Lessons from British Electricity Restructuring
by John E. Kwoka Jr.
John E. Kwoka Jr. is a professor of economics at The George
The sweeping restructuring of the British electric power sector that took effect in April 1990 involved several different changesvertical deintegration of generation, transmission, and distribution; horizontal deconcentration of the power generation stage; a power pool to coordinate transactions between generators and customers; entry into generation and new local supply arrangements; and privatization of all entities but one in the industry.
Those changes appear to have accomplished their intended effects. Costs have declined, power continues to be supplied, and the governments role has been vastly reduced. Yet other results of those changes have been unanticipated and more problematic. By some standards, the price drop is less impressive and perhaps even illusory; regulation has had to expand to address new competitive problems. And, contrary to original reforms, various parts of the industry are now undergoing reintegration and consolidation.
Those reforms are not only important to British consumers, but also to Americans. In states from California to Massachusetts, the electric power industry is being restructured. The British reforms are seen as one possible model. Understanding the strengths, weaknesses, and net effects of their experience can facilitate the design of good policy in the United States. While the British experience has some very positive features, it also has encountered significant problems that cautions about the special difficulties of transforming the electric power sector.
A bit of background will prove helpful for evaluating the British reforms. From nationalization in 1947 until 1990, the Central Electricity Generating Board (CEGB) operated all generation and distribution facilities in England and Wales as a vertically integrated statutory monopoly. Fossil fuels contributed more than 70 percent of power generation, with coal accounting for more than two-thirds of that. Coal was guaranteed a minimum level of use in generation at prices well above world levels, contributing, in turn, to excessive electricity prices. Nuclear plants represented 20 percent of generation capacity; hydro sources comprised 4 percent to 5 percent.
Power was supplied to twelve state-owned regional distribution companies, area boards, that in turn distributed power to all customers within their service territories. Area board actions were tightly coordinated with the CEGB, so that, as described by John Vickers and George Yarrow, "the resulting behavior [was] little different from that to be expected from a single fully integrated public corporation."
A number of factors contributed to the United Kingdoms decision to restructure its electric power sector in the 1980s. The Thatcher government disdained public ownership on both ideological and pragmatic grounds. The electric power sector was performing poorly, with productivity lagging the all-manufacturing average. Plant cost overruns were very large and the CEGB was planning yet more construction.
In early 1988, the Government published a White Paper entitled "Privatizing Electricity," detailing the planned restructuring of the sector. With minor modification, the report constituted the core of the Electricity Act adopted by Parliament in July 1989. An Office of Electricity Regulation (Offer) was established shortly thereafter, with Prof. Stephen Littlechild as its Director General.
The transformation of the British electric power sector in April, 1990 proceeded along three paths. First, the traditional industry was dismembered both vertically and horizontally. High-voltage transmission assets were transferred to a new National Grid Company (NGC). Coal and oil fired units were divided among two companies National Power and PowerGen. Nuclear Electric retained control of all nuclear units. The initial plan was to assign the nuclear units to National Power, but that plan foundered in the face of investor concerns about nuclear safety and decommissioning costs. At the outset, National Power had 52 percent of total generating capacity, PowerGen had 33 percent, and Nuclear Power had the remaining 15 percent.
The second set of changes involved ownership. Both National Power and PowerGen became private companies in 1991, whereas the difficulties associated with nuclear power resulted in continued government ownership of all nuclear units. Approximately 30 percent of shares in National Power and PowerGen were sold to the public, an equal amount to foreign and institutional investors. The remaining 40 percent was held by the government until 1995.
The CEGBs other creationNational Gridwas also privatized. All twelve state-owned area boards were privatized and became Regional Electric Companies (RECs). Until 1995, the government held a so-called "golden share," a single equity share with the right to prevent acquisitions involving the RECs without government approval. Each new company performed both the natural monopoly distribution function as well as "supply"the potentially competitive tasks of delivery, metering, and billing. Concerns over the need for coordination and for additional supply led to two compromises with the objective of vertical deintegration. The RECs were allowed to generate power up to a maximum of 15 percent of their own requirements, and the twelve RECs collectively owned the National Grid.
