Washington Vs. Big Tech: Should You ‘own’ All Your Social Network Data? An AEIdeas Online Symposium

In a provocative New York Times op-ed last June, the University of Chicago’s Luigi Zingales and Guy Rolnik wrote about the “increasing monopolization of the technology industry” and offered a pro-competition policy solution other than the usual suspects of antitrust action or turning the big platform companies into heavily regulated utilities. Definitely read their entire note, but here is the core of their argument in favor of a “Social Graph Portability Act”:

For a 21st-century problem, we suggest a 21st-century solution: a reallocation of property rights via legislation to provide more incentives to compete. In fact, the idea is not new. Patent law, for example, attributes the right to an invention to the company a scientist works for, to motivate companies to invest in research and development. Similarly, in the mobile industry, most countries have established that a cellphone number belongs to a customer, not the mobile phone provider. This redefinition of property rights (in jargon called “number portability”) makes it easier to switch carriers, fostering competition by other carriers and reducing prices for consumers.

The same is possible in the social network space. It is sufficient to reassign to each customer the ownership of all the digital connections that she creates — what is known as a “social graph.” If we owned our own social graph, we could sign into a Facebook competitor — call it MyBook — and, through that network, instantly reroute all our Facebook friends’ messages to MyBook, as we reroute a phone call.

If I can reach my Facebook friends through a different social network and vice versa, I am more likely to try new social networks. Knowing they can attract existing Facebook customers, new social networks will emerge, restoring the benefit of competition.

Today Facebook provides developers with application-program interfaces that give them access to its customers’ social graph, Facebook Connect and Graph A.P.I. Facebook controls these gates, retaining the right to cut off any developer who poses a competitive threat. Anticipating this outcome, very few developers invest seriously in creating alternatives, eliminating even the threat of competition. By guaranteeing access to new customers’ data and contacts, a Social Graph Portability Act would reduce the network externality dimension of the existing digital platforms and ensure the benefits of competition.

Certainly an intriguing and creative approach. And at least on first take, one both less intrusive and more practical than those other frequently mentioned policy fixes. To explore it further, I asked for the insights of experts both in and outside of AEI.

Mark Jamison, an AEI visiting scholar and director of the Public Utility Research Center at the University of Florida’s Warrington College of Business.

Should information technology companies be required to give you what they know about you, and then forget you? Not if you want great information technology.

One of the flaws in Zingales’ and Rolnik’s thinking is an assumption that data porting is worth the cost. This is demonstrably wrong: Software that gathers and exports information from social media already exists, meaning that it would be a small step to have an app that does this on an individual basis. Since no entrepreneur has found it profitable to develop such an app, it appears that costs exceed value. Of course if such an app does emerge in the future, then the regulations wouldn’t be needed.

The data porting proponents also misunderstand the situation. They think business data is a “21st-century problem.” But the situation is neither new nor a problem. Businesses have been gathering and using facts about customers for centuries and doing so makes businesses better. What has changed is the data value proposition: The cost of data gathering has fallen considerably over the past few decades, and at the same time our analytical capabilities have grown exponentially. These two changes have led companies to gather and analyze large amounts of data. This creates value, so why try to hinder it?

Zingales and Rolnik call this new data economics a “network externality.” It is not. A network externality exists when two or more parties value interacting (a network effect), but fail to do so as much as they should because the market fails to adequately compensate each participant for everything he or she contributes.

Even if Zingales and Rolnik were correct that there are externalities in Facebook, for example, their proposal would not be a solution. The solution to a network externality makes the network bigger, not smaller.

Lastly, the analogies of patent law and phone number portability work against the data porting proposal: The patent analogy shows that economic incentives for creating knowledge are important, so it is hard to understand why it follows that they should be destroyed. And if you cancel your phone service, your old phone company keeps all of the information it gathered on you.

Joshua Gans, professor of strategic management at the Rotman School of Management, University of Toronto, and blogger at digitopoly.org.

Recently, Luigi Zingales and Guy Rolnik proposed to pre-emptively deal with potential market power from social media by advocating for social graph portability. Rather than direct regulation of market power, this proposal takes a rights approach. As they note, this has happened before: “[I]n the mobile industry, most countries have established that a cellphone number belongs to a customer, not the mobile phone provider. This redefinition of property rights (in jargon called “number portability”) makes it easier to switch carriers, fostering competition by other carriers and reducing prices for consumers.”

