Dr. Ellig is Assistant Professor in the Program on Social and Organizational Learning at George Mason University, Fairfax, Virginia.
Way back in high school, a far-from-right-wing teacher made our class sit through a short film called “The Incredible Bread Machine.” The story’s protagonist invented a bread-baking machine so productive that it vastly reduced bread prices and promised to end world hunger. But instead of praising his achievement, the government pilloried him as a monopolist, claiming that he drove competitors out of business and charged monopoly prices. He could do both because his wondrous invention dramatically slashed the cost of baking bread. Despite this boon to society, the movie implied that the inventor landed in the pokey for a Sherman Act violation.
Even during the malaise years of the late 1970s, the movie sounded like a stretch of the libertarian imagination. Surely, if someone invented a technology that is superior, even politicians and regulators would have to recognize him as a benefactor and let him go about his business.
Recent events in the airline industry have shown otherwise. During the past 20 years, several airlines developed a new, computerized technology to process reservations and issue tickets that is the equivalent of the “Incredible Bread Machine.” Unfortunately, the developers of airline computerized reservation systems have fared little better than the ill-fated inventor of the bread machine.
What Is a Computerized Reservation System?
The Airline Deregulation Act of 1978 freed airlines to enter new routes and charge whatever fares they could induce customers to pay. In this new, deregulated environment, an airline of any size has thousands of different route combinations, generating tens of thousands of different airline fares. Before the invention of computerized reservation systems, a travel agent making a reservation had to look up flights in a book called the Official Airline Guide, then phone the different airlines to find out if they actually had seats available at a particular date and time. If the customer wanted a hotel room or rental car, the agent had to make another round of phone calls to find out what was available and make reservations.
Computerized reservation systems automated and streamlined this process, often permitting travel agents to consult with the customer, search for available seats, and book the reservation within minutes. When a customer phones, a travel agent begins searching for suitable flights on a computer screen. The screen is linked through phone lines to a mainframe computer owned by the company that operates the reservation system. The mainframe contains information on millions of available airline flights and fares. As soon as the customer and agent select a flight and agree on payment, the agent can print out the ticket. Similarly, a travel agent can access information on available hotel rooms and rental cars, making reservations within minutes.
Computerized reservation systems slashed travel agent and airline costs while dramatically reducing the time it takes to make reservations. By 1981, use of these systems had become widespread, and a Harris survey in that year indicated that computerized reservation systems had increased travel agents’ productivity by an average of 41 percent. In fact, “One travel agent estimated that his employees could make a reservation using a computerized reservation system in one-third the time it would take to look up schedules in a book and make reservations over the telephone.” United Air Lines estimated that a reservation system’s booking fee is less than half the cost of making a reservation through the airline’s own reservation staff. Although travel agents retain the option of returning to pre-computerized reservation system technology, 95 percent subscribe to at least one system.
Of course, if computer systems cut the time and effort needed to make reservations, they no doubt eliminated some jobs in the travel industry, and possibly in the airline industry’s reservations departments as well. In effect, these systems substitute human brain power and the sand in a microchip for the brute force labor previously required to process airline reservations. They provide a textbook example of what the economist Joseph Schumpeter called “creative destruction”: the replacement of old technologies and ways of doing things with new arrangements that drastically reduce the amount of resources required to get the job done. In so doing, the new methods and technologies may cause some temporary dislocations, but they vastly increase our standard of living by enabling us to satisfy more human desires with fewer resources. Schumpeter argued that the process of creative destruction is in fact the driving force behind economic progress.
The Competition Issue
In the case of computerized reservation systems, some critics saw a great deal of destruction and very little creativity. Instead, voices in both industry and government perceived computerized reservation systems as a threat to competition in the airline industry.
There are currently four major computerized reservation systems in the United States, but most travel agents subscribe to only one system. Typically, a travel agent signs a long-term contract with a computerized reservation system provider, who then installs computer terminals and other equipment in the travel agent’s offices. The contracts have included rollover provisions, so that when a system vendor adds equipment or features, the contract is automatically extended. Minimum-use provisions have required travel agents to book 50 percent of their flights on a given system, and liquidated-damages provisions have discouraged travel agents from switching to a new computerized reservation system.
There is also some evidence that this is a difficult industry for new firms to enter. The computer systems and software needed for a reservation system may not have much resale value if their developer chooses to leave the industry. Analysts at the Department of Transportation also found evidence of economies of scale, economies of scope, and learning-curve effects, which suggest that the first firms to enter this industry may enjoy a substantial advantage over new competitors. The existence of long-term contracts helps augment this advantage.
