Comments by John Besant-Jones, The World Bank
Much of the paper is devoted to a review of the history of natural monopoly in the United States, especially the demand for regulation by suppliers to protect their monopolies. But any treatment of natural monopoly – including that given in the paper - rapidly transposes into treatment of other economic concepts such as competition, regulation and economies of scale. Hence this commentary focuses on the treatment in the paper of competition, regulation and economies of scale, as well as natural monopoly. The paper’s application of these concepts is less than rigorous to the case of electricity, and reflects an absence of current thinking on the subject.
The title of the paper asserts that the concept of natural monopoly is a myth and was formalized to justify government intervention (page 43). Rather than being mythological, the true nature of natural monopoly may be an abstract concept whose effects can only be observed indirectly through observable causes and effects of competition, regulation and economies of scale. This feature may motivate the skeptics of natural monopoly’s existence.
The true test of monopoly power is the presence of market power that supports monopoly prices, and a natural monopoly exists where market power is derived through investments that enable an incumbent to keep rivals out of the market but still charge higher than competitively set prices.
I must admit that I don't get it. What is "competitively set prices"? I understand that it is a price charged by competitors – i.e., those who find it profitable to compete with the incumbent. Prior investments are always necessary, whether you set up a lemonade stand or built a power plant, so if would-be competitors decide not to invest in this particular business, I can interpret it only one way – existing prices are already too low to allow for profitable competition (provided, of course, there is no external pressure by the government).
Viewed from this angle, the very notion of "market power" seems misleading and empty.
Market power is not immutable, however, and can be eroded over time through the advent of new technologies (see paras. 0 and 0 below) or changes in consumer tastes, for example. A natural monopoly can be weakened, therefore, and destroyed eventually by market forces (a similar point is made in the paper on page 44 – second para.). But an awful lot of rent can be extracted from consumers and deadweight losses incurred while market power lasts. The paper hardly mentions market power and the dangers from its abuse.
Again, what can be construed as "abuse" of this "market power" – as long as price level is kept low, no competitors arrive; as soon as an incumbent starts charging too much, they are here to undercut his gains.
The indisputable example of a natural monopoly to be found in the electricity sector is the function of co-ordination of an electricity network. This function is critical because electricity cannot be stored economically and electricity flows can become unstable and cause a system breakdown without this control. In other words, system co-ordination is a public good. But control of this co-ordination provides huge market power over entry and use of the network. This power is certainly not a myth, nor can it be successfully challenged by a rival. Note that the issue is not monopoly price for this service, because its cost accounts at most for around one percent of the total cost of electricity supply. The issue is that an electricity producer that controls the network can charge monopoly prices for electrical energy to electricity distributors and users.
Coordination is well too often a means to increase productivity etc. However, it does not mean that it must be provided for in a compulsory way, by the state fiat. One can find coordination everywhere – in the airline, railway and bus schedules, production-wholesale-retail networks, between phone companies across the globe (with no regulating authority). Basically, if coordination is good and productive, participants voluntarily adhere to it and agree to follow necessary rules.
Actually, economic theory points to at least two great example of voluntarily-arranged coordination which involves forfeiting of some freedom of movement by participants in exchange for greater efficiency and access to wider market. The first is the institution of firm – with shareholders, management, and labor voluntarily submitting themselves to a set of complex rules. Another is "organized market", like commodity and stock exchanges as well as interbank payment systems. If these institutions were not available for observation, one could also claim that the only way to get people coordinated is to set up state enterprises and assign people to them in a compulsory manner.
In short, if coordination is good for electricity market, what would prevent electricity generators, consumers, transmitters and distributors to abide by the rules set up by regulators and pay for their services accordingly? Moreover, we will have a (potentially) competititve market for coordination networks, just like in the US there are competing commodity and stock exchanges.
On principle, most economists would agree that competition is preferable to regulation, partly because of the well-known risks of creating sub-optimal incentives under regulation. But this principle needs elaboration which is not found in the paper, firstly about what is meant in practice by competition, and secondly what is meant in practice by regulation. Without this clarification, the paper is free to slide over and around issues and experiences that clash with its general thesis about the myth of natural monopoly.
