead. With these data and the ISO's independently-made load = forecast in=20 hand, the seller must by 9:10 a.m. identify the units it intends to = operate and=20 the quantities it will produce at various prices. n16 At that time, the = seller=20 may also submit "schedule adjustment bids" to change the operations of = its=20 unit(s) in the event transmission is congested. With resources thus = identified,=20 the Cal-ISO examines the proposed dispatch for transmission congestion. = If=20 congestion is a concern, an iterative process of bid-taking and schedule = adjustment follows. n17

A seller must evaluate transactions in=20 alternative venues on the basis of opportunity costs.=20 Generation dedicated to the day-ahead PX is unavailable for bids into = the=20 day-ahead ancillary services markets, the various = hour-ahead=20 markets, and new bilateral transactions. Just after the = generator makes its day-ahead commitments at 9:30 a.m., its bids to = supply the=20 various ancillary services for each hour of the next day are due at the = ISO.=20 [*341] An ancillary services bid must specify (1) the amount = of a=20 generating unit's capacity that will be at the ISO's disposal; (2) the = minimum=20 price it will accept for that capacity in each ancillary service = (regulation or=20 load-following, spinning reserves, non-spinning reserves, and = replacement=20 reserves); and (3) the minimum it will accept for energy produced if the = capacity operates. The Cal-ISO uses industry standards and load = forecasts that=20 are known to the bidder (regionalized in the event of transmission = constraints)=20 to determine its ancillary services requirements. It matches capacity = bids with=20 its requirements for each service in sequence. During the next day's = operations,=20 it only runs those units whose energy bids are below the price then = prevailing=20 in Cal-ISO's real-time energy market. Units that = operate to=20 produce a given ancillary service receive both the=20 market-clearing capacity bid for that service and the = real-time=20 energy price for their output.

Day-ahead PX and ancillary = services=20 transactions are alternative opportunities that require commitments to = operate=20 or be available to operate. Both exclude the generator from numerous=20 shorter-notice transactions for which bidding remains open. The seller's = capacity can be bid into the PX's day-of markets (for = delivery=20 during three multi-hour intervals within the next twenty-four hours) or = the=20 hour-ahead markets for ancillary services, in which the = same=20 multi-part bidding structure prevails as for the day-ahead=20 markets. Changes in system conditions not anticipated = when=20 day-ahead commitments are made can change the rewards available in = shorter-term=20 markets.

Finally, the real-time = market=20 allocates what energy remains. Bids of otherwise unused capacity = ("supplemental"=20 bids) can be made up to forty-five minutes prior to the hour of = operation. In=20 addition to these last-minute bids, real-time supplies come from units = being=20 operated to supply ancillary services, from generators producing more = than their=20 initial PX schedules (to profit from higher real-time prices than = expected) and=20 from loads choosing to take less than they have scheduled (to lessen the = impact=20 of high real-time prices). Demand for real-time power comes from the ISO = as=20 purchaser of the ancillary energy, from generators producing less than = their=20 scheduled amounts, and from loads taking more than their scheduled = amounts. An=20 important increment of real-time demand during the last half of 2000 was = also=20 motivated by opportunity cost considerations. Utilities chose to = under-schedule=20 their day-ahead loads in the PX to take advantage of caps on real-time = prices=20 that were lower than caps on PX prices. n18 The real-time price is the = outcome=20 of numerous buyer and seller decisions based on opportunity costs. Prior = to the=20 dislocations of summer 2000, arbitrage between the day-ahead and = real-time=20 markets yielded high [*342] correlations = between=20 their prices and near-equality between their average values. n19=20

Referring to the bid-matching venues operated = by the PX=20 and Cal-ISO as markets does not make them relevant=20 markets for the analysis of competition. A bid into any = of the=20 venues inexorably carries an opportunity cost that = often cannot=20 be precisely known at the time of the bid. With the unfolding of the day = comes=20 information that alters the costs and benefits of what choices remain. = In the=20 day-ahead, hour-ahead, and real-time venues for = ancillary=20 services and energy, over the space of twenty-four hours a generator = must make=20 decisions on at least 480 price bids, 146 capacity commitments, 192 = capacity=20 reservation prices, and forty-eight ancillary services energy price = thresholds.=20 n20 The choice to abstain from operating in some venue = can be=20 as important as the choice of amounts and quantities to bid in the=20 venues one does enter. Beyond all of these choices are=20 bilateral transactions. Since early 2001, state government has become a = buyer of=20 both long-term and short-term power in response to the utilities' loss = of=20 creditworthiness. The choices of a generator that has not otherwise = committed=20 its output, however, are as complex as before.

III. The = California=20 Collapse and The FERC's Response

A. The Summer of 2000 =
=20
The PX and ISO began operation on April 1, 1998. Over the next two = years=20 prices at the PX usually varied predictably with market = conditions and generation costs, both seasonally and over the course of=20 individual days. n21 Prices at the Cal-ISO were less stable. Within two = months=20 of its opening, the prices for certain ancillary services had exceeded $ = 10,000/=20 MWh on three occasions. The FERC allowed the Cal-ISO to impose $ 250 bid = caps,=20 which were raised to $ 750 in October 1999. n22 The PX price was=20 [*343] capped from the outset at $ 2,500/MWh and never = reached that=20 level. By June 1999, a combination of high prices for divested plants = and low=20 prices at the PX and Cal-ISO had allowed San Diego Gas & Electric = Company to=20 collect all of its transition costs and replace its frozen rates with = rates that=20 passed PX and ISO prices through to its retail customers. n23 By early = 2000,=20 Pacific Gas & Electric and Southern California Edison were on = schedule to=20 complete their transition cost recoveries within the legislated = deadline. That=20 expectation ended in May of that year as PX and ISO prices rose well = above=20 historical levels and stayed there. n24 Still subject to the rate freeze = of AB=20 1890, the two companies' wholesale energy bills soon exceeded retail = revenues.=20 n25

The Board of Governors of the Cal-ISO lowered real-time and=20 ancillary services price caps twice, from $ 750 to $ 500/MWh on June 28, = and to=20 $ 250 on August 1. n26 In response to a complaint by Morgan Stanley = Capital=20 Group, Inc., the FERC validated the Cal-ISO's authority to lower caps on = its own=20 volition, but also determined that the ISO could not force sellers to = bid into=20 its venues without amending its tariffs. The FERC "put = the ISO=20 on notice that any amendment to mandate sales must be accompanied by a=20 demonstration that this extreme measure is the proper response to low = supplies=20 in the [Cal-]ISO markets." n27 Later in August, the = FERC=20 acceded to San Diego Gas & Electric's August 2 request for a hearing = into=20 the reasonableness of California prices. n28

B. The DMA Analysis =

1. Market power and price caps
=
With the=20 ISO's powers to independently impose price caps set to end = [*344] on=20 November 15, 2000, its Department of Market Analysis = (DMA)=20 issued an August 10 report supporting a request for an extension of that = authority. n29 Perhaps inadvertently, the request pointed out conflicts = between=20 competition and expense-based regulation. Assuming that the Cal-ISO's = ancillary=20 services and real-time venues constituted the relevant=20 market, DMA noted:

The presence = of=20 market power can be verified by bids setting=20 market clearing prices significantly over the variable = costs of=20 the highest cost supplier in the ISO's markets. The = highest=20 variable cost of in-state generators is approximately $ 100/MWh and many = suppliers have routinely bid a significant part of their capacity at the = ... [$=20 750] price cap level. This shows that during high load periods, many = bidders are=20 pivotal, meaning their bids are guaranteed to be selected and thus they = can=20 influence the market clearing [real-time] price. n30 =
=20
At the same time, however, the DMA stated that California's = inadequate=20 supplies resulted from their ability to operate in a wider relevant=20 market extending beyond the state's boundaries:=20

[A] significant amount of thermal capacity = divested to=20 merchant generators is being scheduled through bilateral contracts or = through=20 regional block forward and spot markets. Of capacity = scheduled=20 through these bilateral or forward market contracts, a=20 significantly greater amount appears to have been purchased for = out-of-state=20 markets. n31

