The latest bills currently debated in Congress regarding Big Tech—the US Innovation and Competition Act that was passed by the Senate this month and the five antitrust bills introduced in the House—are a mixed bag. In addition to welcome measures meant to increase merger enforcement, some of these bills could affirmatively harm competition, hurting both consumers and labor.
Congress is entertaining several important legislative proposals relating to technology. The US Innovation and Competition Act, passed in the Senate by a bipartisan majority, is “the most expansive industrial policy legislation in US history,” explicitly targeting China as a tech competitor. The other is a set of five House antitrust bills focused on large digital platforms.
The 2,300-page Innovation and Competition Act would promote the development of technology hubs and increase funding for the National Science Foundation and support for 5G technologies. It would also fund research and development in medical technology, robotics, and artificial intelligence. The Act would develop a task force to pursue antitrust violations by Chinese firms and fund “actions against anticompetitive business practices by Chinese companies.” It does not propose substantive changes to antitrust law.
Identifying antitrust violators should be an enforcement decision. Legislation targeting a particular country creates needless bias and rigidity. Rather, the statute should identify particular anticompetitive practices. If China is guilty of these and causes competitive injury in United States markets, then of course antitrust enforcement would be justified.
Nonetheless, the overall goal of the Innovation and Competition Act appears to be the same as that of antitrust laws: to promote maximum output consistent with sustainable competition. “Output” can be measured by quantity, quality, or innovation. Increased output benefits not only consumers but also labor, resellers, and other input providers.
The five antitrust bills introduced in the House are a mixed bag. The proposed Augmenting Compatibility (ACCESS) Act mandates data portability and interoperability. Networks become more valuable as the number of users grows larger, and competition among two or more rivals on the same network can be much more intense than competition between different networks. The classic example is the telephone network, which interconnects virtually everyone, but is operated by thousands of competing businesses. Properly executed interconnection arrangements can increase competition, encourage new entry, and give users the benefits of a larger network. In competitive markets interconnection is often voluntary because it can be in everyone’s best interest. Email is one example, where many different clients interconnect so smoothly that we rarely pay attention to the identity of our correspondent’s email provider. By contrast, dominant firms or firms facing pressure from firms in adjacent markets frequently resist interconnection. They would rather force customers to join their own network. For example, AT&T strenuously resisted interconnection prior to its breakup.
Interoperability requirements can be technically complex to design and enforce, but a quarter of a century of experience of policing similar requirements under the Telecommunications Act shows that it can be done and that it can facilitate a much more competitive marketplace.
Another House bill, the Platform Competition and Opportunity Act, increases merger enforcement against large digital platforms. One of the greatest impediments to platform competition has been their acquisitions of hundreds of small, nascent potential competitors, effectively preventing them from ever emerging as viable rivals. The FTC’s complaint against Facebook cites evidence that Facebook acquired Instagram for just that reason. This bill wisely permits mergers to be treated as exclusionary practices, rather than looking merely at the opportunities they create for collusive behavior or price increases.
One objection is that mergers can create efficiencies that benefit consumers. As acquiring firms are bigger and more diverse, however, they are more able to achieve needed efficiencies without merging. A rule that simply bans acquisitions by the dominant platforms might be a little overdeterrent, but not by very much. This provision permits an acquiring firm to prove by clear and convincing evidence that the merger will be harmless. That seems to be about the right balance. In any event, the desirable assets of small acquired firms are typically intellectual property rights and human capital (people). People can be hired away, and a covered platform that wants access to intellectual property can acquire a non-exclusive license. That gives the platform everything it needs to improve its own technology, but not the power to exclude others.
Another of the five bills increases merger filing fees, providing the Justice Department and the Federal Trade Commission with more resources for evaluating mergers. The agencies need more funds, but also for non-merger enforcement, which this bill does not address. Most of the contemplated actions against the large digital platforms are monopolization cases, which have always been particularly costly to bring.
