How Ownership Structure of Electric Utilities Effects Their Efficiency

Central European University

Department of economics

‘Economics of regulation’ course paper

How ownership structure of electric utilities effects their efficiency?

Professor: Andrzej Baniak

Student: Willen Lipatov

Budapest, 2001

CONTENT

Introduction 3

Historical outline 3

Theories of private versus public enterprise 5

Empirical studies of differences in performance of public and private electric utilities 8

Illustrations for pricing behavior, allocative and productive efficiency 8

Other studies of efficiency of private and public electric utilities 9

A closer look at empirical studies: different approaches 11

Conclusion 14

Introduction

The problem of efficient functioning of natural monopolies in general, and electric industry in particular has been an object of great interest for economists already centuries ago. The importance of this problem is difficult to overestimate, since distortions caused by monopolies lead to substantial economic inefficiencies and consequently to the undermining of economic development. This paper considers one aspect of the problem, namely how efficiency of electric utilities is effected by ownership structure. Recently interest to the ownership structure of natural monopolies has been given a new impetus, especially after the series of successful privatizations in Great Britain.

Firstly, we will look at how views on ownership structure of utilities in general changed in the historical context. Secondly, some theories of enterprise that shaped opinions about natural monopolies will be surveyed. Next we’ll turn to empirical studies of electric utilities, which aim at comparing performance of private versus public utilities. Conclusion will stress the importance of regulatory methods in interpreting results of these comparative studies.

Historical outline

In what follows in this section we follow Newberry (2000, chapter 3). Firstly, one must admit that public ownership is not rare among the network utilities nowadays. But its existence depends mostly on how widely it is spread geographically, as well as on whether property owners or mass interests are more influential in the government. The other factor is importance of a particular utility on local or national level.

Historically utilities started to develop on local level, and already at that time there was a choice whether to own them privately or publicly. The decision depended on cost-benefit relation of introducing certain kind of ownership. This relation included price of service, its quality, and possible dividends of municipal companies. These factors, in turn, were determined by how large a municipality was and how many areas of economic activity were covered by it. Successful experience of other municipalities also mattered, as well as availability of utility managers. And certainly ability of utility customers as opposed to owners to influence local politics played an important role.

One common way to deal with local natural monopolies was offer of limited life franchises. A private firm could get a right to operate as a natural monopoly if it proposed the best (among rivals) conditions of operating a utility. Advantages of franchises are usually considered to be ability of a firm to return its investment and of municipality to reduce costs meanwhile. At the same time, high likelihood of being denied prolongation of franchise makes incentive for the firms to reduce quality or supply of services at some point of time before the expiration of the contract.

Nationwide network utilities first emerged only as private firms. Great Britain failed to regulate such enterprises, relying only on franchise competition. As a result, it experienced big difficulties caused by duplication of services, lack of coordination between companies, and cartelization of industries. The United States were more successful in accommodation of new trends, but here performance of network utilities largely depended on industry and even macroeconomic dynamics. Apart from it, firms had large profits, ownership was mixed, and the government had to implement profit regulation.

In 60s there could be observed a shift in attaining higher weight to consumers’ surplus as opposed to profits. In 70s utilitarian approach prevailed. Both paradigms pointed on ability of public capital to generate faster economic growth, inspired by Soviet capital accumulation. But 80s brought a lot of doubts on this matter. The point was that government has objectives that differ from those of the voters. As a consequence public sector does not generate adequate investment, lowering savings rate instead. Current situation is characterized by the fact that only a very limited number of countries has sound system of regulation of privately-owned natural monopolies. A wider circle of states has some mixture of private and public property in network utilities. Finally, less developed countries have no other option except for public ownership because of loosely-defined private property rights.