The third set of changes sought to open the system to competition, where possible, while continuing necessary regulation. Vertical and horizontal restructuring of power generation was based on the assumption that generation had become workably competitive and would become increasingly so with new market entrants and with the advent of small-scale combined cycle gas turbine (CCGT) technology. Thus, even if there were few incumbents, their pricing might nonetheless be constrained. As if to underscore that prospect, the first new generator, Lakeland Power, was granted its license within days of the inauguration of restructuring and began operation of its 230 MW CCGT station a year and a half later.
Both distribution by the RECs and transmission by the National Grid Company were treated as natural monopolies and subject to price cap regulation by Offer. For the first three years, NGCs cap was set at the inflation ratemeasured by the Retail Price Index or RPIand later reset at RPI minus a 3 percent offset for the four years commencing March 1993. The tightening of the cap was the result of high profits during the first price cap cycle. Notably, the cap applies to NGCs revenue per kW of maximum demand, even though its revenue derives from actual transmission services priced according to a use-of-system tariff. As a consequence, the capped price may not bear a close relationship to actual service price at any point in time (although it is adjusted periodically).
The twelve RECs are individually price capped. In addition, each has a separate cap for its distribution operation and its supply function. All of the latter were initially subject to a limitation of RPI, later revised to RPI minus 2 percent, together with direct pass-through of certain cost items. By contrast, the offsets for distribution varied among the RECs, and initially ranged from zero to a positive 2.5 percent, the latter permitting actual real price increases over time. Those generous offsets were intended to generate capital for improvements in the distribution networks. The variation among RECs was based on the particular capital-expenditure requirements and load growth of each. Because of the excessive profitability that resulted, a one-time adjustment was imposed and the cap was reset at a uniform RPI minus 2 percent for all RECs in 1994.
The distinction between "distribution" and "supply" is of growing importance over time. Initially, the RECs were granted monopoly franchises over all customers with peak loads less than 1 MW. That is, all such customers had to purchase powerat regulated ratesfrom their local REC. The threshold freed only about five thousand customers throughout Britain, although those represented 30 percent of total load. Of the large customers, those with loads in excess of 10 MW typically had to contract for power directly with any of the three generating, companies a distant REC, an independent power producer, an independent supply broker who resells power purchased off the grid, or even a supplier in Scotland or France that is interconnected to the grid. Users of intermediate loads could either purchase at regulated rates or contract for power, at their option. The local REC was, and is, required to distribute any contracted power to the end user on a nondiscriminatory basis.
As of April 1994, the threshold defining franchise customersthose without access to the unregulated marketwas lowered from 1 MW to 100kW. In 1998, the threshold will be abolished entirely, at which point the RECs will have no captive customers whatsoever and full retail competition will occur. At that time, the RECs will continue to be the last resort supplier to small customers, an obligation they may satisfy simply by reselling power at the pool price.
The Power Pool
Deintegration of the CEGB created three companies that generate power plus one other providing transmission service. Traditional vertical economies, such as least-cost dispatch and maintenance of system reliability, were left to an entirely new institutionthe Electricity Pool of England and Wales. That power pool is operated by the National Grid separately from its transmission function. All generators over 50 MW, as well as smaller generators that sell directly to customers, are required to be members of the power pool. Most power is therefore sold through the pool.
Operation of the pool is an exercise in textbook economics. Each generating company offers to supply power from each of its generating units for half-hour increments during the next day, at company-chosen prices. Those bids are aggregated in merit (that is, least-cost) order into an upward sloping supply curve, examples of which are shown in Figure 1. In principle, dispatch is carried out by the pool calling on plants in order up to the point at which demand is satisfied. The system marginal price or SMP is the offer price of the highest-cost plant called upon to supply power in each half-hour period and is therefore intended to closely resemble the short-run marginal cost of meeting demand. All generating units actually called upon for power production receive SMP regardless of their offer price.
The full price at which power is purchased from generatorsthe "pool purchase price"consists of SMP plus a capacity payment. The latter is designed to ensure adequate declared capacity at all times by creating a price that better reflects long-run marginal cost. In practice, the capacity payment is the product of two factorsthe loss-of-load probability and the value of lost load. The loss-of-load probability (LOLP) is an engineering determination of the probability of a system voltage reduction due to inadequate available capacity. The value of lost load (VOLL) is a per-kilowatt hour (kwh) measure of the social cost of failure to meet demand, derived from utility planning models and subject to adjustment by Offer. Each kwh of generating capacity that is available during a half hour period receives a capacity payment regardless of whether or not it is actually called upon. At the outset of reforms, the capacity payment represented only about one to two percent of the total pool purchase price.
