The Great Antitrust Breakup: Often Threatened, Rarely Executed
The breakup is something that has been normalized in our conversations about large companies. One may be forgiven for thinking that a breakup is a relatively simple and common remedy. However, the non-merger breakup is actually something we have little experience with. There are only three instances of a breakup being used in non-merger cases, with the last being the breakup of AT&T in 1982.
Breaking up a company is very difficult to do and the results could make everyone – not just the company – worse off. It’s a little like brinkmanship in global politics: the strategic threat of an extreme policy can sometimes get a country what they want but the plan is to always back away from the policy without executing. In antitrust, the threat of breakup could encourage good behavior from companies that have grown large. But actually executing a breakup could undermine efficiencies, stifle innovation, and harm stakeholders.
Breaking up a company for Section 2 violations, which are monopolization claims under the Sherman Act, is a dramatically different remedy than when applied to merger cases. Mergers create one company from two or more distinct companies, so it is far easier to see the lines where breakup should occur. But even the lines in merger cases are only clear for a limited time. This is why antitrust enforcement agencies fight so hard for injunctions to stay mergers pending judicial review. It can become very difficult to “unscramble the egg” of a consummated merger once the companies have consolidated operations. In a non-merger case, there are rarely clear lines between business units that allow an enforcer to break off a fully functioning company from the larger whole.
Setting aside U.S. v. United Shoe Machinery, which was an unusual case where the company was forced to sell off assets after a court-ordered conduct remedy failed, there are only two examples in U.S. history where a company was broken up as an antitrust remedy in a non-merger case. Both of these cases had an important factor in common – they had well-defined operational units that made it easier for them to be carved up into smaller companies. These cases shared other similarities as well. Importantly, they not only had high market shares but had control of vital infrastructure and engaged in clearly anticompetitive behavior.
Standard Oil: Standard Oil was a holding company that controlled over 90% of oil related-assets in the U.S., including the production, shipping, refining and selling of petroleum and its products. The company also engaged in a lot of anticompetitive behavior. It bought out competitors and was accused of bribery, espionage, and blackmail. The fact that Standard Oil was a holding company made it easy for enforcers to break it up into its subsidiaries.
AT&T: AT&T was a giant at the time of its breakup. Not only did it control local phone service, it was also a major player in the manufacture and sale of phone equipment through Western Electric and research and development through Bell Labs. The government alleged that AT&T used its power over local interchanges and local phone service to prevent customers from installing rival equipment and prevented competitors from offering long distance services. What makes the AT&T breakup interesting is that it voluntarily entered into the settlement that divided it up and helped the government to determine how the breakup should occur. The breakup of AT&T was hailed as a success story, but the company has been largely put back together:
History gives us some clues as to how the breakup remedy should be used. First, it should be a remedy of last resort. It is an extreme remedy that could do more harm than good to a market. Second, the monopoly in question must present a combination of durable power enabled by control of vital infrastructure and a flagrant anticompetitive use of that power to stifle all emergent competition. Third, the monopoly must be easily split apart, and the split needs to actually address the identified harms. Arbitrary divisions are more likely to be harmful than help consumers.
Learning from Microsoft
Some wonder if large companies in tech fit the mold of a breakup remedy. We have a good indicator of this in U.S. v. Microsoft, where the Department of Justice argued that Microsoft needed to be broken apart into an application business and an operating system business. This breakup, the DOJ argued, would ensure that the applications company would “have the incentive to ensure that Microsoft’s applications . . . can work with competing PC operating systems,” which would “reduce the barriers to entry protecting Microsoft’s monopoly.” The district court agreed with the DOJ, but the breakup order was eventually tossed out in favor of behavioral remedies.
The decision to go with behavioral remedies in the Microsoft case was probably correct. If the breakup is a remedy of last resort, then history has proven it wasn’t needed. While it was perhaps inconceivable that Windows would ever face competition in 1995, we know today that Apple was able to capture a significant share of computer users – especially in the graphic design industry – even before the advent of the mobile platform. When mobile is included, Microsoft actually fell to third place in 2015 in installs of devices shipped worldwide (including laptops, PCs, smartphones and tablets). The stated goals of the DOJ were also accomplished without the need for a breakup. Today, Microsoft makes applications for macOS, Android, and iOS. This also proves that Microsoft did not control vital infrastructure that its competitors could not get around. Competition moved past Windows dominance of desktop computing and created a powerful market in mobile computing that Microsoft has been largely left out of.
The Microsoft case shows that competition in the tech industry is resilient enough to not need a breakup remedy, even when a company seems to completely dominate the landscape. Microsoft is still a powerful company, but its influence is waning as consumers turn more to mobile computing. Indeed, we don’t hear many calls for Microsoft to be broken up today, something one would expect if they had the kind of stable dominance and incentive to engage in anticompetitive behavior that would call for such a remedy.