I am familiar with these types of policy having initially proposed it for cellphone numbers 18 years ago and saw it since become adopted to great success throughout the world.

The basic premise of the argument is that network effects make it hard for consumers to switch to new entrants. Switching costs was the basis of cell number portability but you could call anyone on any network, the costs were simple. A social network like Facebook would be a different story. If you were to switch to another social network, let’s call it Newbook, you could not read your friends posts on that network and your friends would be unable to read your posts on Facebook. Even if it were far superior to Facebook for some (or all) consumers — Newbook would not get many (or any) customers.

The barriers are not as strong as that. Facebook, for instance, gives third parties access to consumers’ social graph. But they can cut that access off at any time (something Twitter had done).

What Zingales and Rolnik want is for that social graph to be controlled by consumers who can choose where to re-route their communications on any network. So, if I want my posts to go to you and to see your posts, so long as we have a link, it would not matter which network we were actually on.

In practice, what would this mean? You might have a “social identity” (akin to a phone number) and would then form bilateral links with others and those links would be recorded publicly. You would then modify those links as need be. That said, the last time we did something similar we got marketing calls! The point being, one value of Facebook, in particular, is the way permissions work. It is not readily obvious how this would work in their absence.

It gets even more complex when we think about people’s private and public identities. Some of my Facebook posts are public and I read many public posts from the media, fan groups and companies. That is all part of my social graph but how would we work all of that? There may be solutions there but the larger issue is that this is constantly evolving yet having a consumer controlled social graph may make it difficult to be responsive. After all, think about how you manage the social graph that is your pre-programmed fast dial numbers on a phone (if you even do those things). They quickly go out of date and you can’t be bothered updating them.

The practical issues of porting an entire social graph aside, Zingales and Rolnik are on to something. There is, in today’s world, a need to clarify what data a consumer owns. In terms of social graph, consumers surely have a right to share information they have provided Facebook with others and Facebook should probably make that easy even if it falls short of some portability proposal. True consumer power would come if you could also exclude how your data is used. The porting proposal doesn’t even touch that . . . yet.

Daniel Lyons, an AEI visiting fellow and an associate professor at Boston College Law School.

Zingales and Rolnik deserve credit for thinking creativity about competitive challenges facing the digital economy. They are right that switching costs can give incumbent social networks a significant advantage over upstarts. Net neutrality proponents argue that paid prioritization might deter the “next Facebook” by increasing the cost of entry. Yet this hypothetical toll pales in comparison to the 1-billion-user head start that Facebook has over potential competitors.

But I do not share their pessimism regarding the law’s ability to discipline these incumbents. Facebook’s size is not inherently problematic. The law should be concerned only if the company exploits its dominant position to engage in anticompetitive behavior. Zingales and Rolnik argue that because antitrust law focuses on consumer harm, it is ill-equipped to sanction companies who give their wares to consumers for free. But Facebook is far from the first company to barter entertainment for valuable consumer information. Radio and broadcast television pioneered the advertising-based revenue model long ago. And when these companies behave anticompetitively, the law corrects them just as it does any other segment of the economy.

Nor do I share their optimism that social graph portability is a superior solution. Admittedly, number portability improved competition among telephone companies by reducing customers’ switching costs. But social media is different from the telephone network, with larger technical challenges to interoperability. Each telephone company sells the same service — voice communication over the public switched telephone network — which facilitates interconnection between competitors. While social media companies offer similar services — the ability to exchange messages with peers — each platform is slightly different. Zingales and Rolnik envision a world where I can sign in to a Facebook competitor and “instantly reroute all [my] Facebook friends’ messages to” that new service. But it’s hard to imagine how a Facebook message can be rerouted to Twitter, for example, with its 140-character limit, or Instagram, which trades primarily in images.

This observation raises a more fundamental concern: the effect of portability on innovation. The authors discuss empowering a new “MyBook” that delivers essentially the same product as Facebook. Social graph portability encourages this homogenization so that messages can port easily between platforms. But society may be better served by competitors who differentiate their products from Facebook, offering services to niche customers unsatisfied by incumbents. Differentiation expands the market and ultimately drives innovation, but this competitive dynamic could be retarded by the pressure to conform to portability standards.