Evidence in support of one or other theory is difficult to rely on, as the data (gathered by the World Bank in 1978-91) includes not only natural monopolies, but all state-owned enterprises. However, absence of correlation between per cent of public enterprises in GDP and real GDP per head points on lack of convincing argument for private enterprises to be superior in terms of efficiency to the public ones. On the other hand, larger share of public enterprises in Africa and Asia than that in Latin America is consistent with the view that less developed countries often simply have no other choice than public ownership.

Another interesting observation is that public enterprises have larger share in investment than in GDP, but not only because they are more capital intensive, also because of underpricing of output. Electricity, in particular, had a very poor performance in middle-income countries. Prices were set below long-run marginal costs despite existing excess demand, so that only 60% of costs were covered by revenue. Investments were financed out of taxes that led to huge inefficiencies. In general, public utilities earned low return and required high transfers. In order to improve situations some steps were made, but mostly by lowering investment rather than raising prices and revenues.

Theories of private versus public enterprise

As historical outlook does not provide non-arbitrary conclusions about relative performance of public and private network utilities, although it serves as a useful tool to become familiar with the problem, let’s switch our attention to some theoretical models. Our speculations will follow Viscusi, Vernon and Harrington (1995, chapter 14) in this section.

The basic idea of modeling behavior of a firm-natural monopoly is how to induce this firm to price its output efficiently (first, second-best or close to it). Two ways to do it are establishment of a regulatory agency and franchise auction. First one is not ideal, since such an agency does not have full information about the firm’s actions, that leads, for example, to Averch-Johnson effect in case of rate-of return regulation. Second has the same disadvantages, although maybe to less extent.

Such imperfections give room to existence of public enterprise. The striking difference of it as compared with a private one is that the public enterprise does not have to maximize profit. Instead, it is supposed to maximize social welfare. However, because of the agent-principal problem, the public firm maximizes manager’s utility subject to constraints imposed by government.

Similar problems certainly arise in a private enterprise. Usual methods to solve them are to reward manager with shares of the company and to pay bonuses based on performance of the enterprise. Reputation of the manager on the labor market also plays an important role, but the most effective incentive for the manager is considered to be a possibility to be fired. This threat becomes credible only with existence of the capital market. Unfortunately, it is absent in case of public enterprise.

Moreover, social welfare is much more difficult to measure than profit. This gives the manager of public company undesirable from the point of view of society freedom. Indicators used in assessing performance of the enterprise always cause some distortions, since managers achieve high values of these indicators undermining other characteristics of production.

Sam Peltzman (quoted in Viscusi et al 1995, p. 460) suggested another approach to modeling behavior of managers of public firms. According to him, manager maximizes political support. Two ways through which it can be influenced are price (the lower - the better) and tax burden (or subsidy necessary to run the business with given price). As subsidy is equal to profit with negative sign the problem of manager consists in maximizing profit with given price or minimizing price with given profit (zero). If price is set to the level of average costs these problems are equivalent and bring about the second-best solution.

This setup implies at least three results. First, public enterprise will set prices below those that maximize profit, unless manager attaches no importance to minimization of price, possibility that we are ruling out. Second, with possibility of discrimination price for non-voters will be set on profit-maximizing level, and for voters it will be minimized. Third, discrimination among customers-voters will be less than in case of private enterprise, since discrimination alienates voters who are worse-off with its introduction.

The problem of higher degree of discretion possessed by manager of public in comparison with that of private enterprise still stays. It has its deepest roots in non-transferability of public ownership and hence impossibility to impose the restrictions on the manager that otherwise would be provided by the capital market and threat of owner change. And all kinds of difficulties, as inefficiently high wages, more simple pricing schemes and overcapitalization, stem from this problem.

All in all, theory predicts that public enterprise will be less efficient than unregulated private enterprise, since it maximizes political support rather than profit, does not have the capital market constraint. Moreover, even regulated private enterprise (that has A-J inefficiency) is likely to be more efficient than a public enterprise, because behavior of managers leads to the similar overcapitalization, but also causes other inefficiencies.