Within California, the DMA = saw=20 generators viewing the Cal-ISO and the PX as alternatives and used = opportunity=20 cost to explain bidding behavior, noting, "high real time and = replacement=20 reserve purchase prices and quantities have created a significant = opportunity=20 cost that may have led suppliers to withhold or bid higher prices in the = PX Day=20 Ahead market." n32 The DMA thus claims three relevant=20 markets in which to analyze the competitive behavior of = generators: (1) the Cal-ISO's venue(s); (2) the West; = and (3)=20 the Cal-ISO and PX venues. Because the relevant=20 market cannot simultaneously be all three of these, its = case=20 for generation market power is incomplete at best. =

As=20 price caps at the Cal-ISO were lowered during mid-2000, utilities = migrated=20 toward power traded there in accordance with opportunity costs. With a $ = 2,500=20 price cap in the PX and a $ 250 price cap on real-time power, the = utilities=20 began underscheduling their demands in the PX and moving them to the = Cal-ISO's=20 real-time market. In prior years, approximately 5%=20 [*345] of Cal-ISO throughput traded in real-time, but as = demand=20 migrated the figure went over 25%. n33 Generators followed utilities to = the=20 real-time market, but if not enough power was = forthcoming at=20 capped prices, the Cal-ISO would make "out-of-market"=20 purchases, possibly outside of California, at whatever prices prevailed. = ISO=20 staff estimated that out-of-market purchases in = response to=20 underscheduling raised its expenses by over $ 100 million during summer = 2000.=20 n34

2. Venues and markets

The = real-time=20 and ancillary services markets are only relevant for = analyzing=20 market power if they contain the alternative = opportunities=20 available to buyers and sellers. If buyers and sellers can migrate among = venues, arbitrage will tend to equalize their prices as = it has=20 for short-term power at delivery points in the Western Interconnection.=20 Analyzing the effects of concentration on price requires examining the = abilities=20 of both sellers and buyers to migrate among venues. = Migration=20 may be between locations and also between transactional arrangements = such as=20 long-term and short-term contracts, each of which has attributes of a=20 venue. On the other side, some generators may have = long-term=20 contract obligations that leave little of their capacity available for=20 short-term transactions. Calculation of a generation owner's=20 market share in short-term power without adjusting for=20 precommitments may overstate its ability to influence=20 market-clearing prices.

The DMA acknowledged = the=20 existence of alternative transactions outside of the PX and Cal-ISO, but = its=20 lack of a well-defined relevant market led to = implausible=20 inferences about the effects of seller concentration. The DMA noted that = the two=20 largest owners of divested generation each had "market = shares"=20 of approximately 9%, and if 91% of "available capacity" were in use, = each could=20 "profit from causing price spikes." n35 The DMA did not define the = relevant=20 market, but one candidate consistent with its figures = is=20 California generation that is not owned by municipal utilities, state = agencies,=20 or cooperatives. The DMA's report provides no rationale for = non-governmental=20 power produced in California being a relevant market, = and its=20 discussion of transactions outside the state is consistent with a = West-wide=20 market. The state is not a relevant geographic=20 market under a commonly accepted standard of = cross-border flows=20 that are under 10% in both directions. n36 Even if California is a = relevant=20 market, a 9% share of capacity in it would not normally = cause=20 market power con- [*346] cerns under = federal Merger=20 Guidelines. n37

If a market contains two = bid-based=20 venues that operate simultaneously, the opportunity = cost of=20 bidding into one is the revenue foregone by not bidding into the other. = To cover=20 that cost, the amount a generator will bid into one must equal or exceed = its=20 expectation of what the equilibrium price will be in the other. If = regulators=20 monitor the venues, they will see bids in excess of = operating=20 costs and possibly conclude that market power is being=20 exercised. The disparity between bids and expenses will occur even if = both=20 venues satisfy the assumptions of a perfectly = competitive=20 market model. In effect, the existence of multiple=20 venues turns each of them into an exchange where = generators are=20 paid as-bid. Equilibrium prices are equalized and equal to opportunity = costs.=20 n38

If the market consists of a single=20 venue rather than a number of them, a competitive = seller will=20 rationally bid its operating costs. n39 Bidding more is risky because = that bid=20 may be above the equilibrium price, in which case the plant does not = operate.=20 Had it bid its true costs, the idle plant would have been earning more = than its=20 variable costs, leaving some amount to apply against fixed cost. A = seller that=20 can affect price in a venue may leave some capacity = idle and=20 bid the rest in at more than operating cost in order to raise price, if = doing so=20 increases total revenue. Prior to recent reforms, the United Kingdom = (U.K.) pool=20 market was essentially the sole venue=20 available to generators there. Some researchers have found that bids = into the=20 pool that exceeded operating costs indicated the exercise of=20 market power. n40

C. The FERC's Proposed=20 Market Order: Soft Caps and Opportunity Costs =

On=20 November 1, the FERC Staff issued a report the Commission had requested = in=20 August. Instead of referencing "California markets" or = singling=20 out certain venues, the report's title was "Western=20 Markets and the Causes of the Summer 2000 Price = Abnormalities."=20 n41 The Staff viewed underlying competitive conditions and the = associated power=20 flows as respon- [*347] sible for unifying a regional=20 market. Its report began with a discussion of the past = decade's=20 imbalance between demand and capacity growth throughout the region, and = went on=20 to show the effects of heat in the Southwest and low water levels in the = Northwest on California's imports and exports. n42 After accounting for = higher=20 costs of fuel and emission permits, the Staff concluded that "prices in = some=20 hours appear to be above those that would have prevailed in a = competitive=20 short-term (hourly) market, if the competitive prices = were=20 determined from short-term marginal costs." n43 The last statement = carries an=20 implication that competitive prices might be determined by factors other = than=20 operating expenses. n44 An alternative is that they are set by = opportunities=20 open to sellers, an implication of the Staff's finding of very high = correlations=20 between prices at various Western delivery points. n45

The = report noted=20 problems in determining market power when sellers = operate in=20 linked venues:

When market power = is=20 exercised, the market clearing price exceeds the price = that=20 would have been set under competitive conditions. It is important to = note that a=20 generator's true marginal cost is the generator's opportunity cost of = selling=20 into a particular market. That is, the next highest = value of=20 the resource. If the running cost of a unit is $ 40 per MWh, but that = unit is=20 physically able to sell into a market in which the = price would=20 be $ 80 per MWh if that generator participated in that = market,=20 then the opportunity cost of selling into another = market is $=20 80 per MWh. As long as the generator bids its true opportunity cost into = a=20 market, it will never receive less than the true value = of its=20 output. n46

It is, however, possible to determine = whether=20 market power has been exercised in a system of multiple = venues:

A generator could = exercise=20 market power through either economic or physical = withholding.=20 In the case of economic withholding, a generator would submit bids in = excess of=20 its opportunity cost in order to raise the market = clearing=20 price. In the case of physical withholding, the generator would not = supply all=20 of its available energy in order to increase the=20 market-clearing price. n47

=
Also=20 on November 1, the FERC issued its Market Order = Proposing=20 [*348] Remedies. n48 The order proposed a "soft cap" of $ = 150/MWh=20 (or per MW of ancillary services capacity) on all transactions in the PX = and the=20 ISO for the next twenty-four months, with no adjustments for (e.g.) = changes in=20 fuel or emissions costs. n49 If a venue clears at less = than $=20 150, the soft cap is irrelevant. If the PX or the ISO must accept bids = over $=20 150, the highest bid below $ 150 sets the price received by all who have = bid=20 less. Those bidding over $ 150 are paid their individual bids but must = file=20 weekly reports at the FERC that include data on incremental generation = cost.=20 Most importantly, "the filing may also identify legitimate opportunity = costs=20 that are known and verifiable that the seller considered in developing = its bid,=20 i.e., prior to the transaction." n50 The Proposed Order did not detail = the=20 consequences of justification by opportunity costs. The most important = was that=20 prices contemporaneously prevailing over the remainder of the West would = also be=20 those paid by Californians. If the entire West constituted a competitive = market then California utilities would pay competitive = prices=20 for most of their power supplies. Since retail rates were frozen, = California=20 ratepayers would not benefit from any type of cap, but the abilities of=20 utilities to collect their stranded costs would be affected.

