Optimal Regulation: The Economic Theory of Natural Monopoly

by Kenneth E. Train

Optimal Regulation addresses the central issue of regulatory economics - how to regulate firms in a way that induces them to produce and price "optimally." It synthesizes the major findings of an extensive theoretical literature on what constitutes optimality in various situations and which regulatory mechanisms can be used to achieve it. It is the first text to provide a unified, modern, and nontechnical treatment of the field. The book includes models for regulating optimal output, tariffs, and surplus subsidy schemes, and presents all of the material graphically, with clear explanations of often highly technical topics. Topics include
The cost structure of natural monopoly (economies of scale and scope) • Characterization of firstand second-best optimality • Surplus subsidy schemes for attaining first-best optimality • Ramsey prices and the Vogelsang-Finsinger mechanism for attaining them • Time-ofuse (TOU) prices and Riordan's mechanisms for attaining the optimal TOU prices' Multipart and self-selecting tariffs, and Sibley's method for using self-selecting tariffs to achieve optimality • The Averch-Johnson model of how rate-of-return regulation induces inefficiencies • Analysis of regulation based on the firm's return on Output, costs, or sales • Price-cap regulation • Regulatory treatment of uncertainty and its impact on the firm's behavior • Methods of attaining optimality without direct regulation (contestability, auctioning the monopoly franchise.)

Genres: Economics

Hardcover

First published August 12, 1991

Reviews

[Nick Crowley · Rating 4 out of 5]

I read through Train’s Optimal Regulation. I read the first 200 pages fairly closely, as all of that material was fairly new to me. The final 100 pages covered tariff design, which I have some first hand experience with, so I did not read that part as closely. Overall, I think the book is a great resource.

The book begins with a definition of natural monopolies. A natural monopoly exists when an industry is better served by one firm than by many (a phenomenon known as “subadditivity”). The key feature of a natural monopoly is that average cost declines as output increases. In other words, the company exhibits economies of scale. The problem with this situation from a regulatory standpoint is that pricing output at marginal cost fails to cover the total cost to serve. If the regulator forces the natural monopoly to price at marginal cost, the firm will go bankrupt. Much of the book focuses on how to deal with this problem.

Before addressing solutions to the marginal cost problem, Train covers some of the historically important economics behind monopoly regulation. Most prominent among the historical literature is the Averch-Johnson effect, which states that a firm regulated under rate of return regulation has an incentive to over-allocate capital in its production, with respect to other factors of production (eg, labor), as long as the regulator sets the firm’s rate of return above the true cost of capital. The reason is simple: as long as the firm’s rate of return on rate base (capital infrastructure) yields some margin of profit, it will profit from increasing that rate base. Despite the production inefficiency this yields, a firm will not indulge in “pure waste” by purchasing useless capital. It will only employ productive capital. It is worth noting that when the rate of return is set by the regulator to be below the cost of capital, the firm will employ no capital (and ultimately produce no output). Think about it: if cost of capital > rate of return, the firm loses money by employing capital. Over time, the firm would go bankrupt under this paradigm. It’s a sticky situation.

The AJ-effect has additional complications in the real world—ie, a world with uncertainty. Train offers stylized examples of a firm facing “good” and “bad” luck scenarios, showing that the core findings of the standard AJ model holds. The ratio of capital to labor will be inefficiently high under ROR regulation.

Train presents a number of possible solutions to the problems posed by rate of return regulation, all with their own thorns. They include: return-on-output; return-on-sales; return-on-cost; and price discrimination. In my time as an industry economist, I have never seen any of these employed in real life. At least not in a pure form (some amount of price discrimination does exist in utility pricing).

The next portion of the book returns to the problem posed at the beginning. How should a firm set prices if marginal cost pricing is unsustainable? The “second best” option is to price at average cost. This means that the firm meets its costs, and earns no profit. When the firm sells more than one product, as is often the case with a gas and electric utility, the optimal quantity produced of each product is the quantity where each product’s marginal cost curve intersects that market’s demand curve. Producing at average cost does not inherently result in optimal output for each good. Note that setting the combined average cost of each good equal to price results in zero profits just as setting price equal to the average cost of each good separately results in zero profits. In this way, there are an infinite number of price combinations that can lead to zero profits, with one product subsidizing the other. The way to achieve optimal output at “second best” pricing (average cost pricing) is through Ramsey Pricing. Ramsey Pricing maximizes social surplus by setting the price such that profits equal zero and ((P1-MC1) / P1)* ε1 = ((P2-MC2) / P2)* ε2. In other words, if the percentage by which price exceeds marginal cost in each market is multiplied by the elasticity of demand in that market, the resulting product is the same for all markets. Marginal cost and demand set the price and output, much like the “first best” pricing scenario.

Inducing Ramsey Prices is non trivial. A regulator does not necessarily know the demand curve or cost curve facing a firm. All it can see is the price, output, and quantity cost. In 1979, Vogelsang and Finsinger devised a mechanism that the regulator can use to induce Ramsey Prices, employing only these three variables. The basic process works like this: in period one, the firm charges whatever prices it wants. In period two, the regulator constrains the firm such that its second period prices * first period quantities do not exceed first period costs. Over each period, the firm will decrease price and increase outputs, eventually resulting in Ramsey Prices. This works on both downward and upward sloping Average Cost curves. The most prominent drawback of this mechanism is that it incentivizes pure waste—but this can be mitigated with random audits.

The remainder of the book discusses tariff design, which I deal with frequently in my day-to-day business. Train shows how customer selection of tariffs and TOU rates can increase total surplus, in various settings. The book ends with an appendix on the theory of price cap regulation, which is a big part of my job. I wish there was more on that! Alas, the book was published in 1991, though, when the world of price caps was still young.

Benefited from the read. Definitely accessible to anyone with a cursory background in economics.

See here for text: https://eml.berkeley.edu/~train/regul...

[psb · Rating 4 out of 5]

AKA "The PG&E Problem".