A successful resource allocation mechanism is one that provides households with the goods and services they would choose for themselves if they knew all of the production recipes for transforming inputs into outputs and also the size of the economy's stock of primary inputs. The notion of a group of consumers deciding among themselves the best menu of commodities is an ambiguous one. Let us suppose, briefly, that it is sufficient for the mechanism to cater to the preferences of just one consumer. We might call this consumer a dictator. Let us see what difficulties the dictator would have in designing a resource allocation mechanism to suit his preferences. Then we will see what additional difficulties arise when an economy is required to cater to individual wants generally.

The limitations implied by production recipes, or techniques, and by the stock of primary inputs will constrain the dictator's choice. But no individual can be expected to know all of the recipes and all of the details concerning input availability. Each important fact relevant to production is known by someone, but no one individual, even if he is a dictator, is privy to even a substantial fraction of the relevant information. In other words, knowledge is dispersed. The key to the successful operation of a resource allocation mechanism is its ability to organize and exploit essential information by means of a communication process.

Introduction

In this chapter we bring together the demand side and the cost structure of the regulated firm to start on our main theme: efficient pricing by the regulated firm for its services. The discussion is designed to bring out the main concepts as clearly as possible. These concepts are:

1. – Efficient prices are those which lead to the highest possible level of welfare, defined as the sum of consumer surplus and producer surplus.

2. – Moving from some given set of prices to efficient prices makes it possible for the “winners” from the price change to compensate the “losers” and yet still remain better off than before the change. Thus, potential economic welfare rises for all individuals.

3. – If the regulated firm must break even out of its own sales revenues, potential welfare of society is lower than if the regulated firm were not required to break even.

We first discuss these points in a simplified scenario where the demands for the services of the regulated firm are independent of each other, and where marginal costs are constant. Subsequently, the analysis is broadened to include cross-elasticities of demand between services and to consider non-constant marginal costs. The discussion will cover marginal cost pricing, so-called Ramsey pricing and the accounting-based methods known as Fully Distributed Cost (FDC) pricing. In addition, we will discuss the literature on cross subsidy and the recent axiomatic approach to the allocation of common costs.

In recent years there has been much debate within the United States on the regulation of public utilities. This debate has been coincident with a burgeoning of interest on the part of professional economists in public utility pricing issues and in the theory of regulation. The purpose of this book is to show how some of the recent advances in the theory pertain to the policy discussion. We hope that by making these advances accessible to policy makers we may foster a mutually beneficial interchange between economists, regulators and the public utilities.

Our interest in this area dates back to the mid 70's and was actively supported and encouraged by Elizabeth Bailey, Gerald Faulhaber and Edward Zajac who were responsible for the economics research activity at Bell Laboratories. We owe a debt of gratitude to our present and past colleagues at AT& T Bell Laboratories and Bell Communications Research for stimulating discussions on the subject matter of this book. Our debt to Peter Linhart, Patrick Marfisi, John Panzar, Jeffrey Rohlfs, William Sharkey and Robert Willig is particularly strong. We have benefitted not only from their research results, but from conversations extending back a period of years. Panzar, Sharkey, Leonard Mirman, Peter Grandstaff, David Mandy and Donald Brown also provided us with detailed comments on portions of the book. We would also like to acknowledge helpful conversations and feedback from Stanley Winkler, Roger Noll, and David Sappington. Finally, we acknowledge a particular debt of gratitude to Ms. Geraldine Moore who set the type for this book.

This book aims to make recent developments in public utility pricing theory accessible to the non-technical reader and to show how they can be usefully applied to major policy issues in ratemaking. Several policy issues have arisen within the last fifteen years or so which cannot be analyzed correctly without these developments. The classic treatise of Kahn [1970], although offering a wealth of institutional detail and breadth which we cannot match, summarizes the relevant economic theory at a point in time just short of a series of major advances which began to take place shortly after the appearance of Kahn's book. It is useful to sketch the policy issues and the research advances to which they led.

One policy question has to do with the rationale for declining block tariffs. Traditionally, such tariffs have been justified on two grounds. In the first place, utilities must cover large fixed costs of operation. In electricity, for example, the maintenance of a line from a pole to a customer's meter is a fixed cost that may be attributed to that customer. In addition, there are costs that may not be so readily attributed to a particular customer such as the cost of maintaining storage facilities and a portion of the reticulation system in the case of a gas utility. Some of these costs are covered by fixed charges that a customer must pay regardless of usage.

Introduction

In the previous chapter we developed the major theme of second-best pricing: to cover total costs of the regulated firm with minimum deadweight loss. The term Ramsey prices was used to refer to the set of uniform prices which maximize total surplus - minimizing deadweight loss - subject to the breakeven constraint. Ramsey prices do this by charging different prices to the regulated firm's various markets with the aim of generating the largest amounts of contribution from markets in which a high markup of price over marginal cost will perturb consumption levels least from what would be achieved with full marginal cost pricing. In this chapter we will broaden the analysis to include price structures which permit us to vary prices not only between markets, but also between consumers in the same market.

The device for doing this is called the nonuniform price schedule. A nonuniform price schedule is a tariff for one or more goods in which the consumer's total outlay does not simply rise proportionately with the amounts of the goods he purchases; quantity discounts and quantity premia are permitted. Analogously, we call a tariff a uniform price schedule when total outlay is simply proportional to the amount purchased. Thus, Ramsey prices are the uniform prices which maximize total surplus subject to the constraint that the regulated firm break even. An example of a nonuniform price schedule is given in Figure 4.1, where we depict total outlay per month at different consumption levels under a residential electricity tariff used by the Commonwealth Edison Company in 1976.

