If It Ain’t Broke, Don’t Break It Up

Robert W. Crandall
Robert W. Crandall Adjunct Senior Fellow - Technology Policy Institute

June 14, 2000

Judge Thomas Penfield Jackson’s decision to break Microsoft into two parts in order to “create” competition is not only scary for software users, but virtually unprecedented as an antitrust remedy.

Anglo-Saxon common law has long frowned on practices that led to monopoly, but the courts generally couldn’t break up the offenders. Even after the first U.S. antitrust law, the Sherman Act, was passed in 1890, divestiture was a rare remedy reserved for entities organized to fix prices or for attempts to control an industry by buying competitors. Nobody has accused Microsoft of either offense.

Worse, when applied, the ultimate penalty generally hasn’t worked. In the few cases in which the government succeeded in breaking up a single-firm monopolist, divestiture has proved ineffective, irrelevant or—in one major case—costly overkill.

Sound familiar?

Begin with the stunning government victories in 1911 against Standard Oil and American Tobacco. Both “trusts” had been assembled by combining scores of competing firms and related enterprises. Both companies were broken up by court edict, but with very different results.

At the time of the divestiture, the oil industry was becoming competitive because it was moving into Kansas, Oklahoma, Texas, Louisiana and California—states where Standard Oil didn’t have much power. The price of petroleum was actually falling in the years before the case, because these new fields were pumping so much crude. The breakup had no measurable effect on oil production, crude-oil prices or refined-product prices.

American Tobacco was quite another matter. It had put together enough companies to control the production of most tobacco products, particularly cigarettes. Divestiture established three major cigarette producers. For the next 30-odd years, three companies battled for market share through advertising and marketing, but they didn’t compete on price. The government finally conceded that competition wasn’t working and filed another antitrust case.

Antitrust became somnolent after World War I when the government lost a monopoly case against U.S. Steel. In the late 1930s, however, the Roosevelt administration went after Alcoa, the only producer of primary aluminum in the U.S. It also launched an investigation into the motion picture industry, which distributed and exhibited movies through a range of collusive agreements.

The movie investigation culminated in a major case brought against eight Hollywood studios, five of which also owned a large share of the country’s first-run theaters. To deter collusion, the five “majors” were forced to sell their theaters. This turned out to be a blessing for stockholders, as a new medium called television would soon deal a fierce blow to the movie-theater business.

Despite the decline in demand for movies, the real price of tickets actually rose after the 1948 antitrust case, because distributors reduced output and raised their share of box-office receipts. There were no new entrants for nearly two decades after the decrees. No matter how you slice it, divestiture didn’t increase competition.

Alcoa belongs in a different category. The antitrust suit filed in 1937—not the best of economic times—accused Alcoa of maintaining a monopoly through a variety of business practices that weren’t illegal in themselves, but that kept potential entrants at bay. Sound familiar?

But unlike the Microsoft case, the government couldn’t persuade the court to break up Alcoa. This was in part because the government had sold the aluminum production plants it had built during the war to Alcoa’s would-be competitors, Kaiser and Reynolds Metals. There wasn’t any measurable increase in the rate at which aluminum prices had been declining under the Alcoa monopoly.

Alcoa’s domestic monopoly had largely been the result of economies of scale in refining alumina, the first step in producing aluminum—not its alleged nefarious anticompetitive practices. And it wasn’t until the late 1950s, when the market had grown big enough to accommodate six competitors, that competition became enough to affect prices. No antitrust decree could have established competition in the prewar market. The case was simply wasted effort.

The next futile pursuit of a monopolist was the United Shoe Machinery case. United produced most of the machines shoemakers used to produce leather footwear, and it did so from a single plant in Massachusetts. In addition to supplying a full line of machines, United did a lot of hand-holding for legions of small shoemakers. The government filed a Sherman Act case against United in 1947, alleging that its leasing practices kept smaller firms from competing and inhibited the rise of a used-machinery market. The government wanted to break up United, but the trial judge simply ordered the firm to make its machines available for purchase as well as lease, and to change some of its leasing practices.

Ten years later the judge decided the decree was “working” because a used-machinery market had developed and competitors had gained market share. But the Supreme court reversed him, and in 1969—22 years after the complaint—the judge was forced to order divestiture. By then the shoe industry had begun a long decline that would result in the marginalization of United and its ultimate sale to a larger company. There is no evidence that either the first or the second antitrust decree had any effect on shoe machinery prices.

In 1969 International Business Machines was charged with monopolizing the computer industry. But whatever IBM had done to incur the Justice Department’s wrath became irrelevant in a world in which personal computers and minicomputers were making deep inroads into the “mainframe” computer business. Thirteen years later, the government simply dropped the case.

Finally, in 1974, the government brought the mammoth AT&T case, alleging that AT&T had used its government-protected local phone monopolies to frustrate competitors’ entry into the long-distance and equipment-manufacturing businesses. Perhaps the federal and state regulators should have been co-defendants in the case because they had required AT&T to keep long-distance rates far above cost in order to subsidize local service. AT&T, not surprisingly, also felt it had to defend its long-distance business and aggressively did so by making it difficult for MCI and Sprint’s predecessor to connect their calls through local AT&T facilities. AT&T finally agreed to divestiture in 1982 because it felt it was about to lose the case.

The 1984 AT&T divestiture has become very important to the proponents of a Microsoft breakup because, to all appearances, it “worked.” Long-distance competition flowered and rates fell. Three relatively large newcomers, two of which have since merged and are proposing to acquire the other, captured nearly 40% of the long-distance market over the next 12 years.

But divestiture was not the reason for this increase in competition. It was the provision in the decree that required the divested local Bell companies to provide “equal access” to their circuits for all long-distance companies.

If you don’t believe this, look to Canada. In 1993, Canada required telephone companies to offer equal access to all long-distance entrants, but it didn’t break up the established companies. Bell Canada still offers nearly all the local service in Ontario and Quebec, as well as a national long-distance service that has some 60% of the market. Yet today its rates are as low as U.S. companies’, or perhaps slightly lower. A simple rule from the Federal Communications Commission could have managed what it took decades and a wrenching court-mandated reorganization to accomplish.

A Page of History

Justice Oliver Wendell Holmes once observed that a page of history is worth a pound of logic. What can be learned from this page of antitrust history that is relevant to the Microsoft case?

The Standard Oil case illustrates that government is often incapable of reorganizing when the market is changing on its own. The American Tobacco failure shows that the courts are able to turn monopolies into oligopolies, without any measurable change in prices. The IBM fiasco makes clear that antitrust is a slow and cumbersome weapon in the fast-paced world of high-technology. The AT&T case suggests that when it comes to remedies, less is often more.

Nothing in the historical record of antitrust should make us confident that the court’s dismemberment of one of the most successful companies in history would increase competition.