Hal Singer, a senior fellow at the GW Institute for Public Policy, principal at Economists Incorporated, and an adjunct professor at Georgetown University’s McDonough School of Business.

To evaluate whether data portability is a good remedy for dealing with dominant tech platforms, we should step back and assess the two broad approaches to dealing with monopolies generally: (1) attack the monopoly itself or (2) protect ancillary markets that are vulnerable to monopoly extension. Data portability fits squarely in the former, alongside other remedies being floated such as a breakup that would follow on the heels of an (uncertain) antitrust litigation.

To its credit, data portability has worked remarkably well in other contexts, including in mobile telephony. Indeed, according to one estimate (my own), the annual savings for US consumers reached $10 billion since the Federal Communications Commission mandated number portability in 2004.

But I am skeptical of its efficacy here for two reasons. First, the network effects of a dominant social-media platform are more powerful than are those for a dominant phone operator. There is nothing important about the mobile network to which my mother subscribes; data portability in mobile was about reducing switching costs for consumers — a tractable problem. In contrast, data portability won’t solve the massive coordination problem associated with moving all of my friends in unison to a new social media platform. The prospects for breathing life into what Zingales and Rolnik call “MyBook” seem dim.

Second, even the antitrust laws tolerate a monopoly that was legitimately obtained. What’s offensive in antitrust is conduct that artificially maintains that monopoly or extends it to an ancillary domain. Data portability is an attack on a monopoly for essentially being a monopoly. Assume arguendo that Google and Facebook engage in no anticompetitive conduct. Should a firm that invents the best mousetrap — and then refrains from engaging in anticompetitive conduct — be penalized by having to surrender its secret ingredient (here, its data) to its rivals?

Accordingly, I am willing to tolerate Facebook’s or Google’s monopolies in social media and search, respectively, so long as the government can protect innovation at the edge of their dominant platforms. This is another way of saying that regulation should attack the anticompetitive conduct (the sin) and not the underlying monopoly (the sinner). Facebook and Google both make life miserable for app and content providers that operate at the edge of their platforms. To address those bad acts, we could create a new forum (the “Net Tribunal”) to police discriminatory conduct by dominant online platforms. As I’ve written here and here, such a forum would plug a gap in antitrust protection, which tolerates the mild forms of discrimination practiced by the tech platforms.

Bret Swanson, an AEI visiting fellow and president of Entropy Economics.

Even if we agreed on the need to constrain Silicon Valley’s tech titans, Professors Zingales and Rolnik are correct that price regulation and antitrust would not likely be effective tools. They propose instead a “Social Graph Portability Act,” which would allow users of Facebook, for example, to take their data to another social network. They liken it to porting your phone number to another mobile provider.

Portability is superficially more attractive than heavy-handed regulation. But it’s not clear what problem it attempts to solve, or that it would yield more benefits than complications. I can think of lots of reasons not to legislate in this arena:

1. Do we really need to regulate one of the most innovative sectors of the economy? Perhaps Silicon Valley’s dynamism, relative to other sectors, is a result of experimentation without Washington’s direction.

2. Facebook already provides an Open Graph API (application programming interface), which allows other apps access to much (if not all) of the information Zingales and Rolnik seek. You’ve probably noticed new apps asking if you want to sign in with your Facebook or Google login and invite your connections.

3. Would social graph portability apply only to Facebook, and perhaps Twitter and Snapchat? How would it resolve the broader Silicon Valley tech titan “problem,” to the extent one exists? Would the professors force Amazon to share customers’ purchase histories for books and gadgets? Should Google and Microsoft share search histories, and Netflix share video preferences? Must Apple open its App Store?

4. A social graph is more complex than a 10-digit telephone number. It’s a vast web of information accumulated over time. Portability of the social graph — if it contains all your interactions with other users — would present some problems. What if user A wants to take her graph to “Mybook,” but user A’s connections — users B, C, D — do not want their information to appear on Mybook, or for Mybook to see it? Does user A then have to acquire consent of B, C, and D? Portability makes sense for a simple identifier that’s attached just to you, like a phone number; but for a vast collection of data? If the Social Graph is just the list of people to whom you are connected, then it’s not very powerful. I can make a list of my Facebook connections today and invite them to join another network.