D.=20 Opportunity Costs Abandoned: The FERC's December Order
=
The FERC=20 finalized its Market Order on December 15, 2000, = explicitly=20 rejecting opportunity cost as justification for bids above the $ 150 = soft cap.=20 n51 The Commission found little merit in most comments on the November = Proposed=20 Order that opposed opportunity cost. It nevertheless rejected the = concept using=20 a rationale that had been suggested by one of the concept's supporters. = Several=20 commentators understood that an opportunity cost justification would = render the=20 soft cap a nullity. The PX's Market Monitoring = Committee urged=20 rejection of opportunity costs by arguing that if prices were similar = throughout=20 the West (presumably set by competition) then the use of opportunity = costs in=20 the PX and the ISO would subject California to = market-based=20 pricing. n52 Somewhat inconsistently, that Committee was also concerned = that=20 parties selling into the PX would enter other venues = unless the=20 FERC treated opportunity cost justifications permissively. n53 The = California=20 ISO saw that caps applying only to venues operated by = it and=20 the PX would encourage exports to higher price areas, [*349] = with a=20 consequent rise in out-of-market transactions and = above-cap=20 prices. n54

The FERC took opportunity costs out of the picture = with a=20 discussion of the real-time market:

=
In=20 recognition of the unworkable complexities that the opportunity cost = concept=20 introduces in the ISO real-time imbalance market, we = will=20 eliminate it. As Dynegy states (at 21) the major cash = markets=20 in the west ... close one hour before the California PX Day-Ahead=20 market. This market in turn closes before the ISO = ancillary=20 service markets. Therefore, a seller's opportunity to = sell in=20 these other markets has already passed. This is = particularly=20 true with respect to the ISO real-time energy imbalance = market.=20 n55

The FERC's apparent understanding is that last-hour=20 "supplemental" bids submitted to the ISO are the sole bids that will be = matched=20 against demand to determine the real-time price. If true, a generator's=20 opportunity cost of using the real-time market will = equal the=20 expenses incurred to produce the power. n56 Since no other=20 venue is available at this late hour, the real-time = price=20 should equal the marginal cost (i.e. expense) of the least efficient = generator=20 whose bid is accepted. Under the FERC's reasoning, if = market=20 power is present that price will exceed operating expense.

The = Order=20 says little more about relationships between real-time and other=20 venues. One possible line of reasoning analogizes the=20 sequencing of markets to a paradox in game theory. n57 = Assume=20 for simplicity that only two sequential venues exist in = each=20 hour. Separate prices are determined in a real-time = venue and=20 an hour-ahead venue. Real-time is the only place that = trading=20 can occur after the hour-ahead venue closes. Under = competition,=20 a generator will bid any capacity not otherwise committed into real-time = at=20 operating expense, since its only alternative is idleness. A generator = with=20 market power will bid more than its operating expense, = and in=20 either case, the real-time price will equal the highest accepted bid. As = closing=20 time for bids in the other venue approaches, generators = know=20 that the outcome in the real-time market will equal = marginal=20 expense (under competition) or exceed it (if market = power is=20 exercised). Hence, the hour-ahead venue becomes the new = "last"=20 market and generators will bid into it = [*350] at=20 their expectation of the real-time price. Sellers will arbitrage between = the=20 venues and equalize their prices. Prices in all prior=20 venues will converge on the real-time price, which = under=20 competition equals the operating expense of the least efficient = generator. The=20 real-time venue anchors prices in all = venues=20 that close earlier, and price in excess of operating expense in any=20 venue indicate that market power is = being=20 exercised.

Unfortunately for this theory, last-minute = "supplemental"=20 bids are not the sole determinants of the real-time price. The real-time = market also incorporates supplies of energy from units = that are=20 chosen to supply ancillary services.

The ISO = acquires=20 energy from any one of five sources: from the four hourly [hour-ahead = and=20 day-ahead] ancillary services markets and from = suppliers who=20 bid to provide "supplements" to their day-ahead energy schedules. As = described=20 earlier, suppliers who have committed capacity to one of the ancillary = service=20 markets, excepting regulation, and who also produce = energy,=20 receive the imbalance energy price in addition to their respective = ancillary=20 service capacity payment. Suppliers who provide energy through = supplemental=20 energy bids receive the imbalance energy payment only... the ISO = combines these=20 bids into a system-wide bid curve for supplemental energy. n58=20

A bidder in the day-ahead or hour-ahead ancillary = services markets must weigh possible energy and = capacity bids=20 against the revenue it could expect to receive in alternative=20 venues. It must earn sufficient revenue from its = combined bid=20 of capacity and energy (weighted by the probability of operating) to = recover at=20 least the amount it could have earned by operating elsewhere. If so, = real-time=20 price will be determined not by operating expenses, but by opportunity = costs=20 associated with other venues. The convergence of = real-time and=20 the PX prices observed during most of the PX's existence was driven by=20 opportunity costs. n59 Approximately half of the ISO's supplemental bids = in 1999=20 were from outside of California, presumably made by comparing local=20 opportunities with the expected value of the real-time price. n60 The=20 opportunity cost basis of the real-time market is = important=20 because some frequently cited estimates of overcharges start from a = perspective=20 similar to the FERC's. n61

[*351]

IV. Can=20 Opportunity Costs Be Operational?

A. Diagnosing Competition =
=20
If product flow data indicate that a region is a relevant geographic = market then persistent price disparities between = locations that=20 exceed the cost of shipment may be worthy of antitrust or regulatory = attention.=20 n62 If a vertically integrated transmission owner has attempted to = advantage its=20 own generation by refusing to wheel competitive power to high-price = locations,=20 price data are useful in themselves and as supplements to factual = allegations.=20 n63 For example, the evidence may show that price generally rose at = certain=20 delivery points around the times when refusals to wheel were alleged. = Since=20 neither complaints of exclusion nor defenses of refusals are always to = be=20 believed, prices and related data will still require statistical = analyses on=20 which experts might differ. n64

Regarding generation, uniform = prices can=20 characterize market equilibrium under both competition = and=20 under monopoly or collusion. No known measure of supply concentration = indicates=20 that the West is a near-monopoly. n65 Although environmental and siting = delays=20 are common, there are no other substantial barriers (e.g. patents) = facing new=20 generators, and power can be imported into environmentally sensitive = areas from=20 plants located in less sensitive ones. A successful collusion of = generators in=20 the West would require inducing a large number of normally competitive = firms to=20 act cooperatively while risking discovery and antitrust. n66 A generator = wishing=20 to break the collusion can enter confidential bilateral transactions, = and even=20 if other generators can detect increases in its output they have no = obvious=20 sanctions at their disposal.

The most essential economic = difference=20 between competition and monopoly is that a competitive = market=20 trades more output than an other- [*352] wise identical=20 market served by a monopolist. If utilities in the past = built=20 enough capacity to produce competitive levels of output, then = anticompetitive=20 behavior by owners of divested generation requires keeping some formerly = operating plants idle and attempting to block the construction of new = ones. Some=20 research claims to have found withholding of capacity from the PX and = Cal-ISO,=20 while other research questions the finding. n67 Even if the generators = avoid the=20 PX and ISO, they can continue to sell outside of California. n68 If the = PX and=20 Cal-ISO are themselves relevant markets rather than=20 venues, withholding should make their prices higher = than=20 elsewhere in the West. Price data appear to indicate equality, but a = complete=20 study has yet to be made. Other seeming cases of withholding may be = competitive=20 responses to opportunity costs. As an example, some California = generators face=20 environmental limits on their annual operating hours. Operation of such = a plant=20 during a given hour carries the risk that it cannot be operated when = (if) prices=20 rise later in the year. n69

Some commentators objecting to = opportunity=20 costs were concerned that sham transactions might provide a general = avenue of=20 escape. Since "opportunity costs" can be used to justify an = above-the-cap bid,=20 all it would take is one reported transaction somewhere in the WSCC = (which might=20 even prove to be a sham or affiliate transaction) at a higher price than = $ 150=20 to almost immediately negate the cap. Only if the soft cap were = WSCC-wide would=20 it even begin to offer some semblance of adequate consumer protection. = n70=20

The authors of the quoted comments (two advocacy organizations = for small=20 consumers) do not make clear why any buyer would pay more than a = publicly=20 visible price. If an isolated transaction somehow occurs at a price=20 significantly higher than the equilibrium prevailing elsewhere in the = relevant=20 market, a single seller who attempts to match that high = price=20 will lose business to competitors. This will also hold in the event a = seller=20 transacts with an affiliate. If collusion is possible and all sellers = raise=20 their prices at once, there is no need for an initial anomalous = transaction=20 since each can claim that its relevant opportunity cost is the price at = which=20 other sellers are selling to their customers.