The pool selling price differs from the purchase price by a so-called "uplift." The uplift is the per-kwh sum of all capacity payments made for declared but unused capacity, transmission losses, ancillary services (spinning reserves, reactive power), pool operating costs, and out-of-merit running, for each half-hour period. Samples of pool purchase and selling prices are given in Table 1.
|Table 1: Electricity Pool Prices (#/Mwh)|
|System Marginal Price (SMP)||17.37||19.5||22.64||24.16||20.7|
|Pool Purchase Price (PPP)||17.42||20.82||22.8||24.44||24.0|
|Pool Selling Price (PSP)||18.34||22.43||24.19||26.2||26.38|
|Source: Adapted from S. Stevens, "Privatization of the Electric Power Industry of the United Kingdom," Energy Studies Review, 1995.|
As a hedge against volatile and unpredictable prices, generators and customers can enter into options contracts, known as "contracts for differences" or CfDs. In a typical CfD, an REC and a generator agree on a fixed pricethe "strike" pricefor a specified quantity of power at a designated future time. At that time, the contract is consummated by payment to the relevant party of any difference (positive or negative) between the strike price and the actual pool price that occurs. Under that system, while generators continue to supply power to the pool, the effective price of power to buyers and users becomes the strike price. The pool price serves to control dispatch and minimize the total costs of meeting demand, but it is decoupled from the contract price of the transaction.
In principle, merit-order dispatch together with pool and strike prices might constitute the foundation of a well-functioning power market. In practice, however, the process is greatly complicatedperhaps fundamentally compromisedby transmission constraints. The essential problem arises since power generation and consumption have a geographic dimension as well as time and quantity dimensions. It is often impossible to dispatch the globally least-cost generating unit, since transmission capacity constraints may preclude use of that power to satisfy requirements elsewhere in the system. Instead, some other non-least-cost (that is, least generation cost) unit may be the optimal source of power. Such "out-of-merit running" is simply a recognition that there are, at least temporarily, geographically distinct power markets rather than a unified whole.
The British system addresses those contingencies with administrative actions of the pool. Some units are taken out of service despite lower generation cost. Those units are termed "constrained-off" and are compensated by the difference between the pool price and their bid price. Other units required to supply a region despite higher cost ("constrained-on") are compensated at their bid price on the theory that the bid price represents their actual marginal cost. The excess costs resulting from those two types of compensation are recouped through the uplift, as described earlier. The problems caused by such transmission constraints are discussed after a review the sectors performance under reform.
Given the magnitude of the changes to the British electric power sector, a threshold question is how well the new system has performed. Of course, the original reforms mandated certain changes in industry structure, but over the past several years changes have continued. In the generation sector, both fossil-fuel generators have experienced declines in their market shares, with National Powers share falling substantially. More remarkable is the renaissance of Nuclear Electric, which is 50 percent larger than in 1990. As shown in Table 2, the other major net gainers have been new entrants and distant suppliers through the interconnectors. Relying predominately upon CCGT technology, entrants have planned capacity additions equal to more than 15 percent of the British market. Nonetheless, it should be noted that the generation industry remains highly concentrated, with three firms accounting for 82 percent of output.
|Table 2: Generator Market Shares|
|Interconnectors and Pumped Storage||4.82||7.7||8.4||8.7||8.4||9.2|
Change has occurred at the distribution stage as well. While RECs provided supply to 57 percent of their local customers with load greater than 1 MW in 1990, that percentage fell to 43 percent within three years. REC supply to distant large customers increased from 4 to 16 percent in the same time period. Those facts suggest gradually increasing competition at the distribution and supply end of the market.
