11/23/19

           In America, policy-driven discussions over the past few years have become largely passé. In fact, close to zero of CNN’s “Most Popular Stories of 2018” are discussions of policy. These popular stories instead focus mainly on natural disasters, shootings, personal/corporate/government scandals, and celebrity deaths1. There is one scandal, however, that was big enough and far-reaching enough to bring policy discussions back into the mainstream.

This particular scandal broke out in March 2018, involving a UK-based consulting firm called Cambridge Analytica. The company was exposed for underhandedly collecting mass amounts of Facebook data from both online survey respondents and the Facebook-friends of these respondents2. Although Facebook was quick to distance itself from Cambridge Analytica3, the controversy resulting from this instance of misconduct has sparked a broader debate about the power of large technology corporations. Encouragingly, this debate mostly has not devolved into tribal grudge matches or petty personal attacks.

Many participants in the debate over the power of large technology firms are in favor of increased regulatory restrictions. Some prominent figures have even called for the breakup of certain big tech companies. A few of these figures include US Senators Ted Cruz4 and Elizabeth Warren5, and even Facebook co-founder Chris Hughes6.

           The argument that those in favor of breaking up these firms make is that large technology platforms (examples of which include Google’s search platform and Facebook’s social media platform) are so big, that they ought to be classified as monopolies under the Sherman Anti-Trust Act. The Sherman Act was passed in 1890 and is meant to protect market-competitiveness by banning cartels, trusts, and monopolies from operating in the US free market7. In the past, companies that have been found by a court to be in violation of this act have been broken up.

In order to evaluate whether the case for breaking up Facebook has any merit, it is crucial to examine companies that set anti-trust legal precedent.

STANDARD OIL: “THE DRAGON SLAIN8”

The first such company, the John D. Rockefeller owned Standard Oil, was broken up in 1911 because the Supreme Court found that the trust was “a combination in unreasonable restraint of inter-State commerce9.” At its peak, Standard Oil controlled as much as 87% of oil production in the United States10. The company’s sheer magnitude meant that it could secure preferential shipping deals and engage in allegedly unfair practices, such as price-cutting with the intent to drive smaller competitors out of business11.

Although scholars continue to debate whether Standard Oil was truly in violation of the Sherman Act12, it is hard to argue with the results of this breakup. After the dissolution, the 34 companies that Standard Oil split into became more innovative and all of them doubled in value within a year1314.

The increase in value that came from competition made J.D. Rockefeller, who retained shares in all 34 spinoffs, America’s richest man of all time—a record which remains unbroken15. In 1912, considering the success that increased competition brought to the market, Theodore Roosevelt remarked: “No wonder that Wall Street's prayer now is: 'Oh Merciful Providence, give us another dissolution.’16”

THE NETWORK EFFECT

Standard Oil was a traditional monopoly, and as a result, its breakup was fairly simple: there were now 34 companies selling oil where only one existed before. Modern online social media platforms would not be able to break up as easily. This is because the nature of how networks gain and lose value results in an inherently different economic incentive structure.

The ‘network effect’—or Metcalf’s law—states that the more users networks have, the higher their value17. Interestingly, this idea is thought to have been first espoused by AT&T president Theodore Vail in the one of the company’s annual reports. He argues that companies like AT&T (American Telephone and Telegraph) naturally lend themselves to market monopolization because of the network effect. His argument entails the idea that so called ‘natural monopolies’ are good for consumers because they provide standardization (ex. Anyone with a telephone can call anyone else with a telephone because all phones operate on the same AT&T network).

A telephone—without a connection at the other end of the line—is not even a toy or a scientific instrument. It is one of the most useless things in the world. Its value depends on the connection with the other telephone—and increases with the number of connections18.

                                                     -AT&T President Theodore N. Vail, 1908

Vail went on to call AT&T’s Bell telephone system “the nervous system of the business and social organization of the country19.”

           Although Vail was speaking about more traditional networks, his theories can also be applied to online social networks: the more people who are on a network, the more attractive that network becomes to everyone who uses and joins it. This is because the number of connections that can be made on that network increases exponentially with each new entrant20.

This is true on Facebook, for example, where people commonly join because their friends, family, potential employers, teams, etc. all use it to connect. In other words, Facebook is a natural monopoly because having everyone on one platform is useful.

           Real-world market share is consistent with the ‘natural monopoly’ theory. In 2017, it was estimated that 94% of social media account holders had an account on Facebook or one if its subsidiaries, like Instagram21. At its peak in 1939, 93% of phone calls in the US were made using AT&T’s Bell system22.

           Given the similarities between these two networks, it is useful when considering anti-trust arguments against Facebook to look at what happened to AT&T.

ANTI-TRUST ACTION AGAINST AT&T

           The US government first considered taking anti-trust action against AT&T in 1913 after it began to consolidate the telephone market through a series of acquisitions. By 1921, however, Vail had convinced congress that “telephoning is a natural monopoly23.” As a result, an earlier ban on competitive acquisitions was reversed and AT&T began mass consolidation of the US telephone industry24.

For more than 60 years, AT&T was permitted to operate as a ‘regulated monopoly,’ but in 1974, the company again drew government anti-trust scrutiny. This time, the Justice Department charged AT&T with violation of FCC directives that were meant to preserve market competitiveness. One charge had to do with refusal to allow non-Bell made telephones to operate on the Bell network (consumers were forced to rent telephones from Bell itself), while another charge involved AT&T’s refusal to allow specialized carriers access to the Bell network25.