Two of the House’s antitrust bills could harm competition. Both are vaguely drafted, and their precise effects are unclear. First, the American Choice and Innovation Online Act does some good things. It addresses preferential exclusionary agreements. One step that could make the large platforms more competitive is prohibition of distribution agreements that require suppliers to charge higher prices to competing firms. There is also evidence that Google uses its DoubleClick advertising auction site to favor its own assets, such as YouTube. These practices fall within the traditional definition of contracts in restraint of trade. Unless the agreements involve competitors, however, current law would require proof of power and anticompetitive effects. This statute imposes a harsher standard, although it permits defendants to prove that a challenged restraint is competitively harmless. A simple, focused injunction against such practices is usually the best remedy.
But the bill does not stop there. It also attempts to enforce a conception of “fairness,” ignoring why the United States has historically been a favorable environment for technology startups, including all of the large digital platforms in their infancy. The bill’s “nondiscrimination” provisions are reminiscent of the Robinson-Patman Act, passed in the 1930s during a deep depression. That statute was a misplaced and unsuccessful reaction to chain stores—another important commercial development that Congress did not understand at the time. Until the Supreme Court put an end to it, the Robinson-Patman Act required firms to treat their trading partners equally in both prices and other terms of sale. Congress simply did not understand the important role of markets in picking winners. Firms that deal in the goods of others must be able to encourage competitive performance. A firm that rewarded a successful supplier or dealer with better prices or terms invited a Robinson-Patman Act challenge from firms who did not get the same reward. The same thing would happen under this statute. It would be bad enough if enforcement were limited to the Federal Trade Commission and injunctive relief. This bill also provides for private challenges and treble damages.
Finally, the Ending Platform Monopolies Act is the most anti-consumer and anti-labor bill in this collection. While it does not expressly target Amazon, that is the principal platform whose business model falls within its coverage, although it also reaches discriminatory advertising provisions. The title targets “monopolies,” but just as the Robinson-Patman Act, it lacks a market power requirement. The only relevant requirement pertains to the aggregate size of the covered platform.
This provision would all but eliminate practices such as Amazon’s sale of its own batteries in competition with Duracell (owned by Berkshire Hathaway). It also requires that if Amazon sells a product, it must host on a nondiscriminatory basis any actual or potential competitor. This provision goes even further than the Robinson-Patman Act, which limited a firm’s power to discriminate among existing suppliers. It would require a firm to accept other suppliers as well. For example, there are currently 300,000 makers of handcrafted bar soap in the United States and 397 manufacturers of garden fertilizer. Because Amazon already sells these products, this bill would require it to sell the competing products of all of these other firms and on the same terms. It could no longer select, as all retailers do, among third party sellers. Imagine telling Walmart (which is a bigger retailer than Amazon) that if it sells kitchen cutting boards it must carry the competing cutting boards of all 165 American manufacturers. The only practical way a firm could comply would be to exit from markets with a large number of sellers. It would also create silos that would significantly limit a platform’s ability to compete on one of the other platforms. For example, Amazon could not offer Amazon Music on Apple devices in competition with Apple Music. This bill could do more harm to the US internet economy than China ever could.
There is evidence that Amazon, Facebook, Apple, and Google are committing anticompetitive abuses that require correction. A non-antitrust possibility is theft of intellectual property rights if a firm steals trade secrets or copies the products of third-party sellers. Another, which is covered by the antitrust laws, is most favored nation clauses. These require suppliers to ensure that competitors’ prices are as high or higher than the firm that is imposing the restraint. Such clauses directly harm consumers, labor, and others who benefit from high output, because they impose higher prices on everyone who is affected. The best antitrust tool for these, if proven, is the restraint of trade provisions in the Sherman Act. A good assist would be legislation applying to large network sellers that reduces the market power requirement for illegality and switches the burden of proof for defenses.
The ultimate goal should be to make digital sellers of all sizes more competitive, not to destroy them.
Disclosure: Herbert Hovenkamp serves in an unpaid capacity on the international board of advisors for the Global Antitrust Institute (GAI) at the Antonin Scalia Law School at the George Mason University, where he debates against the Federalist Society on antitrust issues. GAI has received funding from tech companies, including Google, Amazon and Qualcomm.
Editor’s note: The author’s disclosure has been updated after publication.