The mantra of “breaking up Big Tech” is increasingly a rallying cry for members of both parties, but it specifies neither the actions required to dismantle companies nor the complexities and costs associated with doing so. Unfortunately, antitrust is a slow, messy remedy that would require ongoing and costly regulation that, at the end of the day, might not solve the underlying concerns about today’s tech giants.
This continues our series of essays about how a possible antitrust case against Facebook in particular might work. The first installment focused on the question of what the “relevant market” is, while the second installment explored the question of consumer harm.
Previously, we discussed that Facebook is not a monopoly under the traditional understanding of the term, and we concluded the firm offers the public benefits that likely outweigh any alleged “harms.” We also cautioned against assuming that differing preferences regarding technology and its use amount to harm. In this installment in the series, we consider the challenges of applying traditional antitrust remedies to Facebook in an attempt to address the various concerns raised by critics.
While the current “techlash” may not be as intense among the general public as current political rhetoric would suggest, many people are genuinely concerned about privacy, content moderation policies, and the influence of social media on society more generally. Even if we assume those issues warrant greater attention, it is unlikely that an antitrust-based approach would properly address them.
If history is any guide, an antitrust case against Facebook will also take a long time to settle and may never reach a practical remedy. For example, earlier antitrust crusades against IBM and Microsoft took 13 and 11 years respectively, and nothing much came out of those cases in the end. In both cases, the most import “remedies” were the unexpected changes that happened in the computing markets.
In the IBM case, antitrust officials focused on market power in the mainframe market and missed the personal computing revolution with the rise of firms like Apple and Microsoft, as well as the emergence of many Japanese computing competitors. When the antitrust spotlight later fell on Microsoft, the rise of the internet and emerging online markets were a secondary consideration. But by the time the case was winding down, Google, Amazon, and others were starting to take off, and the smartphone and social media revolutions were about to begin.
If a Facebook case takes a decade to settle, whatever remedy comes out of it will likely be irrelevant with the ever-increasing pace of technological change. Like the IBM and Microsoft cases, markets and technology will certainly evolve and shift power more in a decade than the law will.
While the IBM and Microsoft cases took a long time and are largely considered failures, consider the somewhat speedier and more “successful” antitrust action against AT&T which began in 1982. It, too, holds lessons for how well-intentioned structural remedies sometimes fail to improve competition and consumer welfare.
The AT&T breakup focused on structurally separating long-distance (AT&T) and local telephone service (“Baby Bells”). Antitrust officials hoped to create more competition in the long-distance market while containing the regional Bell monopolies and the supposed monopoly in local wireline telephone service. This antitrust decree gave Americans a little more long-distance competition for a time, but true communications competition came later with the rise of cellular telephony and internet-based services. Everything fought over in the AT&T antitrust case became a sideshow thanks to technological change. Nobody talks about “long-distance” phone service today, after all.
While the case did not produce much new innovation or competition, it did produce an epic amount of paperwork in the form of ongoing court and regulatory agency filings (at the federal and state level). For the decree to work as designed, the artificial firewall between local and long-distance services had to be preserved, and new innovations and integrations by the local Baby Bell companies had to be strictly limited. For example, there were epic legal and regulatory fights over everything from pay phones and paging services to phone books! Much of the ongoing enforcement effort fell to a single US District Court judge (Harold H. Greene) to determine what was and wasn’t permissible under the AT&T consent decree. Industry lawyers were paid handsomely to figure out how to make winning arguments in front of that judge, as well as before the Federal Communications Commission and state regulatory agencies that oversaw other post-divestiture regulations.
It is hard to know what specific remedy antitrust officials would choose for Facebook, but let’s imagine they broke up Facebook into several smaller entities—let’s call them “Baby Books”—or at least forced the spin-off of specific apps or services, like Instagram or WhatsApp. Following divestiture, officials would need to make sure that the old Facebook—let’s call it “Facebook Classic”—would not be allowed to reintegrate spun-off functions on the grounds that the firm would leverage its supposed core monopoly to crush spun-off competition. In practice, that requires the imposition of line-of-business restrictions on Facebook Classic, similar to the firewalls the government created in telecom markets following the AT&T breakup.
What might this look like? Facebook might be forced to discontinue its encrypted messaging on its Messenger (or end the product altogether) in order to clear the stage for a freshly divested, independent WhatsApp. Similarly, courts and regulators could be tied up by the minutiae of ensuring updates to Facebook’s photo features do not impose upon an independent Instagram.
The problem with this plan is that markets and technologies often evolve in unexpected ways. Preventing the re-integration of certain capabilities (or even the creation of entirely new ones) requires sweeping, heavy-handed regulatory permissions and endless litigation. Antitrust officials and the courts (or perhaps a new specialized social media regulatory body) would need to constantly police innovation in digital media markets (just as Judge Harold Green did for AT&T and Baby Bells).
The enforcement problems witnessed in the analog era would be even more pronounced given the nature of digital technology. The new services or capabilities that consumers might want from Facebook Classic (perhaps because they already had them before) would be forbidden by the antitrust decree. That will have collateral damage. “Breaking up tech companies means breaking up teams and the underlying technology,” notes Will Rinehart of the American Action Forum. “Tech companies are tightly integrated firms, not holding companies, so the practical act of breaking them apart is far more difficult.”