Introduction

In the last chapter we showed that a properly chosen set of n+l selfselecting two-part tariffs could Pareto dominate a set of n self-selecting two-part tariffs. The reason for this welfare dominance is that the set of n+1 two-part tariffs gives each type of consumer the ability to find a two-part tariff more closely attuned to his willingness to pay than under the original set of n two-part tariffs. Because of the relationship between a set of n self-selecting two-part tariffs and a single nonuniform price schedule with n rate steps, this result suggests that welfare can be made progressively higher by constructing nonuniform price schedules with more and more rate steps, each one progressively smaller in length, obtaining in the limit a continuously varying nonuniform price schedule.

In the upper panel of Figure 4.14 we depicted four self-selecting two-part tariffs which form the shaded piecewise-linear outlay schedule. The associated marginal price schedule is the declining block schedule in the lower part of that figure. If we were to let the number of optimally chosen self-selecting tariffs tend to infinity, then the resulting outlay schedule would be the smooth lower envelope of the two-part tariffs given as R(Q) in Figure 5.1. The corresponding smooth marginal price schedule is given as the curve P(Q) in the lower panel.

In this chapter we will explore the properties of smooth nonuniform price schedules which maximize total surplus, subject to the firm breaking even.

Introduction

The aim of this chapter is to introduce basic concepts of welfare economics and industrial organization that shall be used throughout the book, and to introduce the paradigm of the regulated firm that we use throughout our discussion. We intend neither a full and complete treatment of these concepts nor an elementary textbook treatment of the material. Such is available elsewhere. Our intent is to provide a brief and nontechnical introduction to those ideas used most frequently in chapters subsequent to this.

We assume that the objective of regulators in setting prices is to maximize social welfare, broadly defined. Pricing policy serves this function in two ways: directly, by redistributing wealth in society, and indirectly, by signalling a reallocation of resources in society.

So called lifeline service that offers minimal service at low rates to poor people and the elderly is often a redistribution to these groups from the company and other consumers. It achieves the social purpose of providing public utility service to those who would not otherwise be able to afford it. This is an example of the redistributive function of pricing policy. As an example of the reallocative effect of pricing, imagine that the utility were to offer an optional day/night tariff constructed to yield the same revenue to the company as some existing tariff. Consumers cannot lose by such an offering to the extent it is indeed optional.

Introduction

The optimal nonuniform price schedule P* (Q) generates more consumer surplus for a given revenue requirement than any other self-selecting tariff. In operational terms however, it is important to know just how much better it does over arguably simpler Ramsey and Fully Distributed Cost pricing rules. To answer this question, we need to be able to compute these alternative pricing rules and evaluate the consumer surplus and revenues generated for a variety of assumptions about demand and cost conditions.

These computations represent a potentially difficult numerical problem where the willingness to pay reflects differences in tastes across a population of individuals. The consumer surplus and revenue integrals have themselves to be integrated over consumer types. In the case of the optimal nonuniform pricing rule, there are functions to be twice integrated that involve an optimal price schedule P*(Q) that is only implicitly defined.

This implicit definition of the price schedule P*(Q) arises from the fact that the optimal price is derived from the first order conditions specific to a given customer type θ. The optimal price schedule is then defined in terms of the maximum quantity consumer type 6 will purchase. In other words, given a willingness to pay function p(Q,θ) decreasing in Q and increasing in consumer type θ, the maximum type θ will consume is given as the Q for which P = p(Q,θ). Thus, provided the self-selection constraint is satisfied it is possible to define (implicitly) the unique function P = P*(Q) for all consumer types θ.

Introduction

In this chapter we will apply efficient pricing theory to policy questions relying heavily on a special, but important, case study: the market for message service on AT&T's public switched network. The particular interest of using this market is that the relative merits of FDC pricing, Ramsey pricing, and nonuniform pricing in this market were vigorously debated for years, with few clearcut results. Although our analysis will probably raise more questions than it settles, it will bring a useful perspective to that debate.

For some years, AT&T and the Federal Communications Commission (FCC) debated the rationale for offering Wide Area Telephone Service (WATS) as a service distinct from ordinary Message Toll Service (MTS); the latter is sold under a predominately uniform price tariff and the former under a tariff displaying quantity discounts. AT&T argued that WATS is fundamentally a different service from MTS, so that it can be offered under a different tariff. The FCC took the view that WATS and MTS had so much in common that they should be regarded as the same service in all functional aspects and that any tariff differences between the two should be due only to cost differences. The FCC argued that, absent such cost differences, the existence of WATS as a separate service with a bulk discount tariff constituted price discrimination against small users, who bought on the MTS tariff, which has a higher usage charge.

Introduction

A key ingredient in most discussions of efficient pricing theory is the assumption that the regulated monopoly's customers are independent of each other. That is, it is assumed that the amount consumed by one customer has no impact on the surplus which can be earned by another customer. For this reason, it is possible (and convenient) to ignore interactions between consumers in designing efficient prices.

In reality, there are many reasons why interactions between consumers could be important; in such cases, efficient pricing rules should take account of them. One especially important type of interaction occurs when the utility sells both to business customers and residential consumers. Residential consumers buy the utility's services as final products to be consumed. The business customers buy the utility's services as inputs into their own production processes, which produce outputs that they sell to other businesses and to final consumers, including, possibly, the utility's residential customers. The prices that the utility's business customers charge for the goods and services which they produce link their welfare to those of the utility's residential customers, who buy both the service of the utility and the outputs produced by the utility's business customers. Individual business users' profits are linked to the extent that they compete with each other. This type of interaction also occurs when a utility sets a nonuniform price schedule for business customers. Changes in marginal prices at given points on the schedule will affect total outlay for firms which consume larger amounts of the utility's services. This, in turn, affects prices in these industries.