Perhaps the most impressive result of restructuring has been the enormous gains in labor productivity in the traditional generation sector. Whereas CEGB productivity gains averaged about 3.5 percent per year, Table 3 shows that the three generating companies created out of CEGB have achieved productivity gains on the order of 25 to 30 percent per year since reform. The largest increase has been recorded by state-owned Nuclear Power. In four years, Nuclear Power has increased output by nearly 50 percent while reducing staff by the same percentage, literally doubling its output per employee. Its plants ran at just 59 percent of capacity prior to reforms, but at 71 percent in 1995. Since it has remained publicly owned, that experience strongly suggests the power of competitive pressures on enterprisesregardless of ownershipto improve efficiency.
|Table 3: Labor Productivity in CEGB Successor Companies
|Source: Vickers, 1994.|
The effect of reform on prices is all-important. Table 4 shows that tariffed customers have seen prices fall by an average of 12 percent between 1990 and 1997, a number that ranges from about 6 percent to three times that among the RECs. While the decline appears favorable, some critics suggest that there is less to the declines than meets the eye. Yarrow, for example, has noted that the 1990 benchmark price is distorted by a price increase of more than 3 percent in 1988-89. The increase was apparently intended to improve the prospects for privatization. In addition, the costs of coal and gas declined precipitously during the 1990s, suggesting that in put cost declines rather than efficiency gains account for price drops. Using the 1987-88 price as a benchmark and controlling for fuel cost decreases, he concludes that prices actually increased in the first four yearsresidential price by 13 percent and industrial price by 11 percent, and even more if efficiency gains are factored. As further support, Yarrow compares declines in U.K. electricity prices with those in neighboring EU countries during the 1990s, and finds the U.K. experience inferior.
|Table 4: Standard Domestic Tariff|
|Source: U.K. Office of Electricity Regulation (OFFER), 1996.|
For nontariff, that is, unregulated customers, price reductions in the 1 MW market have averaged about 15 percent. Even here, the larger customers have not benefited nearly as much as anticipated. In fact, some of the very largest customers, who have no access to tariff and whose contract prices are often tied to the pool price, (and who previously had enjoyed some subsidies) have experienced outright price increases. In short, the effects on residential and industrial prices actually attributable to reform remain a critical area of continuing debate.
Further controversy surrounds the pool price. Table 5 shows that the pool purchase and selling prices have both increased by 40 to 45 percent since the inception of reform. The uplift, which represents the difference between the two, has risen 250 percent. The capacity payment constitutes 20 percent of the pool price. Both the uplift and the capacity payment componentsand hence the pool selling price itselffluctuate over the course of the day and the year as intended, but they have also exhibited some wild gyrations. In one extreme episode in late 1991, the pool pricetypically about 2p/kwhpeaked at 33p, sixteen times its initial value.
|Table 5: Average Demand Weighted Pool Prices by Component (p/kWh)|
|System marginal price||1.81||1.99||2.31||2.59|
|Pool purchase price||1.82||2.16||2.33||2.63|
|Pool selling price||1.92||2.34||2.48||2.86|
Finally, privatization and deregulation (or looser regulation) of the electric power sector have resulted in substantial profits for most segments of the industry. As shown in Table 6, supply is the least profitable part of the business, although the effects of Nuclear Electrics recovery are evident. By any measure, NGC and the RECs have enjoyed enormous profitability, each about 30 percent of revenue and far above pre-reform levels. Exacerbated by the fact that some of the profits have been used for large executive bonuses, REC profitability has become a very controversial outcome of reform.
|Table 6: Prefix Profits and Revenue of the Power System in England and Wales - 1992/1993 (#million)|
*The companies had small profits or losses
Source: M. Armstrong, S. Cowan, and J. Vickers, Regulatory Reform: Economic Analysis and British Experience, 1994.
Issues and Problems
Operation of the restructured British electric power industry has exposed a number of defects in the design of the system. Some of the defects have been corrected, but others remain unresolved and some may even be inherent in the model.
Perhaps the most serious initial defect in restructuring was the creation of only two private generating companieswhich grew to three when Nuclear Power was added. There are over 75 generating facilities in England and Wales, and it would have been both possible and prudent to create a significantly larger number of companies. Two factors appear to have played a role in the initial decision. The first was a belief that highly aggressive bidding between similar sellers would characterize the market, obviating the need for multiple companies. Second, from the vantage point of the government, "successful" privatization meant ensuring interested buyers, which in turn required assured profitability for the generators even at the expense of long-run competition.