Transmission = scarcity=20 makes price comparisons less straightforward. In a = transmission-constrained=20 system energy prices will vary by location. If [*353] = transmission=20 capacity into a consuming area is scarce, price there can exceed = operating=20 expenses of local generators, whether those generators are competitive = or=20 monopolistic. n71 A constrained transmission line will affect prices at = points=20 not directly attached to it. The difficulty of diagnosing monopolistic = conduct=20 in a transmission-constrained system can be eased by instituting = transmission=20 pricing that accounts for its scarcity and some system of exchangeable = rights to=20 reallocate constrained lines in accordance with their economic value. = n72 There=20 are as yet no straightforward rules available to regulators that provide = bright-line distinctions between monopolistic and competitive outcomes = in=20 systems with locationally differentiated prices. As one example, an = owner of=20 generators at several locations can at times exercise = market=20 power by operating some of them at or above competitive levels. Doing so = can=20 constrain transmission links to advantage other of its generators or=20 disadvantage those of competitors. n73

Opportunity cost = reasoning may=20 not yet be fully implementable in complex networks with transmission=20 constraints, but the alternative of price-expense comparisons is biased = toward=20 finding market power when it does not exist. There is = no=20 evidence that opportunity-cost reasoning, supplemented as necessary with = analyses of structure and conduct, favors findings of competition when = in fact=20 market power prevails. As venues = multiply in=20 relevant markets, comparisons based on price-expense=20 comparisons at single venues become riskier, and the = difficulty=20 of comparisons will further increase with the variety of transactions.=20 Short-term energy has dominated California as a consequence of = legislated=20 exchange institutions and requirements. Elsewhere in the West diversity = of=20 contracts remains the rule. Although the details of contract mix for all = of=20 California's divested generators are not yet public, one has stated that = 90% of=20 the power from its plants has already been committed under longer-term = contracts=20 that bypass the California institutions. n74 As access to the=20 market becomes more general, it will be easier for = transactors=20 to respond to onerous regulation of short-term energy by changing their=20 contractual arrangements, in effect moving to more lightly regulated=20 venues.

[*354]

B. Setting=20 Mitigation Standards

The FERC's December = Market=20 Order fixed a $ 150/MWh soft cap on all bids into the PX and ISO. A=20 progressively more detailed and inclusive set of mitigation orders has = followed.=20 Their expanding geographic scope provides evidence that the California=20 venues are not relevant markets = because they=20 contain only a minimal subset of the trading opportunities that are = actually=20 available to most buyers and sellers. The recent evolution of mitigation = policy=20 illustrates the difficulties of policy formulation absent prior = definition of=20 relevant markets and opportunity costs. The = Commission's most=20 recent (June 2001) Order appears to be at the outer limit of easily=20 implementable mitigation policies, and gives no clear path to follow if = tighter=20 controls are deemed necessary.

The FERC's December=20 Market Order acknowledged that its $ 150 soft cap could = be only=20 an interim measure, and mandated development of a mitigation plan that = would=20 replace the cap with one more closely linked to generation expenses. n75 = On=20 March 9, the FERC ordered that generators pay $ 69 million in refunds = for sales=20 during January 2001 that were in excess of a calculated $ 273/MWh "proxy = clearing price" during periods when the ISO had declared a Stage Three = (under=20 1.5% reserves) emergency. n76 The proxy price equals the estimated = marginal cost=20 of the most inefficient generator operating. Without reference to = opportunity=20 costs, the March 9 Order states that the proxy establishes the=20 market clearing price that would have occurred had the = sellers=20 bid their variable costs into a single price auction, as would have = occurred had=20 competitive forces been at work. n77 Sellers wishing to contest their = refunds=20 could submit additional justification based on expenses but not = opportunity=20 costs.

The California utilities, the ISO, and others found the = March=20 Order inadequate. Its restriction to Stage Three emergencies left 81% of = January's PX and ISO transactions unexamined. n78 They were also = concerned about=20 exports of power when prices elsewhere were above the California caps. = This=20 arbitrage, dubbed "laundering," sometimes forced the ISO to purchase at=20 uncontrolled prices elsewhere in the West. The FERC could choose either = to=20 extend price controls throughout the West or to compel sellers to offer = power to=20 the ISO. In a complex April 26 Order, the FERC extended the policy to = all times=20 when ISO reserves fell below 7.5%, and to all California generators, = including=20 non-public utility sellers doing business with the ISO. All generators = were to=20 submit heat rate data for proxy price [*355] calculations = and to=20 offer uncommitted power to the ISO at all hours. n79 Exports from = California=20 were still permitted. Not yet willing to place the entire West under the = same=20 rules at all times, it instituted a section 206 investigation of spot = prices in=20 the WSCC when reserves for any control area fall below 7%. n80 = Commissioner=20 Massey's dissent noted that "many of the transactions that are driving = the high=20 prices in [the Northwest] are for terms well exceeding twenty-four = hours." n81=20

May and June were marked by very high short-term prices and = threats of=20 West-wide price control legislation. n82 On June 19, the FERC extended=20 mitigation of spot prices to all hours and all locations in the WSCC = through=20 September 30, 2002. n83 Since there is no centralized price-determining = process=20 in general use outside of California, the FERC used California's = real-time price=20 to set the Western ceiling. If reserves are over 7% at the ISO, nowhere = in the=20 WSCC can price exceed over 85% of the highest hourly clearing price = during the=20 last Stage One alert in California, absent evidence of cost-based = justification.=20 If the Cal-ISO's reserves are below 7%, the calculated California = competitive=20 proxy price becomes the cap everywhere in the region. n84 The Order also = generalized the requirement that unscheduled California generation be = offered to=20 the Cal-ISO with a requirement that all public and non-public generators = in the=20 West (hydro excepted) offer their energy in "the spot = market of=20 their choosing." n85 The June 19 Order affects energy prices throughout = the West=20 at all hours, but its allowable prices are entirely set by events at the = ISO's=20 real-time market. n86 The FERC gives no rationale for = so broad=20 an application of prices that are determined in a venue = that=20 handles under 15% of what it calls the "California wholesale=20 market," and an even smaller fraction of total power = produced=20 in the West. n87

Like most price regulation, mitigation is = historical=20 rather than forward-looking. If the mitigated maximum price is low and = demand=20 rises unexpectedly at some location, a uniform cap on bilateral spot=20 transactions will induce localized shortages. Generators with higher = expenses=20 than California's proxy that must declare themselves available under the = Order=20 can be required to sell at a loss, which will be increased under = provisions=20 requiring the seller to pay for transmission. n88 Producers such as = industrial=20 [*356] cogenerators that need not declare themselves = available will=20 choose not to do so if the cap fails to cover their opportunity costs.=20 Non-California utilities have protested the absence of cost-based = exemptions=20 from the mitigated ceiling and have suggested that the appropriate proxy = be set=20 on a West-wide basis. n89 The Commission may compel existing generators = to make=20 themselves available and possibly sell at losses, but cannot order new=20 construction or enjoin project abandonments. n90 Utilities that = increased the=20 fraction of their supplies purchased under long-term contracts have = claimed that=20 mitigation punishes decisions that the FERC previously viewed as = prudent. It=20 leaves those who gambled on short-term markets paying = low=20 mitigated prices, while those with long-term contracts sometimes pay = more and=20 can only resell unwanted power at a loss. n91