The paper espouses the theoretical concept of a dynamic, rivalrous process of competition (page 44 – top para.). This concept requires numerous buyers and sellers, perfect information available to all participants, and total (“costless”) freedom of entry and exit from the market. In sum, there is no possibility of abusive market power under this type of competition. This concept has its uses for theoretical analysis, but it has limitations in practice since it is seldom realized in any market, let alone one served by a network industry such as electricity.
The reference to “ruinous competition” on page 48 does not reflect practical considerations for a network system. Networks are particularly vulnerable to local congestion and sabotage of competitors. The lax enforcement of law and order found in many developing countries with poor electricity supply would create ripe conditions for ruinous competition in this sense, resulting in considerable loss of economic welfare.
The paper also attacks regulation in addition to the concept of natural monopoly, because natural monopoly leads to regulation. It is noticeable how the paper finds that the evils of natural monopoly lie mainly in the risks of regulation rather than in the standard concern about abuse of market power. This stance implies a touching faith in the ability of competition to rise and flourish anywhere and under any circumstance if only governments refrained from imposing obstacles. The necessary conditions for such an outcome, however, do not exist in most countries because of their poor investment climates and/or insufficient market size. Even in countries that do offer such conditions and have implemented well-designed electricity markets, establishing competition in the electricity market has been difficult and required repeated adjustments to transaction design.
The paper’s treatment of regulation draws solely on the United States, but it fails to point out that the United States has adopted a cost-of-service approach to regulation for public utilities that has produced well-documented perverse investment incentives – such as “gold-plating” and “after the event” prudence tests. Furthermore, the paper does not refer to the rise of incentive forms of regulation in the UK and elsewhere that cap prices or average revenues under formulae such as the “RPI – X”, and which do not share the disadvantage of the heavily intrusive and data demanding US approach. Incentive regulation can overcome many of the problems noted in the paper with traditional US-style regulation. But incentive regulation is also far from perfect, and more experience is needed with this relatively new form to resolve problems that have arisen.
The paper is also fuzzy on another conceptual point – namely that of the services provided by an electricity supplier. It treats the demand for electricity in the same way as demand for a consumable such as food i.e., as an economic good whose consumption directly enhances the consumer’s welfare. But the consumption of electricity only enhances welfare indirectly by facilitating more efficient supply of services such as lighting, heating and communication. This derived demand for electricity relies on the acquisition by consumers of appliances to enable electricity to penetrate the markets for these services. The significance of this characteristic for competition theory is that there are non-electric means (technologies) that can also supply these services, and their existence provides a contestable form of competition for electricity – albeit weakly in most cases once electricity networks are built. This form of competition is therefore most relevant to serving demand in areas not covered by the main power grid.
New forms of contestability and even direct competition between technologies are now emerging. The paper fails to make this point even in the more obvious in cases of telecoms and cable TV where alternative technologies (wireless mobile phones and satellite TV, respectively) are providing strong competition to some - but not all - of the services and markets served by the established service technologies. In the case of electricity, network supply is facing nascent but increasing competitive pressure from distributed forms of electricity supply located on or near to consumers’ premises. But distributed electricity supply is also being used to complement network supply to overcome bottlenecks in the network. Network electricity supply also faces direct competition from other forms of energy such as natural gas for the space heating market, provided that common ownership of the means of supply of gas to consumers and to gas-fired electricity producers doesn’t transfer the concern about market power to the natural gas market. The contestability provided by these competitors, however, is likely to remain too weak for many years to adequately discipline a network service provider’s market power.
Economies of scale
The paper (pages 44-46) underplays the significance of how economies of scale in electricity production have evolved considerably over the lifetime of the electricity industry. The paper gives the impression that because power supply was somewhat atomistically competitive initially, it was wrong to create monopolies with the advent of large-scale electricity production facilities. In fact, technological developments over time that underlie electricity production explain much of the changes observed in the structure and regulation of the electricity supply industry.