Worse still, for many purposes, market operation was and is almost completely determined by the fossil-fuel duopoly. Nuclear Electrics plants provide the base load, that is, the electricity needed to meet minimal power demand even during off-peak hours. Therefore they bid in a manner that ensures their continuous operation. The mid-merit-order units that establish pool price most of the time, belong to National Power and PowerGen. It is that duopoly that has caused a host of further distortions and difficulties. Some of the problems have been straightforward exercises in market power. For example, in early 1992, system marginal price (SMP) rose by about 25 percent. Offer investigated and concluded that the rise was due to mutually reinforcing bid price increases by National Power and PowerGen.
Somewhat less obvious has been some deliberate gaming of the pool price system by the generators. Several techniques have been employed. One relies upon the fact that with only two generating companies, each knows the likely costs and hence offer prices of its rival and therefore, over time, can construct a good approximation to the pool supply curve. Knowledge of demand then allows each to anticipate which generating facility is likely to be the marginal unit. The bidder that believes his unit will be the marginal operating facility has an incentive to raise his offer price on that unit to just short of the next higher bid. That action unambiguously increases the likely pool price without altering the rank order.
Another strategy has targeted the capacity payment. In the late 1991 episode just mentioned, PowerGen declared certain plants unavailable. That declaration raised the loss-of-load probability and thus the capacity payment. On the actual day, however, PowerGen redeclared the capacity as available, making it eligible to receive the increased compensation. To prevent such practices, Offer imposed requirements of more advanced capacity declarations and employed auditors to monitor the actual availability of generators facilities.
The generators have also devised methods of strategically manipulating the uplift. Recognizing likely transmission constraints, they can anticipate, certain of their units becoming constrained-on, that is, required by the grid to operate out of merit order to ensure adequate supply in a region. Since constrained-on units are paid their bid prices rather than SMP, the generators often have offered such units at very high bids, ensuring a lucrative return when the grid instructs them to produce. In an analogous fashion, constrained-off plants collect the difference between SMP and their bid prices, inducing generators to bid those units at zero.
Offer has had to closely monitor the practices of National Power and PowerGen and at one point even threatened to seek action from the Monopolies and Mergers Commission. To head off that action, the generators agreed to two conditions. The first requires them to divest about 15 percent of their capacity within two years, an amount that could form the basis for an additional sizeable generator. Second, they agreed to bid their facilities into the pool for two years at a level that represents an average seven percent price reduction. Many observers found the capping of pool prices at odds with the deregulatory principles of reform. More generally, the capping of pool prices seemed to imply that Offer held an increasing skeptical view of the pool itself.
Treatment of out-of-merit running represents another major problem confronting British reforms. The frequency with which out-of-merit running occurs implies that the premise of the original designa unified power poolis often violated. That is not necessarily surprising, since transmission constraints are commonly believed to motivate vertical integration in the electric power industry. By contrast, the British system seeks to resolve vertical interdependencies, even in the face of constantly changing transmission conditions, through pricing and administrative actions by a separate pool.
Two specific problems have arisen in the British system. The first involves the method of cost recovery. Clearly, constrained operation entails costs to the pool, but under the present system, those costs are combined into a uniform uplift charged to all users. An average surcharge does not confront each user with prices that reflect the particular costs that it imposes on the system. Rather, high-cost users are undercharged and low-cost users overcharged. Such charges also fail to provide incentives to sellers to develop alternatives to circumvent transmission constraints. Instead, it encourages inefficient dispatch and uses of the grid.
A further and longer-run problem lies in the division of tasks between NGC and the pool. The National Grid originally had responsibility only for the fixed costs of the grid, the pool, through its control of out-of-merit running and the resulting payments, measures the cost of system constraints. If the NGC invested in the grid to ease those constraints, the benefits would accrue to pool users rather than to the grid. Thus the grid has lacked incentives to contend with transmission constraints. In response, an uplift management system was implemented under which the NGC is allowed to keep 20 percent of congestion cost savings. How successfully that allowance addresses the relationship between long-run investment planning and short-run pricing of transmission services is unclear.
A number of issues have arisen at the distribution stage of operations as well. One major concern involves the degree of competition that exists versus the degree of regulation that may be necessary. At the outset of reforms, the RECs were granted a substantial base of franchise customers who could be charged prices subject to very loose cap regulation. The subsequent conversion of those customers to totally unregulated status has proceeded slowly. Only about 20 percent of the users who initially could choose between tariff and unregulated contracting in fact chose the latter, although that may have been partly due to pool price manipulation by the generation duopoly.