C. Why Relevant=20 Markets Matter

Opportunity costs have = driven the=20 FERC's expansion of mitigation in the West, and they also delineate the = set of=20 substitutable contracts that mitigation does not reach. Since the June = Order=20 favorable weather, recession, and conservation have reduced demand in = the face=20 of increased generator availability, rendering mitigated prices = irrelevant at=20 most times. If mitigated prices become binding more often the = consequences are=20 less certain. Sellers continue to have options in the form of substitute = contracts whose terms cannot be reduced to the simplicity of a single = energy=20 price to compare to a spot price. They will wish to migrate from the = mitigated=20 venues to uncapped ones, and buyers will wish to move = in the=20 opposite direction. If mitigation encourages the use of mitigated=20 markets, the FERC will have to choose between extending = it to=20 new types of contracts and placing more onerous restrictions on the=20 venues in which market participants = can=20 operate. The FERC's most recent response to requests that mitigation = apply to=20 all contracts was a statement that "critical interdependence" among=20 venues implies that "any mitigation applied to the = ISO's=20 real-time markets will, over time, impact bilateral and = forward=20 markets as well." n92 There was no further discussion = of that=20 impact, the associated incentives, and the policy options that might = remain.=20

Mitigation policies are best rationalized by a belief that a=20 market is only competitive if its prices usually stand = in the=20 same relation to booked costs as would have occurred under regulation. = If the=20 relevant market subsumes venues not = subject to=20 mitigation, there is no theoretical basis for as- [*357] = suming that=20 its performance will be more competitive under selective mitigation than = with=20 none. The FERC instituted mitigation without giving noticeable attention = to=20 market definition. n93 No economic expert has found = measures of=20 seller concentration or barriers to entry that would lead to a = conclusion that=20 the broad Western market is uncompetitive. Even narrow = studies=20 of the California PX and ISO often do not produce substantial structural = measures of market power. Reliability requires that = energy=20 production match demand instantly. If demand is inelastic, price must go = to=20 levels well above operating expenses for a small number of hours to = justify=20 investments in peaking plants. If mitigation extends beyond its intended = September 2002 termination, the consequences for investment may begin to = merit=20 attention.

Contrast the FERC's treatment of = market=20 power in mergers and mitigation proceedings. The Commission's merger = procedures=20 rely largely on the same guidelines that apply to mergers in other = industries.=20 n94 Applicants, intervenors, and staff follow the procedures to = construct and=20 critique measures of concentration, geographic extent, and barriers to=20 competitive entry. The FERC's treatment of short-term energy and = capacity as=20 relevant markets reflects a reasoned determination that = mergers=20 are most likely to adversely affect competition in these products. n95 = If the=20 Commission decides that the merger must be conditioned, it links those=20 conditions to the state of competition in relevant = markets. n96=20 The FERC approved market-based pricing at the = California PX and=20 ISO using standards previously used in similar determinations and = similar to=20 those for mergers. It broadly accepted the applicants' (California = utilities)=20 definitions of markets that covered the West on the = basis of=20 power flow data, and the shrinkage of those markets = when=20 transmission constraints were binding. n97 The need to extend mitigation = to=20 encompass all short-term power in the West justifies continued use of = the region=20 as the relevant geographic market. Future movements of = buyers=20 and sellers toward and away from products subject to mitigation will = determine=20 whether short-term power alone is relevant or whether it is only a = component of=20 a broader relevant product.

V. Conclusion

If=20 cost-of-service regulation ever lived up to expectations, it happened in = the era=20 when all aspects of electricity production and delivery were in the = hands of=20 monopolies. The more competitive the industry becomes, the greater the=20 difficulty and risk in evaluating its performance using tools de-=20 [*358] vised for the regulation of monopoly. Opportunity = cost is the=20 foundation of economic thinking about productive efficiency and the = benefits of=20 competition. It conditions all types of business decisions, and accuracy = in=20 enumerating and evaluating the alternatives becomes more valuable in = more=20 competitive environments. Although the FERC did not explicitly define = the=20 relevant market as covering the West in its November = Proposed=20 Market Order, its decision to allow an opportunity cost = justification for bids above the soft cap was an implicit acknowledgment = of the=20 market's scope. Whether the Commission's decision to = abandon=20 opportunity costs in the December Order was driven by politics or by=20 misunderstanding, it marked the start of a move to ad hoc mitigation = procedures=20 that concentrated on convenient venues where prices and = costs=20 were almost as easy to compare as they had been under regulation.=20

Relevant markets defined on opportunity cost=20 considerations form a superior foundation for the analysis of=20 market power. Antitrust would probably be more costly,=20 ineffective, and inaccurate if the Department of Justice and Federal = Trade=20 Commission (and private plaintiffs) used the approach that the FERC has = applied=20 to the West. Normally the agencies would define relevant=20 markets by examining trading patterns and volumes, = observing=20 responses to price changes, examining price relationships over regions = and=20 across products, and looking for evidence of restrictive conduct. After = the=20 changeover to the FERC's California methodology, they would concentrate = on one=20 relatively minor member of the complete set of goods making up the = relevant=20 product, coincidentally the element with the most standardized design = and=20 simplest pricing formula. They would then obtain price and volume data = from one=20 or two of the venues in which it is traded. There would = be=20 little or no examination of such aspects of market = power as=20 barriers to entry of new competitors, exclusionary practices by = incumbents, the=20 buyers' side of the market, or evidence for a possible=20 collusion. The antitrust authorities would next acquire information on = variable=20 expenses at individual production facilities and compare them with = prices that=20 their owners bid into one of the previously selected venues.=20 Market power and overcharges would be calculated by examining = the=20 difference between a producer's bids into that venue = and the=20 variable expenses of the most inefficient producer. It seems unlikely = that any=20 aspect of the changeover to the FERC's methods would improve the = accuracy of the=20 antitrust agency's determinations.

Economists often view = regulation as=20 an attempt to produce the efficient outcomes of a competitive=20 market (e.g. marginal cost pricing) in situations where = a=20 monopoly industry structure is unavoidable. Outside of the FERC, they = have=20 seldom suggested evaluating the performance of a competitive=20 market with the tools used to regulate monopolies. = There are=20 important analogies between an efficient, regulated outcome and a = perfectly=20 competitive equilibrium in a single-venue market with a = single,=20 simply-defined product. The analogy vanishes as venues=20 multiply, products become more complex, and geography widens. As the = analogy=20 vanishes, the risk that regulatory tools will produce erroneous = determinations=20 can [*359] only increase.=20

FOOTNOTES:
n1. = Market=20 Order Proposing Remedies for California Wholesale Electrics, San Diego = Gas &=20 Electric Co. v. Sellers of Energy and Ancillary Serv., 93 F.E.R.C. = <para>=20 61,121 (2000), reh'g pending. [hereinafter November Proposed=20 Market Order].

n2. Order Directing = Remedies for=20 California Wholesale Electric Markets, San Diego Gas = &=20 Electric Co. v. Sellers of Energy and Ancillary Serv., 93 F.E.R.C. = <para>=20 61,294 (2000), reh'g pending. [hereinafter December = Market=20 Order].

n3. California Energy Comm'n, California = Electricity=20 Consumption by Sector (1990-2000), available at=20 http://www.energy.ca.gov/electricity/consumption by sector.html (last = modified=20 May 14, 2001); California Electrical Energy Generation, 1991 to 1999 = Total=20 Production, by Resource Type, available at=20 http://www.energy.ca.gov/electricity/electricity generation.html (last = modified=20 Apr. 9, 2001).

n4. Staff Report to the Federal Energy = Regulatory=20 Commission on Western Markets and the Causes of the = Summer 2000=20 Price Abnormalities (Nov. 1, 2000) at Part I, 2-15 - 2-18, available at=20 http://www.ferc.gov/electric/bulkpower.htm [hereinafter Summer 2000 = Western=20 Report].