In the formative years of the electricity supply industry up to around 1920, economies of scale in electricity production were relatively minor, and so power stations were small and highly decentralized around the areas of consumption. Electricity transmission was too primitive to transfer electricity over long distances, In the 1920s and 1930s, however, economies of scale developed substantially in electricity generation – both coal-fired power generation with steam-driven turbines and hydropower generation. This development stimulated the advent of large capacity power stations that could serve many towns and other load centers. In turn transmission technology developed to connect these stations to load centers. So this evolution added a high voltage wholesale wires business to the low voltage retail wires business for distributing electricity. The paper refers to street level networks (page 51 – top para.) - the distribution end – but does not mention the high voltage transmission of electricity. Yet it is transmission that is more closely associated with natural monopoly in the electricity supply chain (see paragraph 0 below).
With the development of gas-fired gas turbines and supplies of natural gas from the 1980s, the direction of change has been swinging back towards decentralized generation of electricity, because economies of scale are much lower for gas turbines than for steam turbines fed by boilers. Nascent technologies for generating electricity based on renewable forms of energy (wind, solar, biomass, photovoltaics) also have lower economies of scale than conventional technologies. If the costs of electricity production from these technologies are reduced to levels that are competitive with conventional technologies, then the dominance of the large central power station may be threatened. But this doesn’t mean that transmission networks will necessarily become superfluous, because such systems will still be needed for creating electricity markets supplied from a variety of production technologies.
Natural monopoly applied to electric utilities
According to the paper, natural monopoly theory states that competition cannot persist in the electric utility industry (page 53). No such claim is made nowadays by the practitioners in this field, because this assertion implies that the electricity supply chain is organized in the vertically bundled structure that typified US utilities in the past. Because the electricity supply chain contains both a natural monopoly and a competitive component, this structure is rapidly disappearing around the world as well as in the US with the unbundling of the monopolistic and competitive components of the supply chain so as to expose the latter to competitive pressures. Perhaps this is what is meant in the paper by the dramatic assertion that “…states …. are abandoning the baseless theory of natural monopoly in favor of natural competition” (top of page 54).
The modern approach to the organization of electricity supply holds that the natural monopoly elements of the electricity supply chain should be regulated to facilitate competition in the rest of the chain. Since the natural monopoly elements of this chain are formed only from the “wires” of transmission and distribution, together with system co-ordination, they account for only ten to twenty percent of the total cost of supplying consumers. In other words, eighty to ninety percent of the total supply cost of electricity can in principle be subject to competition. Where such competition is allowed, this feature explains the statements attributed to Primeaux in a modern context (page 53).
One indicator of confusion about the existence of natural monopoly in the electricity supply industry is “The Problem of Excessive Duplication” (page 50). This “problem” is most apparent in the case of high voltage transmission where regulation prevents the construction of a second, parallel line if the existing line has sufficient capacity to meet network requirements. The application of competition theory in the paper seems to envisage transmission of electricity just between a point of production and a point of consumption, because this transaction would be essentially private between two parties. But an electricity transmission network typically connects many points of production with many points of consumption in a complex web of lines and nodal substations. In this manner the essential purpose of the network is to serve as a common carrier of electricity for many producers and consumers. Hence the operation of the network should be regulated to ensure fair access to it by electricity suppliers and users. In addition, the construction of a line within the network would influence the pattern of electricity flow and thus change the pattern of usage of transmission facilities. Because of this feature, the development as well as the operation of electricity transmission facilities has a significant public good element that should be regulated in the public interest.
If the paper can only identify examples of competition confined to large consumers of electricity shopping around for the best deals among wholesale suppliers (page 54), then competition doesn’t have much of a future in the electricity market. This form of competition is the standard first step in introducing competition, but it only yields some static benefits to a relatively few users. To capture dynamic benefits of competition over time for most consumers, it is necessary to set up a genuine exchange with many suppliers and purchasers competing for business, as in the numerous power pools that have emerged and are emerging around the world.
Finally, the paper’s focus on traditional US-style power utilities reduces its relevance to developing countries and countries in transition. The paper’s advocacy of competitive investment implies the presence of unlimited investment capital – clearly not found in these countries. Likewise, the paper’s presumption of a well-developed electricity supply system is not typical of these countries. Furthermore, competition for the market or in the market for electricity is a distant prospect for most of these countries with their poor investment climates and low per-capita incomes. Hence the onus will be on effective regulation for the foreseeable future, and this again is a major challenge for these countries. In practice, de facto monopolies may be around for longer than desirable in theory on account of these limitations.
jbj\March 1, 2003