5. Zingales and Rolnik are correct that apps and firms in this arena can grow very large very fast, and even “monopolize” a narrow market for a time. Everyone wants to use the service that everyone else uses. Network effects, however, work both ways. The fax machine was also driven by network effects. But when the internet and digital documents arrived, fax machines vanished nearly as quickly as they had proliferated.

6. New decentralized peer-to-peer technologies, such as blockchains, will soon begin disrupting the “walled gardens” of today’s tech titans, especially eroding the advertising-centric models. The idea that today’s titans are impervious to upstart competition is anti-historic, provided we recommit to policies that encourage innovation.

The sad irony is that Big Tech unknowingly invited these new regulatory proposals with its urging, beginning 15 years ago, for Washington to regulate the internet’s physical infrastructure: net neutrality was just another form of mandated open access. Even more recently, Silicon Valley has been asking Washington to block others from getting into the digital advertising business, which is the way it monetizes its social graphs. We predicted Silicon Valley’s calls to impose net neutrality would boomerang into search neutrality and social media neutrality, and now here we are.

Gus Hurwitz, an AEI visiting fellow and an assistant professor at the University of Nebraska College of Law.

Luigi Zingales and Guy Rolnik’s recent New York Times op-ed — in which they articulate establishing property rights in users’ social data as a solution to social media firms’ seeming market power — appears on its face to be both a novel solution and one that should appeal to economically-minded regulatory minimalists. After all, as any good scholar of law and economics can tell you, regulation is best justified by market failure, and markets most often fail due to poorly defined property rights. Consequently, clearly defining property rights in social-media data is an important step toward enabling a robust market in social-media information, and an important bulwark against potentially disruptive and market-distorting regulatory intervention.

There are only two problems: This idea is neither novel nor a meaningful solution to the problems animating Zingales and Rolnik’s concerns. As an initial matter, the idea they propose is basically the same idea that has popped up in the academic literature every five years or so since the advent of social media. Their proposed “Social Graph Portability Act” is essentially a call for data portability, with a whiff of an interconnection obligation lurking in the background. Such ideas were all the rage around 2001 (prompted by arguments that the EC Data Protection Directive required it). They were again popular around 2007–2009, this time following Facebook’s ascent in market share. Around 2012 we started to see scholars looking at these issues through an antitrust lens, frequently considering how antitrust (which is notoriously and myopically obsessed with price effects) could ever address markets which are characterized by non-price considerations. (Hint: Antitrust is quite able to deal with non-price considerations.)

There is a greater problem with Zingales and Rolnik’s proposal: It is actually about de-propertizing existing property rights, not just “reallocating” them. Their core idea is that platforms like Facebook and Google should be required to open their application program interface (API) for accessing their social graphs (the interconnections of relationships between users) to other platforms. (Query whether their proposal would only apply to their disfavored firms, or whether it would apply industry-wide; query, too, whether it would only apply to API relating to social graphs or other API, and who determines this.) By saying that Google and Facebook don’t “own” this data, but must share it with their competitors, the idea goes, users will be able to move between various social networks, taking their social graphs with them.

Putting to the side that its social graph basically is Facebook — this idea wouldn’t so much allow other firms to interconnect with Facebook or limit the network effects of the social graph as it would dissolve Facebook as a going concern into a pool of social media acid. (And who then would invest in developing and maintaining the graph?) This proposal faces the fundamental problem that its net effect is to de-propertize data that currently is propertized: ab initio, this information is the property of individual users, who transfer it (on a non-exclusive basis, in fact) to Facebook through contract. The terms of that contract determine what the Facebooks of the world can (and cannot) do with it.

It is very likely the case that, as the world has been introduced to social media, users have tended to cede too much control over their information to social media and other platforms. It is also arguably the case that US law has traditionally made it too easy for platforms to obtain permission to use this information.

But the response to this — especially in markets that are governed by voluntary contracts and transfer of property rights — is for the market to improve the efficiency of the mechanisms that govern how information is transferred, used, and that use monitored. The complexity of the task is not appreciated by most academics or regulators, but ultimately these platforms have both the ability and responsibility to guarantee their users have the tools necessary to ensure the users’ informational rights are being respected.

On a key point I agree with Zingales and Rolnik: It is clear that if Big Tech doesn’t start tending to these issues, seriously and soon, it risks oppressive regulation. Indeed, Zingales and Rolnik’s proposal is one of the many forms that such oppressive regulation could take.