n5. "Western Interconnection" is an informal = term for=20 the totality of transmission within the Western Systems Coordinating = Council=20 (WSCC), a regional reliability organization. It covers territories = (other than=20 Texas) from roughly the 100[su'th'] meridian of longitude westward, = including=20 portions of Baja California and Canada, See generally = http://www.wscc.com.=20 Because many transactions are sales and re-sales among marketers and = utilities,=20 a summary total volume for the area is not readily available.=20

n6. Summer 2000 Western Report, supra note 4, at Part I, = 3-16.=20 The coefficient of correlation can vary from 1 to -1. A value of 1 ( or = -1)=20 indicates a perfect increasing (or decreasing) linear relationship = between the=20 variables; a coefficient of zero indicates that no relationship exists, = i.e. the=20 value of one of the variables tells nothing about the likely value of = the other.=20

n7. Chi-Keung Woo et al., Electricity = Market=20 Integration in the Pacific Northwest, 18 Energy J. 75 (1997); Arthur S. = De Vany=20 & W. David Walls, Cointegration Analysis of Spot Electricity Prices: = Insights on Transmission Efficiency in the Western U.S., 21 Energy Econ. = 435=20 (1999); Samuel A. Van Vactor, Power Price Spikes: Aberration or = Prophecy?=20 available at http://www.econ.com/pubsindex.html (submitted to Oregon = Public=20 Utility Commission Aug. 14, 2000).

n8. For example, = retail=20 gasoline prices are highly correlated between regions, but since few = retail=20 customers and little gasoline moves across regional boundaries, the = nation is=20 not a relevant market. Instead, supply costs everywhere = are=20 similarly affected by movements in the world price of oil and passed on = in=20 retail prices.

n9. California Pub. Util. Comm'n, Order=20 Instituting Rulemaking on the Commission's Proposed Policies Governing=20 Restructuring California's Electric Services Industry and Reforming = Regulation,=20 Docket No. R.94-04-031 (Apr. 20, 1994).

n10. California = Pub.=20 Util. Code 330-398.5 (2001). A more detailed history and summary of the=20 legislation's content appears in Nicholas W. Fels & Frank R. Lindh, = Lessons=20 from the California "Apocalypse:" Jurisdiction Over Electric Utilities, = 22=20 Energy L. J. 1 (2001).

n11. Neither the rate freeze nor = the=20 utility purchase requirement remains in effect. Municipal utilities, = irrigation=20 districts, and cooperatives serving in total 30% of California's load = were=20 exempt from all major provisions of AB 1890, in particular the rate = freeze, the=20 restriction to PX and ISO transactions, the surrender of transmission = control to=20 the ISO (they own 40% of transmission crossing the state line), and the=20 divestiture of generation. California Pub. Util. Code 365, California = Energy=20 Comm'n, 1998 California Electric Utility Retail Sales, available at=20 http://www.energy.ca.gov/electricity /utility sales.html (last visited = Sept. 28,=20 2001).

n12. The three utilities divested a total of = 20,187 MW,=20 at an overall premium of 175% of book value. California Energy Comm'n, = Electric=20 Generation Divestiture in California, available at=20 http://www.energy.ca.gov/electricity/divestiture.html (last modified = Jan. 28,=20 2000). They continue to own 25,000 MW of nuclear plants, hydroelectric=20 facilities, and contracts with independent generators. =

n13.=20 Mitigation measures included generation divestitures and the creation of = market monitoring organizations. Utilities defined=20 time-differentiated relevant markets to account for=20 transmission constraints, but claimed that 90% of the time the Western=20 Interconnection constituted the trading area. Pacific Gas & Elec. = Co., San=20 Diego Gas & Elec. Co., & Southern Cal. Edison Co., 77 F.E.R.C.=20 <para> 61,265, 62,077 (1996).

n14. A graphic of = the time=20 line appears in Nguyen T. Quan & Robert J. Michaels, Games or = Opportunities:=20 Bidding in the California Markets, 14 Elec. J. 99,102=20 (Jan./Feb. 2001).

n15. The PX filed for bankruptcy in = March=20 2001, but for clarity the text is written in the present tense. It = remains=20 relevant because numerous ongoing investigations and lawsuits are = concerned with=20 bidder behavior in the PX during its existence. California Roundup, = Foster Elec.=20 Rpt No. 209, Mar. 14, 2001, at 5.

n16. The bids are not = simple=20 numbers. A seller can specify a supply curve with up to fifteen = price-quantity=20 points. The points are then connected by straight lines and the seller = must=20 produce the quantity on its graph that corresponds to the=20 market-clearing price.

n17. The process = of=20 pricing congestion by bid and schedule adjustments is complex and also = requires=20 consideration of opportunity costs. See generally Quan & Michaels, = supra=20 note 14.

n18. The FERC's December = Market Order=20 attempted to end this practice, which had become a source of operating=20 difficulties at the Cal-ISO by requiring that 95% of all load be = prescheduled=20 prior to the real-time market and imposing penalties if = it was=20 not. December Market Order, supra note 2. California's = three=20 large utilities are currently liable for over $ 1 billion of these = penalties,=20 which the FERC has refused to waive. For Now, FERC Refuses to Suspend=20 California's Underscheduling Penalty, Foster Elec. Rpt. No. 218, at 5 = (May 23,=20 2001).

n19. California PX, Market = Compliance=20 Unit, Electricity Markets of the California Power = Exchange,=20 Second Annual Report to FERC 42, available at=20 http://www.calpx.com/regulatory/fercfilings/ index = market=20 monitoring.html (last visited Jul. 31, 2000)(website no longer = available;=20 reference material on file with author).

n20. Quan & = Michaels, supra note 14, at 106.

n21. At times during = the early=20 months, the PX cleared at a price of zero; the consequence of a = requirement that=20 utilities bid power under contract and plants that they still owned into = it at=20 that price. Output of these "must-take" resources at times exceeded = demand.=20 California PX, Market Monitoring Committee, Second = Report on=20 Market Issues in the California Power Exchange Energy=20 Markets, Report to FERC 15, available at=20 http://www.calpx.com/regulatory/fercfilings/index (Mar. 9, 1999).=20

n22. Frank Wolak et al., Preliminary Report on the = Operation of=20 the Ancillary Services Markets of the California = Independent=20 System Operator (ISO), at 10-11 available at = http://www.caiso.com/docs/1998=20 /12/18/199812180649493813577.pdf (last visited Aug. 19, = 1998)[hereinafter Cal=20 ISO August 1998 Report]. Frank Wolak, Report on Redesign of California = Real-Time=20 Energy and Ancillary Services Market, at 8 available at = http://www.caiso.com/docs/1999/10/20/199910201045345098.pdf (last = visited Oct.=20 18, 1999)[hereinafter Cal ISO August 1999 Report]. In approving=20 market-based rates for one generation owner, the FERC = noted=20 that "the amount of capacity bid into the ancillary services=20 markets in any hour is largely a function of the = opportunity=20 cost of not using [the] same capacity to sell power in the PX energy=20 markets." El Segundo Power, LLC, 84 F.E.R.C. = <para>=20 61,011, 61,059 (1998).

n23. SDG&E Ends Rate Freeze = 2.5 Years=20 Ahead of Schedule, Stranded Costs Recovered, Elec. Util. Wk. (June 7, = 1999).=20 Although the company could now transact bilaterally, it continued to use = the PX=20 and ISO almost exclusively.

n24. Prices during the first = two=20 years were relatively low. In both years, the Northwest enjoyed a wetter = climate=20 than average, allowing California to import more power than normal. = Summer 2000=20 Western Report, supra note 4, at 2-10, 2-26.

n25. In its = unique=20 situation, San Diego Gas & Electric received special legislative = relief in=20 the form of a rate freeze for smaller customers. The legislature has yet = to=20 devise a method by which the utility can recover its revenue shortfall.=20

n26. New York ISO Seeks to Impose $ 1,300/MWh Cap; = California=20 Lowers Ceiling, Inside FERC, (July 10, 2000). The $ 500 cap applied to = capacity=20 and to energy (i.e. a generator could receive $ 500 for availability and = another=20 $ 500 if dispatched.)