REC tariffs were initially only loosely capped in order to generate the capital required to finance improvements in their facilities. While that objective was clearly achieved, the high profits earned by the RECs resulted in enormous public controversy and the imposition of two price reductions followed by tighter caps by Offer. Another concern is that cap rules permit the RECs to pass all the costs of purchased powerthe vast majority of their costsdirectly to the retail price. Direct pass-through generally weakens incentives for cost-minimization, a concern enhanced in the present context where the RECs generate a nontrivial fraction of their own requirements.
The very high profitability earned by the RECs have made them attractive targets for takeover. Just prior to expiration of the governments "golden share," Trafalgar House bid for one RECNorthern Electricwhich responded by offering its shareholders a substantial cash distribution if they would reject the bid. The offered amounts were so large, however, as to fuel further controversy about excess profits at the RECs. Subsequently, there have been other bids for RECs by both UK and foreign investors, notably from the United States.
Some on-going realignments within the British electric power sector raise questions about both the premises of reform and their future. Significantly, both private generating companies sought to acquire an RECMidlands and Southern Electric by PowerGen and National Power, respectively. Such vertical acquisitions reopened debate about the key competitive concerns that motivated original restructuring. In the National Power-Southern Electric matter, Offer Director Littlechild argued that the merged entity would have an incentive to favor its own divisions for supply and purchase, adversely affecting both customers and competitors and significantly complicating the task of monitoring discrimination and cross-subsidization. His objections failed to convince the Monopolies and Merger Commission, but the commissions approval was later overturned by the Minister of Trade and Industry. This has at least temporarily prevented the emergence of a British electricity industry consisting of large and vertically integrated companies.
Another curious development concerns joint REC ownership of the National Grid. While the arrangement might have aided vertical coordination, in early 1996 the RECs chose simply to sell off their ownership in NGC. The decision was in part due to the market value of the latter, but it certainly belies the view that such coordination economies are too important to be jeopardized by vertical deintegration.
Also in early 1996, the Government split Nuclear Electric into British Energy, consisting of modern nuclear facilities, and Magnox Electric, made up of older plants nearing the end of their useful lives. British Energy was successfully privatized, thus creating a third private generating company of significant size.
The restructuring of the British electric power industry has been premised on the ability of private ownership, markets, and competition, to perform tasks previously addressed by other institutions. While in some sense the experiment has worked, it is more accurate to say the experiment has been made to work. The enormity of the initial changes has been nearly matched by the enormity of the problems that have arisen since.
Among the problems that appear to be the result of initial design defects are the following:
(1) The market power of generators. This has resulted in unwarranted price increases and persistent manipulation of other aspects of the system.
(2) Loose regulation of the distribution stage. The initial specification of price caps allowed for substantial price and profit increases at the expense of customers.
(3) More recently, the threat of reintegration of portions of the industry. Efforts to merge vertical stages of production raise competitive concerns and move the industry into unexpected directions.
In addition, there are problems that might possibly have been ameliorated by alternative strategies, but it is not clear that they could have been fully resolved. Principal among those problems are transmission constraints. Neither the NGC nor the pool has handled them very satisfactorily, resulting in both short-run and long-run distortions in the system. Strikingly, those are precisely conventionally handled by vertical integration, suggesting that the British model does not represent a fully satisfactory alternative to integration.
Thus, despite some operational success, British restructuring of electric power has encountered a number of significant difficulties, including challenges to some of its premises, the inadequacy of key institutions, and the ironic need for rescue by regulation. Some of the problems have been successfully addressed, but others have not. Some issues clearly resulted from poor initial design, but others may well be inherent difficulties presented by electricity reforms, difficulties that go beyond those found in other industries.
Those issues represent serious continuing challenges in Britain and equally serious concerns as restructuring gains momentum in the United States.
Regulation, 1997, Vol. 20, No. 3. Published four times a year by the Cato Institute. Copyright 1997 Cato Institute. Regulation is available on the World Wide Web and readers may subscribe on-line. Subscription rates are $18 per year for individuals and $28 per year for libraries and institutions. Please send changes of address and subscription correspondence to: Circulation Department, Regulation, 1000 Massachusetts Avenue N.W., Washington, D.C. 20001, or call 202-842-0200.
ISSN 0147-0590 INTERNET firstname.lastname@example.org WEB SITE http://www.cato.org