Most of the policy controversy stems from the fact that all four major computerized reservation systems are owned by airlines or groups of airlines. The two systems with the largest market shares are Sabre, owned by the parent company of American Airlines, and Apollo, developed by United Air Lines and now owned by a consortium of domestic and foreign airlines. To be listed on a system, an airline must sign a contract with the system vendor, and it then pays the vendor a booking fee every time one of its tickets is booked using that system.
Critics charge that ownership of a reservation system gives an airline an unfair advantage over competing airlines. Computerized reservation system technology allows the airline owning the system to list its own flights before those of other airlines, update information on its flights before updating the information of rivals, and in general manipulate the information presented to travel agents in ways that promote the owner’s flights over those of rivals. In the absence of regulation, the system owner could also charge discriminatory booking fees, forcing other airlines to pay a higher price when customers choose them.
These types of complaints are common among companies in many types of businesses; the “have-nots” can always be counted on to complain about the “unfair” advantages possessed by the “haves.” But the airline case is unusual, because many defenders of airline deregulation agree that computerized reservation systems pose a threat to competition. Most economists, for example, agree that airline deregulation has lowered prices and greatly enhanced the convenience of air travel. Yet even the most authoritative study on the impact of airline deregulation warned that computerized reservation systems might require regulation because they give the large, established airlines that own them an advantage over new upstarts. An entire mini-industry of scholarly articles focuses on ways of achieving the full benefits of airline deregulation through government intervention to eliminate remaining “market imperfections,” such as those created by reservation systems.
In response to both theoretical and practical concerns, the federal government has issued a number of regulations that govern the conduct of computerized reservation system vendors. “Display bias” is outlawed; vendors are prohibited from giving preferential treatment to any airline’s flights in constructing video displays or operating the system. Contracts between vendors and travel agents are limited to five years. The government prohibits discriminatory booking fees, tie-ins, and exclusive-use contracts. Travel agents must also be permitted to use software that would allow them to access all computerized reservation systems from the equipment supplied by one system vendor.
All of these regulations were adopted not by social planning zealots, but by the Civil Aeronautics Board under President Reagan and the Department of Transportation under President Bush. In both cases, the responsible officials appreciated free markets enough to firmly resist a return to the pre-1978 regulatory regime, under which the government limited entry and regulated airline fares. In fact, they saw the regulations as quite consistent with the goals of airline deregulation, which replaced a government-enforced cartel with a competitive marketplace.
In general, these regulations stemmed from the belief that ownership of a computerized reservation system constitutes a “barrier to entry” that will impede competition and prevent consumers from reaping the full benefits of airline deregulation. This policy position, however, begs a very important question about the true nature of competition in this and all industries.
The Importance of Conceptual Frameworks
What prompts some of the staunchest defenders of deregulation to advocate potentially extensive regulation of a new technology? The answer can be found in the concept of competition employed by most economists who analyze the airline industry. In modern neoclassical economics, “competition” means the abstract model of perfect competition. Perfect competition is characterized by numerous buyers and sellers, homogeneous products, perfect information, and costless entry and exit from the industry. Economists have logically proven that, under such conditions, decentralized markets allocate every unit of every resource to its most highly valued use. This result, termed “allocative efficiency,” is the standard of economic efficiency by which many economists judge policy proposals.
No one seriously argues that regulators or antitrust authorities should impose perfect competition on the airline industry, or on the computerized reservation system industry. However, many common economic policy prescriptions in this industry are based on the related model of “perfect contestability.” A perfectly contestable market may have only one seller, but as long as there are no barriers to entry or exit, the threat of potential entry can force the monopolist to behave as if it had numerous competitors. The end result is the same as perfect competition: prices closely reflect marginal costs, and allocative efficiency is achieved.
A key policy prescription that emerges from contestability theory is that government can promote efficiency by removing barriers to entry. These include not just government-imposed limitations on competition, but also private barriers to entry that take the form of “sunk cost” facilities. A “sunk cost” is an industry-specific investment that a firm cannot recoup if it chooses to leave the industry. In the airline industry, examples of sunk cost facilities include airports, the air traffic control system, and computerized reservation systems. When facilities involving sunk costs exist, government policy can promote contestability by ensuring that all actual and potential competitors have equal access to the facilities. This might be accomplished by government ownership, as with airports and air traffic control, or it could be accomplished through regulation of privately owned facilities, as is done with computerized reservation systems.