n27. Morgan Stanley Capital Group = Inc., 92=20 F.E.R.C. <para> 61,112, 61,431 (2000).

n28. San = Diego Gas=20 & Electric Co. v. Sellers of Energy and Ancillary Serv., 92 F.E.R.C. = <para> 61,172 (2000). The Commission noted that San Diego "has not = documented a single instance of a seller exercising = market=20 power during times of scarcity" and that a $ 250 cap that San Diego had=20 requested "would call into question the ISO's ability to attract = sufficient=20 supply." Id. at 61,606. Regarding San Diego's exposure to short-term = prices, the=20 FERC noted that the company had eschewed hedging tools that the = California=20 Public Utilities Commission had permitted it to use. =

n29.=20 California ISO Department of Market Analysis (DMA), = Request to=20 Extend Price Caps, available at=20 http://www.caiso.com/docs/2000/08/11/2000081110583423551.pdf (Aug. 10, = 2000).=20 [hereinafter Request to Extend Price Caps] DMA is the ISO's internal=20 market monitoring division.

n30. Id. at = 5.=20

n31. Request to Extend Price Caps, supra note 29, at 4.=20

n32. Id. In this context, the DMA defines "economic = withholding"=20 as bidding a price into a venue that exceeds operating = expense.=20 See also Anjali Sheffrin, Empirical Evidence of Strategic Bidding in = California=20 ISO Real Time Market, F.E.R.C. Docket No. EL00-95-012, = March=20 21, 2001, at 4 (submitted by Cal-ISO Dep't of Market = Analysis).=20

n33. Request to Extend Price Caps, supra note 29, at 4. = The DMA=20 refers to this behavior as "defensive."

n34. Ziad = Alaywan,=20 Operation Overview (Oct. 4, 2000), available at = http://www.caiso.com/docs/=20 09003a6080/09/48/09003a60800948e1.pdf. The costs of=20 out-of-market purchases at the time were shared among = the=20 utilities rather than borne entirely by parties that underscheduled.=20

n35. Id. at 5.

n36. Kenneth G. Elzinga = &=20 Thomas F. Hogarty, The Problem of Geographic Market = Delineation=20 in Antimerger Suits, 18 Antitrust Bull. 45 (1973); Kenneth G. Elzinga = &=20 Thomas F. Hogarty, The Problem of Geographic Market = Delineation=20 Revisited: The Case of Coal, 23 Antitrust Bull. 1 (1978). =

n37.=20 In the Guidelines, a market share of 18% [a=20 Herfindahl-Hirschman Index of 1,800 if all firms are of equal size] is = the usual=20 threshold. U.S. Dept. of Justice and Federal Trade Comm'n Horizontal = Merger=20 Guidelines, 57 Fed. Reg. 26,823 (1984), reprinted in 4 Trade Reg Rep. = (CCH)=20 <para>13,103 (June 14, 1984), 1.5. 25,000 MW of California plants = and=20 contracts remain in utility hands, leaving the two largest systems each = holding=20 over 30% in the state's non-governmental generation. =

n38. The=20 point was first made in Scott Harvey & William W. Hogan, Issues in = the=20 Analysis of Market Power in California (Oct. 27, 2000), = available at http://ksghome.harvard.edu/.whogan.cbg.ksg/ HHMktPwr = 1027.pdf.=20

n39. The language of the text suffices for these = examples. The=20 plant can safely bid its operating costs plus a premium sufficiently = small that=20 its bid does not rise above those of the next highest-cost plant.=20

n40. Catherine Wolfram, Measuring Duopoly Power in the = British=20 Electricity Spot Market, 89 Am. Econ. Rev. 805 (1999); = Frank=20 Wolak and Robert H. Patrick, The Impact of Market Rules = and=20 Market Structure on the Price Determination Process in = the=20 England and Wales Electricity Market, Nat'l. Bureau of = Econ.=20 Res.m Working Paper No. W8248 (2001) available at=20 http://papers.ssrn.com/sol3/papers.cfm?abstract id=3D267435. =

n41.=20 Summer 2000 Western Report, supra note 4.

n42. Id. at = 2-1 - 2-18=20 and 2-24 - 2-28. The report presented the first easily accessible = exposition of=20 California's substantial role as a seasonal exporter of power, = particularly to=20 the Northwest in summer.

n43. Summer 2000 Western = Report, supra=20 note 4, at 3-1.

n44. The Staff's study of rising outage = levels=20 between the summers of 2000 and 2001 was limited and inconclusive. Id. = at 2-19 -=20 2-22.

n45. Summer 2000 Western Report, supra note 4, at = 3-15 -=20 3-18.

n46. Id. at 5-17.

n47. Summer 2000 = Western=20 Report, supra note 4, at 5-19. Staff later defined economic withholding = as an=20 offer to sell "at a price that exceeds the marginal cost of production = or other=20 relevant opportunity cost." Staff Recommendation on Prospective=20 Market Monitoring and Mitigation for the California = Wholesale=20 Electric Power Market, San Diego Gas & Elec. Co. v. = Sellers=20 of Energy and Ancillary Serv., F.E.R.C. Docket No. EL00-95-012 at 11 = (Mar. 2001)=20 [hereinafter Staff Recommendation on Prospective Market = Monitoring]. The DMA's definition of the same term, supra note 29, did = not=20 include opportunity cost.

n48. November Proposed=20 Market Order, supra note 1.

n49. Id. at = 89-90.=20 $ 150 was taken as "indicative of high demand" and "informed by the = running=20 costs of the gas-fired generation which is and which we expect to be on = the=20 margin in California."

n50. November Proposed=20 Market Order, supra note 1.

n51. = December=20 Market Order, supra note 2.

n52. = Comments of=20 the Market Monitoring Committee of the California Power = Exchange Corporation, San Diego Gas & Elec. Co. v. Sellers of Energy = and=20 Ancillary Serv., F.E.R.C. Docket No. EL00-95-000, at 4, (Nov. 22, 2000). = The=20 Cal-ISO and PX operated as they did because the FERC had certified their = use of=20 market-based rates.

n53. Id. at 2. The = November=20 Proposed Market Order lifted requirements that the = utilities=20 buy and sell only through the PX and Cal-ISO, but state regulators could = continue to require them to do so. November Proposed = Market=20 Order, supra note 1, at 61,360,

n54. Comments of the = California=20 ISO, San Diego Gas & Elec. Co. v. Sellers of Energy and Ancillary = Serv.,=20 F.E.R.C. Docket No. EL00-95-000, (Nov. 22, 2000), at 10. The FERC Staff = noted=20 that "this correspondence [between falling price caps and increasing = exports]=20 does not necessarily show price caps caused increased exports; however, = other=20 things being equal, lower price caps may provide for greater profits = from=20 exports if conditions outside California lead to high prices and create = greater=20 opportunity costs." Summer 2000 Western Report, supra note 4, at 2-16.=20

n55. December Market Order, supra note = 2, at=20 62,010. "Imbalance market" is an alternative name for = the=20 real-time market. The FERC did not discuss how the = ability to=20 make bilateral transactions at other times might have affected its = reasoning.=20

n56. The FERC Staff also appears to hold a view that the = opportunity costs of most real-time bids equal operating expenses, and = that=20 "with certain important exceptions" bids above marginal cost (expense) = are=20 evidence of market power. The only named exceptions are = generators with limited operating hours due to emission restrictions and = hydro=20 facilities with limited reservoirs, both of which have opportunity cost=20 elements. Staff Recommendation on Prospective Market=20 Monitoring, supra note 47, at 12.

n57. Prajit K. Dutta,=20 Strategies and Games: Theory and Practice 214 (1999). =

n58. Cal=20 ISO August 1998 Report, supra note 22, at 10.

n59. = "These=20 patterns [correlations] track with the energy market = price and=20 reflect the opportunity cost of generators' providing ancillary = services. That=20 is, if generators are holding a portion of their units in reserve, they = are=20 foregoing the revenue they could have earned by providing energy in the=20 day-ahead, hour-ahead, or real-time markets." Summer = 2000=20 Western Report, supra note 4, at 3-17.

