Fragmentation and Synthesis
in the Economics of Industries

Paul W. MacAvoy

Since the Peck courses on industry economics faded away in the mid-1980's, the topic area has broken up into a number of separate centers of activity. The first is focused on strategic analysis of corporate behavior, based on scale advantages, product substitutability, and extent of collusion. Economics has fought a battle with corporate-level application of game theory. Strategic analysis has won most of the battles, because of lack of validation of game theoretic models, but may have lost the war in the long run because of unrealistic assumptions about corporate interactions.

Second, antitrust has become a beehive of activity because of the concern of the courts for the leveraging behavior of dominant firms -- Microsoft, Bell Atlantic, Citigroup -- that extends monopoly powerfrom one product market to another. Again, the games theorists build models to describe leveraging as either pro-competitive or monopolizing, depending on exacting assumptions. The economists look at market behavior, and with applications of price theory, generate contradictory empirical results. That makes for lots of work,numerous journal articles, and voluminous expert witness testimony.

The third cluster of intense activity is the most interesting. Six of the infrastructure industries, all built on star network distribution systems, have broken off from the public sector, in a flurry of deregulatory and privatization events, to come to rest in a more or less permanent "transition" position halfway to the private sector. The economic analyses of the deregulatory phases that each of these industries has gone through before getting stuck are thickets of controversial findings. But to proceed further towards full deregulation the dominant companies have had to demonstrate that "competitiveness" results from what has been done so far. In contrast with antitrust one can now begin to detect some consensus in evaluation of performance of those companies that are half regulated.

The Airline Deregulation Act of 1978, and railroad and trucking legislation in the same time period, took transportation controls of entry and rate setting out of the hands of regulatory agencies and then closed down the agencies. On large volume service offerings, at spokes and hubs in high-density markets, rates declined by more than 25 percent, the number of service suppliers declined and the volume of service increased. Early analytical studies optimisticallyanticipated the emergence of "effective" competition; but as patterns of trade became established, by the early 1990's, it has become recognized that rate-cost margins vary with the (small) number of providers in a spoke, that market dominance by one or few firms exists at many of the major hubs, and that rate structures are enormously complex combinations of discriminatory prices. When the agencies were abolished, the regulators moved over to other agencies, taking limited authority with them. Regulation persists, and suppliers are still in transition, with railroad rate ceilings on single-line service at bottleneck spokes, airline controls on landing slots, and state controls on truck route and equipment size.

In natural gas every policy experiment has been tried in forty years of regulation and twenty years of attempted deregulation. To increase the security of gas supply for consumers, the federal regulatorycommission fixed wellhead ceiling prices, which resulted in sustained, large shortages in the last half of the 1970's. Partial price decontrol of C4 new" gas produced as large a "bubble" of surplus gas the first half of the 1980's. Producers, pipelines, distributors, and consumers all lost substantial (Harberger triangle) surplus. Then industry "restructuring was mandated by a new federal regulatory agency. Pipelines were divested of wellhead gas, gas wellhead prices were totally deregulated, and pipeline space was leased by brokers, retailers and consumers to carry their gas bought on spot and contract markets at a dozen (computer) market "hubs". Prices fell by half, at the wellhead, and by a quarter at the burner tip; price risk was shifted to the consumer, and shortages during extreme cold winters in a national market of spot, contract, and futures gas and separate transportation never emerged. Regulation persists, in the form of price caps on contract transportation, thus adding to the risk of excess demand in transmission, and thus reducing the size of the market. Prices for gas and transportation continue to decline in real terms.

Telephone and electric service have been subject to gas-style "restructuring" with upstream sources of service separated from downstream service to the customer. Where there is purported to be potentially competitive conditions for entry, expansion of smaller providers, etc., the separated services have been at least partially deregulated. Where there are purported to be bottlenecks, with the incumbent service provider the dominant source of supply, then regulation continues to exercise control, whether "light-handed" or according to "cost of service" principles. Transactions between deregulated and regulated providers are monitored and price controlled by regulatory agencies more stringently than ever before. Because the previously integrated service providers were notoriously inefficient, the substitution of market alternatives upstream for the old integrated source has reduced costs and some consumer prices. But these source markets can scarcely be termed competitive, and the regulatory agencies have been sorely tempted to extend their controls back from the bottleneck again to the source.

What market and regulatory processes bind together the dynamics of these six network industries? All have been fundamentally changed by the semiconductor and software revolution. The new multiplicity of transactions, in airline reservations systems, freight inventories, phone intemet services, etc. made it possible to provide service of higher quality only if cost-of-service regulation was eliminated. Deregulatory processes that remove entry barriers reduced prices and increased service offerings at some level of each of these industries. But the regulatory agencies have held on, maintaining jurisdiction over bottleneck loops and nodes in these systems, and as a result presiding over performance below capacity, with constrained introduction of new technology and lower rates of competitive entry. This has been based on the rationale that half regulated is not effectively competitive, so that full deregulation is not justified.