Microsoft, Chicken Delight and Competition for an Imperfect World

The New York Times, "Economic Scene" , June 20, 2002

The biggest antitrust case in recent years is drawing to a close. Yesterday in Washington, lawyers for Microsoft and the nine states pursuing antitrust action against it completed their final arguments. In addition to ruling in the states' case, Federal District Judge Colleen Kollar-Kotelly must decide whether to approve the settlement Microsoft reached in November with the Justice Department.

The Microsoft case showcased recent economic scholarship on imperfect competition, which uses mathematical game theory to understand the strategies companies might pursue to gain advantage in markets that depart from the abstract notion of perfect competition.

In perfect competition, many indistinguishable and anonymous buyers and sellers trade a homogeneous good at the price at which supply matches demand. No buyer or seller can influence that price individually.

While the perfect-competition model offers useful insights, away from the blackboard it is more the exception than the rule. It is not surprising that economists are interested in more complex and realistic arrangements. Hence the growth of theories of imperfect competition, in which companies' strategic choices affect market outcomes.

Under the right conditions, some theorists argue, it's possible to sustain and extend a monopoly that would break down under idealized competitive pressures. Often the way to do this would be for the monopolist to use exclusivity clauses or other "nonstandard" contract provisions.

This theory helped give intellectual impetus to the Clinton administration's antitrust activism, including the Microsoft case.

As economic theory, the work is sound. But if applied naively to antitrust policy, it "is likely to lead to poor legal rules and remedies," argues Paul L. Joskow, an M.I.T. economist, in The Journal of Law, Economics and Organization. He points to a different strand of economic research -- "transaction cost economics" -- that also needs to be taken into account.

This work begins with the insight, first made by Ronald H. Coase in a 1937 paper that would help win him the Nobel in economic science 54 years later, that making and enforcing contracts isn't cost free.

Professor Coase argued that reducing transaction costs helped explain why companies exist in the first place. Some exchanges might be easier and cheaper to manage if the parties are within the same organization.

In the 1970's, the economist Oliver Williamson began to develop this idea into a detailed program of research, with implications for antitrust policy as well as business management. Many economists have since done empirical and theoretical work on how organizational governance can lower contract costs.

Just because a contract is unusual or exclusive doesn't mean it hampers competition. The important question is whether competition existed before the contract. The unusual terms may provide a way to avoid enforcement problems.

Consider Chicken Delight, a franchise operation that was the subject of a 1971 antitrust case. Instead of collecting a percentage of store revenue, Chicken Delight required franchisees to buy all supplies from the parent company. From a transaction-cost perspective, this was an elegant solution to a basic problem: in a cash business, it is easy to lie about sales.

Supplies made a good proxy. If business was good, a store would need more supplies, and it would be harder to cheat.

Some franchisees sued, and a federal appeals court ruled that Chicken Delight had engaged in an illegal so-called tie-in arrangement.

But, Professor Joskow notes, the franchisees didn't have to sign up with Chicken Delight in the first place. There was plenty of competition. He teaches the case as an example of bad antitrust policy.

Or consider vertical integration, in which a company owns its supplier or customer. Despite the suspicions of antitrust regulators, in some cases integration may be the easiest, or only, way to benefit from long-term investments.

"The classic example is when you build a power plant next to a coal mine," Professor Joskow explains. "Before you make the investment in either the coal mine or the power plant, you have lots of choices. You can build a power plant almost anywhere. You can use coal, gas or oil. But once you sink the investment, you're locked into the relationship." The coal mine can exploit the power company by breaking its contract and jacking up prices.

"What my work showed," he said, "was you were more likely to get vertical integration in those cases, and if you don't, you get very, very detailed contracts with specific provisions for how to resolve disputes."

An antitrust lawyer who looked only at the situation after the plant was built might question whether the mine had been illegally tied to a long-term contract. A transaction-cost approach demonstrates why such a contract, or an acquisition, would come out of a competitive marketplace.

The imperfect competition school's "possibility theorems" make good social science. But they create too much uncertainty to make good law.

"They show that it's possible under some set of conditions this could be a bad thing, but it's also possible, maybe even probable, that it could be a good thing," Professor Joskow says.

As for Microsoft, he declines to weigh in on the merits of the case. But his article criticizes the original breakup order, noting that it didn't consider the costs or benefits or even how the divestiture would be done.

"This approach reflects the inaccurate assumption that organizational design does not matter for economic performance and that restructuring complex firms is a 'piece of cake' without the need for careful analysis or any significant economic consequences," he writes. "It ignores everything that we have learned from theoretical and empirical work" in transaction cost economics.