The courts ruled in favor of the Justice Department and starting in 1982, AT&T was required to spin off all 7 of its regional operating companies (known as “Baby Bells”)26. By 1992, the value of AT&T and all of the “Baby Bells” had more than tripled27. Today, the US telecommunications market remains highly competitive and is collectively valued at over 256 billion dollars28.

FACEBOOK: THE ANTI-TRUST ARGUMENT

           There are three questions, according to Harvard Professor Dipayan Ghosh29, that the US Department of Justice asks when deciding whether to take anti-trust action against a company:

1.     Is the industry competitive?

2.     Does the company use its excessive market share to exploit consumers?

3.     Could “fledgling firms” someday compete in the industry?

Because of Facebook’s estimated 94% market share30, Ghosh contends that the answer to the first question is an easy “no.”

           The second question is a little harder to answer. To its consumers, Facebook does not use exploitative pricing (it’s literally free), but Facebook has two ends. Facebook’s main source of revenue is selling user data and analytics so that advertisers can target their campaigns to specific ‘categories’ of people. According to Ghosh, Facebook sells this data with “tremendously high margins31.” Consumers pay a high price for this as well, as they give away their personal data. Because of the network effects at work here, consumers also have little choice but to make this trade. If they don’t, they are effectively cut off because there are few alternatives.

           Question three is especially difficult because it involves making a prediction. According Joseph Schumpeter’s classic economic theory of “creative destruction,” all monopolies must eventually either grow or die. According to Schumpeter’s theory, if monopolies fail to innovate and give consumers what they want, new entrants with new technology will rise to topple them. Schumpeter contends that all businesses operating in the free market are constantly reckoning with an ongoing existential threat. This incentivizes even monopolies to keep growing32.

What’s more, in the social media space, there are not huge infrastructural barriers of entry to the market like in the telecommunications and oil refining businesses. Anybody who can afford to run or rent basic server infrastructure, and knows computer programming, can start a new internet enterprise.

Facebook co-founder Chris Hughes acknowledges this, but his fear is that rather than compete with these startup companies, Facebook will either buy them (like they did with Instagram), block them from accessing Facebook’s network (like they did with Vine), or copy their features (like they did with Snapchat). While this is good for consumers in the sense that Facebook is adding features that they want to use, it is not great for future competition or innovation prospects33.

While venture capitalist firms will certainly fund a startup if they think that it will receive a large buyout, they will not fund a social media startup if they think that Facebook will simply copy its core differentiating features. VCs know that patent protection does not always work in these cases, as it didn’t in the case of Snapchat34. This is bad for innovation because it means that the fresh, new ideas that fledgling startup firms have are much less likely to get traction and funding.

Given the discussion of these three questions, it is certainly conceivable that the US Justice Department could pursue anti-trust action against Facebook. As evidenced by the huge increases in market value following the breakups of both Standard Oil and AT&T, there can certainly be positive effects from anti-trust enforcement. It is important to note, however, that this process needs to be done with care and precision.

A PROPOSAL

If, at some point in the future, the courts order Facebook to break up, the question of execution will be of paramount importance. From the AT&T case, we can see that network effects don’t necessarily preclude a successful anti-trust breakup. To prevent network effects from being a problem, however, implementation of the breakup needs to be very specific.

The single most important part of executing the AT&T breakup—and the policy that allowed the breakup to be possible even with network effects in play—was to make sure that all former Bell systems remained interoperable. Even though control was becoming decentralized, it was extremely important that the various providers be permitted to retain access to the Bell system. If this were not the case, many regionally-isolated phone networks would have existed until a new natural monopoly bought all of them up and connected them back together.

Similarly, if Facebook was split up with no interoperability between the new networks, it is likely—given the nature of network effects—that most users would eventually congregate on one of the spin-off social networks and neglect the others.

Just like interoperability of telephone networks means that anyone can call anyone, interoperability between social media networks would mean that users could log into any social media site and use it to communicate with people on any other social media site. This would promote competition in the industry both for social media users and for the various advertisers that follow these users.

In other words, the social media websites that provide the most desirability through desired features, data-management, user-friendly interfaces, etc. would attract the most users and therefore the most advertisers under this system. These dominant services would be ever-changing because users wouldn’t be locked in. If social media websites were interoperable, people would be free to switch as often or as rarely as they desired. It would be cheap and easy for new social media websites to enter the market, and users themselves would gain a plethora of choices regarding data management and new features.

Interoperability would require a central social networking API (Application Programming Interface). Any websites that wanted access to this central social networking API would have to pay a small fee to an intermediary company or non-profit organization that would exist to maintain it. Social networks would then get the opportunity to partner with the API organization to build in new features. This intermediary company would, of course, have its finances and data-collection practices subject to extensive independent review to avoid corruption.

CONCLUSION

           Before I started researching for this essay, I was unconvinced that an anti-trust breakup of Facebook would do much of anything to change the lack of competition in the social media space. After review of historical/legal precedent and economic theory, however, I have changed my mind. I think that the Justice Department would have a real case against Facebook, and I also think that proper implementation of a breakup could be largely positive for all parties involved. Facebook says that it is unfair to punish success35. I agree, but I don’t think breakup would be a punishment. Who knows? It could make Mark Zuckerberg the next J.D. Rockefeller.