This is bad news for the firms and their employees, who will have to be separated and forced never to work together again. It is also a potential disaster for consumers, who would be forced to look elsewhere to find functionality that they might have already enjoyed in the old Facebook model. This could raise costs to consumers both directly (service costs, i.e., direct charges for some services) and indirectly (search costs, including trying to find previously connected groups or services). Ironically, if these new entities charged prices for service, they would probably be accused of gouging consumers, who previously enjoyed unpriced access to an integrated suite of services.
If the firewalls between structurally separated entities were high and tight, many people—including, ironically, many proponents of antitrust action against Facebook—would likely then complain about the dangers of walled gardens and lack of interoperability among new services. Presumably, advocates of structural separation would suggest that this problem could be solved through data portability or interoperability mandates. That, too, requires complicated ongoing oversight and regulation.
The Federal Communications Commission imposed data interoperability requirements on AOL and Time Warner when they merged in 2001. Regulators were concerned with AOL’s supposed market power with its AIM Instant Messenger and, therefore, required that it be made compatible with other messaging systems to stop this supposed “monopoly.” At the time, FCC Commissioner Gloria Tristiani insisted that, “AOL’s dominant instant messaging platform should be interoperable,” because, “it’s an essential platform.” A group called the Instant Messaging Unified Coalition (which included Microsoft), lobbied for stiffer requirements, arguing that “consumers will not have free-flowing and open communications over instant messaging.”
The results of these regulatory efforts were underwhelming. Years of regulatory filings ensued in this matter, but ultimately led to very little—except, once again, a lot of work for lobbyists and lawyers. Instant messaging competition ultimately came not from rules or central design, but from entirely unexpected sources. Meanwhile, in October 2017, AOL Instant Messenger was sold off and then unceremoniously shut down. While millennials mourned for the good old days of AIM, other messaging products including Facebook messenger, GChat, Signal, and more had long ago taken over and improved on its primary purpose. As our Mercatus Center colleagues Christopher Koopman and Michael Kotrous noted, the AIM case study should “serve as a cautionary tale to those now calling for trust-busting or utility-style regulation of tech firms today.”
With all this in mind, would an antitrust action against Facebook improve consumer welfare? After all, for decades that has been the key purpose of antitrust law. Consumers are not always made better off by structural antitrust remedies, however. In a 2001 study on “The Failure of Structural Remedies in Sherman Act Monopolization Cases,” economist Robert W. Crandall of the Brookings Institution reviewed 95 major Section 2 Sherman Act cases won by the government or ending in consent decrees. He found that there was “remarkably little evidence that these cases and the relief that emanated from them had a positive effect on competition and consumer welfare.”
The same would likely be true if Facebook was structurally separated into smaller units because regulators lack the knowledge to predict or design “optimal” technology products or markets. Facebook has as many users as it does today not because of any nefarious scheme to exclude rivals, raise prices, or rig markets. Rather, Facebook has been successful because it gave the public a low-cost, easy-to-use platform where people could spontaneously come together and create communities of their own choosing. And they could do so backed by the assurances of Facebook’s “real name culture,” which facilitate trusted interactions. For all its faults, this has been Facebook’s crowning achievement—and it would all be forcibly ended by antitrust intervention.
Worse yet, a Facebook breakup does not necessarily solve other supposed pathologies people worry about. A breakup doesn’t solve the so-called “fake news” problem (which may be overhyped anyway); it doesn’t solve the Russian election meddling problem; it doesn’t address the problems with hate speech or extremist views; it doesn’t address privacy concerns; and it doesn’t solve the concerns about overuse or addiction. In fact, there is no guarantee that some of these problems would not get worse in a post-breakup world. As Tyler Cowen points out, a break up might actually make it harder for less well-capitalized spun-off entities to deal with concerns about problematic content or other security issues that critics and consumers alike want addressed.
None of this will likely temper the appetite on Capitol Hill for action. Sen. Josh Hawley (R-Missouri), who sits on the Judiciary Committee's Subcommittee on Antitrust, Competition Policy and Consumer Rights, recently referred to Facebook as “a parasite,” and a “drug that hurts its users.” He blamed the site for a litany of societal problems and said, “Maybe we’d be better off if Facebook disappeared.” If extremist arguments such as these become the basis of antitrust intervention, proponents of a breakup are going to be sorely disappointed when many of these same pathologies continue to exist afterwards.
Antitrust remedies are not costless solutions. They often take a great deal of time to advance, require a significant amount of ongoing regulation to enforce, and typically do not forecast game-changing technological changes (especially in fast-moving markets like technology). They often fail to address specific concerns as they take a broad, blunt, and forceful approach to reordering markets. With that in mind, it is unclear how, if at all, consumer welfare would be improved or specific policy problems solved through sweeping remedies that are akin to using a sledgehammer to do a job better left to a scalpel.
In our next installment, we will consider some more targeted approaches to the concerns raised by Facebook critics that could serve as an alternative to antitrust action.
Photo credit: AFP/Stringer/Getty Images