n60. Cal ISO = October 1999=20 Report, supra note 22, at 54.

n61. Sheffrin, supra note = 31; Eric=20 Hildebrandt, Further Analyses of the Exercise and Cost Impacts of=20 Market Power in California's Wholesale Energy=20 Market, (submitted by Calif. ISO Dept. of=20 Market Analysis), F.E.R.C. Docket No. EL00-95-012 = (2001).=20

n62. U.S. Dept. of Justice and Fed. Trade Comm'n Merger=20 Guidelines, 57 Fed. Reg. 26,823, at pt. 1.2.

n63. = Instituting an=20 ISO does not lower the risk of discrimination and refusals to zero. ISOs = that=20 take the form of for-profit "transcos" might have corporate links to = generation=20 that make such behaviors profitable. The incentives of those who govern=20 nonprofit ISOs can also produce these outcomes. See generally, Robert J. = Michaels, The Governance of Transmission Operators, 20 Energy L. J. 233 = (1999).=20

n64. Typical questions include whether the differences = between=20 locations are statistically significant (i.e. they did not arise by = chance),=20 whether all relevant variables that might account for these differences = have=20 been examined, and whether techniques appropriate for the data and the = sample=20 have been used.

n65. Over ninety active participants = trade=20 regularly in the WSCC area, a highly liquid market that = accounts for 40% of wholesale trades nationally. A monopolist would = shrink its=20 trading to keep price high, but during summer of 2000, volume rose in = the WSCC.=20 Comments of Duke Energy North America, LLC, San Diego Gas & Elec. = Co. v.=20 Sellers of Energy and Ancillary Serv., F.E.R.C. Docket No. EL00-95-000, = at 30=20 (Nov. 22, 2000).

n66. One exception to the consensus = claims that=20 coordination to raise prices is (was) greatly eased by the protocols of = the=20 California PX and ISO and the public availability of real-time operating = data on=20 individual plants. The ISO has since ceased disseminating that data. = Robert=20 McCullough, Price Spike Tsunami: How Market Power = Soaked=20 California, Pub. Util. Fort., Jan. 1, 2001, at 22 .

n67. = Paul=20 Joskow & Edward Kahn, Identifying the Exercise of = Market=20 Power: Refining the Estimates, Mass. Inst. Tech., (July 5, 2001) = available at=20 http://econ-www.mit.edu/faculty/pjoskow/exercise.pdf; Stephen L. Puller, = Pricing=20 and Firm Conduct in California's Deregulated Electricity=20 Market, Univ. Calif. Energy Inst. PWP-080, (July 2001)=20 available at http://www.ucei.berkeley.edu/ucei/PDF/pwp080.pdf; = McCullough, supra=20 note 66. Harvey & Hogan, supra note 38, offer an extended critique = of the=20 Joskow-Kahn study.

n68. In the months when only = California was=20 under price caps, shipments by California producers to uncontrolled=20 venues outside the state coexisted with=20 out-of-market ISO purchases from these = venues.=20

n69. Harvey & Hogan, supra note 38, at 10.=20

n70. Comments of TURN and UCAN, San Diego Gas & = Elec. Co. v.=20 Sellers of Energy and Ancillary Serv., F.E.R.C. Docket No. EL00-95-000 = at 18-19=20 (Nov. 22, 2000).

n71. Judith B. Cardell et al.,=20 Market Power and Strategic Interaction in Electricity = Networks,=20 19 Resource & Energy Econ. 109 (1997).

n72. Two = generic=20 alternatives are "nodal" pricing with financial transmission rights and=20 "flowgate" pricing with physical transmission rights, respectively = outlined in=20 William W. Hogan, Contract Networks for Electric Power Transmission, 4 = J.=20 Regulatory Econ. 211 (1992); and Hung-po Chao et al, Flow-Based = Transmission=20 Rights and Congestion Management, 13 Elec. J. 38 (Oct. 2000).=20

n73. Cardell et al., supra note 71, at 120. In effect = the=20 generator is able to engage in price discrimination. "Residual" demand = for its=20 power at a location is determined by subtracting other supplies from = total=20 demand there. Strategic exploitation of constrained transmission allows = it to=20 make external supplies to the location smaller in quantity and less=20 price-sensitive. This action makes the residual demand for its power at = that=20 location more inelastic and raises the price it can charge if it owns = generation=20 there.

n74. Duke Stresses Bilateral, Long-Term Deals as=20 California Cure, Megawatt Daily, Jan. 12, 2001, at 1. =

n75.=20 December Market Order, supra note 2, at 35.=20

n76. San Diego Gas & Electric Co. v. Sellers of = Energy and=20 Ancillary Serv., 94 F.E.R.C. <para> 61,245, 61,862 (2001), reh'g = pending,=20 95 F.E.R.C. <para> 61,091 (2001). The proxy price for the previous = month=20 is calculated on a monthly basis from heat rates, fuel costs, and other = data,=20 largely as recommended by Staff. Staff Recommendation on Prospective=20 Market Monitoring, supra note 47.

n77. = Id. at=20 61,864.

n78. Calculated from data in FERC's Order of the = Day for=20 California: Refunds, not Price Caps, Power Markets Wk, = Mar. 19,=20 2001, at 3

n79. San Diego Gas & Electric Co. v. = Sellers of=20 Energy and Ancillary Serv., 95 F.E.R.C. <para> 61,115, 61,355 = (2001),=20 reh'g pending.

n80. Id. at 61,365.

n81. = 95=20 F.E.R.C. <para> 61,115 (Massey, dissenting).

n82. = FERC=20 Nominees Remain in Limbo as Feinstein and Smith Introduce Price Cap = Legislation,=20 Foster Elec. Rpt., May 2, 2001, at 10.

n83. San Diego = Gas &=20 Electric Co. v. Sellers of Energy and Ancillary Serv., 95 FERC = <para>=20 61,418 (2001), reh'g pending.

n84. Id. at 62,548.=20

n85. 95 F.E.R.C. <para> 61,418, at 62,549.=20

n86. Id. at 62,568.

n87. December=20 Market Order, supra note 2, at 62,024, . =

n88.=20 PG&E Joins Other Western Utilities to Oppose FERC Actions, Bus. = Wire, Aug.=20 7, 2001.

n89. Id.

n90. The difficulty = may be=20 greatest during interim periods. The controls discourage stopgap = solutions such=20 as "diesel farms" which only intend to operate until more economical = generation=20 becomes operational. Mark Golden, Small Generators Shut Down as FERC = Price=20 Controls Bite, Dow-Jones Newswires, Jul. 13, 2001.

n91.=20 Utilities in 3 Western States Urge FERC to Change Price Controls, Bus. = Wire,=20 Aug. 7, 2001.

n92. San Diego Gas & Electric Co. v. = Sellers=20 of Energy and Ancillary Serv., 95 F.E.R.C. <para> 61,418, at = 62,556=20 (2000). The Commission also stated that the intervenors had not made a = showing=20 that would justify their request. It is unclear how they could make this = showing=20 without a full relevant market study. =

n93. The=20 FERC Staff's Summer 2000 Western Report, supra note 4, provides useful=20 background, but the Commission's reliance on it in forming mitigation = policy is=20 unknown.

n94. Order No. 642, Revised Filing Requirements = Under=20 Part 33 of the Commission's Regulations Final Rule, III F.E.R.C. Stats. = and=20 Regs. <para> 31,111, 65 Fed. Reg. 70,984 (2000). =

n95. Id.=20 at 31,918. Markets to be examined also include = long-term=20 capacity and ancillary services.

n96. Most commonly the=20 Commission has imposed requirements that the merged entity participate = in an=20 ISO. Prior to the coming of ISOs, parties to a merger were sometimes = required to=20 liberalize access to their post-merger transmission systems.=20

n97. Pacific Gas & Elec. Co., 77 F.E.R.C. = <para>=20 61,265 (1996).

	Home =20 S= ources =20 How=20 Do I? Site=20 Map What's = New Help =20