Breaking Up Is Hard to Do—So Don’t Do It

Joe Kennedy October 7, 2020
October 7, 2020

(Ed. Note: The “Innovation Fact of the Week” appears as a regular feature in each edition of ITIF’s weekly email newsletter. Sign up today.)

Congressman David Cicilline (D-RI), chairman of the House Judiciary Subcommittee on Antitrust, Commercial, and Administrative Law, issued a staff report yesterday on competition and the digital economy. The report is the product of more than a year of investigations and seven congressional hearings, including testimony this summer from the CEOs of Amazon, Apple, Facebook, and Google.

The report represents a long-overdue examination of how antitrust law should be applied to the largest Internet companies. It is important that Congress and regulators understand both the challenges and the opportunities information technology poses for competition policy. Unfortunately, many of the report’s conclusions reflect the majority Democrats’ starting point: a hostility toward big companies, a strong suspicion that their success comes from anticompetitive behavior rather than from providing the best services and constantly improving them, and a view that antitrust policy should protect competitors. So, rather than carefully applying antitrust law to address clear problems, Cicilline and his Democratic colleagues would seek to restructure markets to achieve significantly more competition. Such a broad attack would harm consumer welfare and inhibit innovation.

Throughout the report, it is clear that the drafters see more competition as the end goal of antitrust policy and view with suspicion any markets that don’t have a multitude of competitors. This appears to be a result either of a lack of understanding that most innovation industries, as noted innovation economist William Baumol has shown, are somewhat concentrated, precisely because that is how innovation is maximized, or due to a blatant disregard for that fact in pursuit of the goal of limiting the size of U.S. firms.

The majority report contains many recommendations, including legislation to restrict large companies from making acquisitions, increase data portability, and prohibit Internet platforms from selling their own products alongside goods and services being sold by third parties—which would only serve to protect some competitors, not consumers or overall innovation. A company like Amazon thus would either have to sell its marketplace or cease selling its own branded products, including books and music. In the case of Apple, it would mean either not offering free, preloaded apps on its iPhones or giving up control of its App Store. The recommendation mirrors a similar proposal that Sen. Elizabeth Warren (D-MA) put forward during her presidential campaign. She justified her proposed rule by arguing: “You don’t get to be the umpire and have a team.”

The proposal to prevent companies from selling on their own platforms is a mistake for many reasons, most importantly because would reduce total economic welfare and competition. First, consider who benefits from the way platforms operate now. On one side of the equation, Internet companies themselves obviously benefit from their own platforms. The more transactions they facilitate over the same infrastructure—for themselves and for third parties—the more they lower their average operating costs. More product and service offerings also attract more consumers. And more third parties for using the platform mean more sales commissions. (Interestingly, those commissions reduce at least somewhat platform operators’ incentive to compete with third parties, because profits that platforms make selling their own products and services are partially offset by lost commissions they would otherwise earn.)

On the other side of the equation, platforms also benefit third-party sellers ranging from small, family businesses to large, established companies. These providers get access to many more potential buyers than they could ever connect with on their own. The commission they pay is at least partially offset by savings on advertising and running their own websites. Some platforms, such as Amazon’s marketplace, also offer to take over services such as billing and logistics, allowing sellers to scale up faster and concentrate on making the best products they can. The vast majority of third parties benefit greatly from the platform, which is precisely why the platforms have grown so fast. Meanwhile, almost all sellers work through more than one platform.

Finally, platforms benefit consumers by letting them search offerings from a large number of possible suppliers on one site. This makes price comparison much easier and reduces fears about whether the seller is trustworthy. Economist Thomas Philippon has estimated that the benefits of online commerce are equivalent to a permanent increase in consumption of 1 percent. This is partly because the increase in online commerce prices has been running at 1 percent below general inflation, saving buyers millions of dollars.

When a platform operator competes with one of its third-party providers by offering its own product, consumers again benefit. They now have additional options, possibly of better quality or lower price. What Amazon does in competing with suppliers is no different than what most major brick and mortar retailers do when they advertise and sell their own branded product right next to those of their competitors. And in these cases, the retailers collect and analyze real-time electronic sales data of products to decide whether it makes sense to sell a similar product under their own brand. Why is doing this online any different than offline? The committee report anticipated that objection—having heard it many times—but sloughs it off by making the inaccurate claim that big bricks-and-mortar retailers don’t have as much data as online commerce retailers. With point-of-sale terminals and mainframe computers keeping track of sales and inventory, this claim is just wrong.

The majority report’s recommendation to prevent companies from selling on their own platforms stems from the fact that platform operators have information on which products are especially popular and profitable, and they can use this information and their control over the platform to handicap the sellers of those products while they offer their own version. Merchants claim to fear that unfair competition from platform owners will either put them out of business or force them to sell out. These fears are buttressed by a few anecdotes of past cases.

But these fears are exaggerated. First, many of these anecdotes don’t hold up under careful scrutiny. In one of the most widely cited, Amazon’s purchase of diaper merchant Quidsi, Amazon actually lost a great deal of money selling below cost because, despite its size, its margins in what continued to be a highly competitive market were not high enough to recoup the investments it had made. These anecdotes represent only a small fraction of the sales and sellers on the platform. The vast majority of third-party sellers benefit substantially from the platforms. Finally, where sellers have legitimate complaints, remedies far short of breaking up the company are available.

A platform’s long-term interest is in building a reputation for good service and profitability among sellers. Treating sellers poorly will lead the best of them to turn their attention to other platforms.

It is not surprising for-profit competitors want Congress to give them an even better deal than they have now. What is surprising is that members of Congress should favor specific competitors rather than consumers. For 40 years, antitrust policy has been guided by the principle that regulators should maximize consumer welfare by protecting competition even if the workers and owners of other competitive companies suffer. Chairman Cicilline appears to be guided by a new approach to antitrust policy—fashioned in the image of the early 20th century Supreme Court Justice Louis Brandeis—which is inherently biased against large firms and seeks to protect companies from competition. Embracing this doctrine might help a few companies, but it would hurt the overall economy and the vast majority of Americans.

Moreover, much of the subcommittee’s work is motivated by the feeling that Internet companies have become too dominant. But we have heard similar complaints before only to see leading firms lose out to competition. In the retail space, A&P in the 1930s, Sears in the 1960s, and Wal-Mart during the 1990s were frequently described as being too dominant and having permanent market power. Although the latter still holds a large market share, it faces heavy competition from Costco, BJ’s, Amazon, Target and others. Apple’s iOS software for mobile phones has less than a quarter of the global market but still attracts a lot of criticism. Meanwhile, Blackberry and Nokia, which at different times seemed to own the future of phones, are no longer factors.

It is possible that certain changes would create benefits for both third parties and consumers. It seems reasonable, for example, to expect a platform to publish clear and consistent rules and to make those rules transparent to third parties. Perhaps third parties should also have an efficient means of settling disputes. But we don’t have to rewrite existing antitrust laws to accomplish that. So, let’s not.