Empirical studies of differences in performance of public and private electric utilities

Illustrations for pricing behavior, allocative and productive efficiency

To begin with we give examples which illustrate previous theoretical speculations, and also due to Viscusi at al (1995, pp. 463-67). In study using data from 1912-37, George Stigler and Claire Friedland looked for differences in prices of regulated versus unregulated private electric utilities. So, the null hypothesis was actually that regulatory constraint is not binding. The hypothesis was not rejected, and explanation for this was high information advantage of a regulated firm, so the regulatory agency sometimes could only accept a proposed tariff.

A study by Thomas Moore using data from 1962 supports the hypothesis that public enterprises charge lower price than private electric utilities. It claims that regulation had an effect of 5-6% reduction of price, whereas public ownership – 10-22% reduction. Similar results were got by Peltzman in 1971. But he explained it by special status of public utilities that allowed them not to pay taxes. He also found support for the view that public enterprises discriminate less. Namely, private utilities had 8391 rate schedules, whereas public ones – only 5290.

The fact that public enterprises charge lower prices does not necessarily mean that they reach higher allocative efficiency. The thing is that due to finer price discrimination private utilities may achieve lower amounts of deadweight loss. Peltzman’s evidence shows that private utilities were selling more electricity per customer on average, so it is possible that they did not bring about lower allocative efficiency.

Productive efficiency was studied in 1970 study by Moore. He tested for overcapitalization by looking at the ratio of capacity to peak demand. As public electric utilities had larger capacity for a given peak demand, the view that public utilities tend to overcapitalize more seems to be consistent. Thomas Dilorenzo and Ralph Robinson 1982 study considered labor productivity as a measure of technical efficiency. They found that difference between publicly and privately owned electric utilities is statistically insignificant. Donn Pescatrice and John Trapani, however, estimated average costs for 33 private and 23 public utilities and found that they were higher for public utilities 23,5% in 1965 and 32,9% in 1970.

Other studies of efficiency of private and public electric utilities

The next set of evidences comprises a larger number of studies. All of them are taken from Newbery (2000, chapter 3). To make things clear from the very beginning, Newbery gives a definition of the tern ‘owner’ in a sense it is used with respect to natural monopolies: ‘Owner – the authority with the residual rights, that is those rights that are not subject to contract or control’. Choosing between regulating privately owned network utility and directly controlling publicly owned utility, the government faces trade-off between lowering transaction costs and lowering incentives to work.

The major difference in relations of government with public and private enterprises is the degree of commitment to its obligations. In case of private utility it is normally well specified regulation setup, that is regularly revised. With public utilities, to the opposite, any action can be negotiated and obligations can be changed virtually at any time. The other things that is supposed to make public enterprises less efficient is less pressure to set right prices. This pattern appeared in phenomena of underpricing of capital and rent dissipation. The first one is possible because of access of utilities to debt financing or tax exemptions, the other is caused by average pricing when average costs are below long-run marginal costs. In hydroelectric industry, for example, both phenomena were present.

There is a difficulty in empirical evaluation of ownership changes, since usually such changes go in line with a broader set of reform, e.g. change of regulation method. Intercountry comparisons are also difficult because technology varies from country to country, and so does quality of regulation, as well as management. Assessments within a single country are not promising, since co-existence of two forms of ownership means that they yield approximately the same social benefits. Still, empirical studies of all kinds can be helpful if one does not forget about the limitations of their implications. Further there are summarized some of them.

In the United States regulation was and still is rate-of-return, so comparisons are drown between publicly owned and regulated according to the rate of return enterprises. Backer in 1913-15 considered 300 private and municipal electric firms, and found that municipal ones sold at lower price in 1910-11, and also had lower capital and operating costs. They also had 40% higher sales. Partly it can be explained by the fact that they were better placed, got free managerial services and successfully used economy of scale. These explanations seem to be plausible, since by 1925-26 both sectors were getting the same average revenue.