Given this theoretical background, it should come as no surprise that many people who appreciate the virtues of free markets advocate regulation of computerized reservation systems. A good many analysts believe that free markets enhance human welfare by moving us closer to the ideal of allocative efficiency. When private market developments seem to thwart allocative efficiency, it is time for government to intervene.
A Different View of Competition
Such prescriptions make perfectly good sense, if indeed allocative efficiency describes the sum total of human welfare. But many advances in our standard of living occur because of phenomena that the models of perfect competition and contestability assume away. Both of these models describe a world in which all change that will occur, has occurred. The assumption of perfect knowledge rules out the possibility that someone might discover something new tomorrow. To the extent that technological progress is foreseen, market participants incorporate it into their plans today.
In the theoretical world that generates the policy norm of allocative efficiency, Schumpeterian creative destruction is ruled out by assumption. That does not mean that the conventional neoclassical economic models might not be useful for some purposes; they are just not particularly useful for understanding entrepreneurship and creativity. To the extent that entrepreneurship and creativity exist in the real world, we risk misperceiving them if we insist on interpreting them through the lens of the neoclassical model.
Airline computerized reservation systems provide a clear case in point. Several companies became the industry leaders because they were the first to capitalize on the new possibilities that information technology offered the airline industry. Indeed, the first computerized reservation systems began as extensions of American and United Air Lines’ own internal computer systems. These pioneers quickly signed up many travel agents, and before regulators intervened, computerized reservation systems were a powerful marketing tool that gave the owners a competitive advantage in selling tickets on their own flights. The computerized reservation system owners earned high profits—in the form of booking fees and additional ticket sales because they were the first to enter the market with an innovative new service.
These profits, however, are not a sign that government needs to eradicate imperfections in the airline market. Rather, the profits are the reward that the innovators received for entering a new, untried business. When government policy permits entrepreneurs to keep their profits, it promotes the alertness to new opportunities that is the first step toward achieving a more productive use of resources. When policy penalizes such profitability, it penalizes entrepreneurial activity aimed at the discovery of newknowledge. Without such discovery, there can be no replacement of old methods or technologies by new ones. A consistent quest for allocative efficiency will sometimes impede the movement of resources to more highly valued uses, because it can undermine the discovery process that drives this reallocation.
“Where you stand depends on where you sit” is a common saying in the nation’s capital. In the battle over computerized reservation system regulation, plenty of naked self-interest is in evidence. But the policy debate reveals more than self-interest; it reveals a clash of conceptual frameworks. Analysts who take allocative efficiency as their policy ideal will tend to seek regulation that eliminates any competitive advantage an airline might gain from owning a computerized reservation system. A minority will be more skeptical of regulation, because it erodes the incentive for entrepreneurial discovery. As the information age unleashes its creative potential, we can only hope that the minority view gains greater recognition and understanding. Economic progress depends on it. 
- . “New Reservations About Airline Computers,” Frequent Flyer (December 1982), pp. 45- 50.
- Comments of United Air Lines, DOT Docket No. 46494 (November 20, 1989), p. 9.
- U.S. Department of Transportation, Study of Airline Computer Reservation Systems (May 1988), p. 10.
- Joseph Schumpeter, Capitalism, Socialism, and Democracy (New York: Harper & Row, 1942), pp. 81-108.
- U.S. Department of Transportation, Study of Airline Computer Reservation Systems (May 1988), pp. 24-27. Strictly speaking, these may not constitute “barriers to entry” in the economic sense; see George Stigler, The Organization of Industry (Chicago: University of Chicago Press, 1983), pp. 67-70.
- For a general study of the economic literature on airline deregulation, see Jerry Ellig and Wayne Winegarden, “Airline Policy and Consumer Welfare,” Transportation Practitioners Journal, Spring 1994.
- Steven Morrison and Clifford Winston, The Economic Effects of Airline Deregulation (Washington, D.C.: Brookings Institution, 1986).
- See, for example, Severin Borenstein, “Hubs and High Fares: Dominance and Market Power in the U.S. Airline Industry,” Rand Journal of Economics (1989), p. 344.
- See William Baumol, John Panzar, and Robert Willig, Contestable Markets and the Theory of Industry Structure (New York: Harcourt Brace Jovanovich, 1985).
- Elizabeth Bailey, “Contestability and the Design of Regulatory and Antitrust Policy,” American Economic Review (May 1981), pp. 178-183.
- Israel Kirzner, Discovery and the Capitalist Process (Chicago: University of Chicago Press, 1985).