Over the last several years, Facebook has become a public good and an important social resource. But as a company, it is behaving badly, and long term, that may cost it: A spring survey found that almost half of Americans believe that Facebook will eventually fade away. Even the business side has been a bit of a disaster lately, with earnings lower than expected and the news that a significant portion of Facebook profiles are fake. If neither users nor investors can be confident in the company, it's time we start discussing an idea that might seem crazy: nationalizing Facebook.
By “nationalizing Facebook,” I mean public ownership and at least a majority share at first. When nationalizing the company restores the public trust, that controlling interest could be reduced. There are three very good reasons for this drastic step: It could fix the company's woeful privacy practices, allow the social network to fulfill its true potential for providing social good and force it to put its valuable data to work on significant social problems.
Let's start with privacy. Right now, the company violates everybody's privacy expectations, not to mention privacy laws. It also struggles to respond properly to regulatory requests in different countries. In part, this is because its services are designed to meet the bare minimum of legal expectations in each jurisdiction.
When users in Europe request copies of the data Facebook keeps on them, they are sent huge volumes of records. But not every user lives in a jurisdiction that requires such responsiveness from Facebook — U.S. users are out of luck because their regulators don't ask as many questions as those in the European Union and Canada. Privacy watchdogs consistently complain that the company uses user data in ways they didn't agree to or anticipate. There are suspicions that the company creates shadow profiles of people who aren't even users but whose names get mentioned by people who are Facebook users.
True, Facebook recently provided an opportunity for users to vote on changes to the interface. But the program seemed more like a gimmick designed to placate the most opinionated and tech-savvy users: It was not heavily promoted and not a serious effort to educate the public and survey opinion. Indeed, few people voted.
It would be better to have a national privacy commissioner with real authority, some stringent privacy standards set at the federal level, and programs for making good use of some of the socially valuable data mining that firms like Facebook do. But in the United States, such sweeping innovations are probably too difficult to actually pull off, and nationalization would almost get us there. Facebook would have to rise to First Amendment standards rather than their own terms of service. The company could be regulated the way public utilities often are.
With 80 percent of market share, Facebook is already a monopoly, and being publicly traded hasn't made it more socially responsible. The map of its global market dominance is impressive, though some might say this is a map of colonization. In its recent filing to the Securities and Exchange Commission, Facebook declared its goal of connecting all Internet users. The company actually wants to be public information infrastructure, and to that end its tools have been used for a lot of good, like encouraging organ donations and helping activists build social movements in countries run by tough dictators.
But Facebook can also make mistakes with political consequences. The company has come under fire for missteps like prohibiting photos of women breast-feeding and suddenly banning “Palestinian” pages at one point. Facebook communications are an important tool for democracy advocates, including those who helped organize the Arab Spring. Yet the user policy of requiring that democratic activists in authoritarian regimes maintain “real” profiles puts activist leaders at risk. And dictators have figured out how they can use Facebook to monitor activist networks and entrap democracy advocates.
But since the security services in Syria, Iran and China now use Facebook to monitor and entrap activists, public trust in Facebook may be misplaced. Rather than allow Facebook to serve authoritarian interests, if nationalized in the United States, we could make Facebook change its identity policy to allow democracy activists living in dictatorships to use pseudonyms.
While most U.S. citizens and most global citizens treat Facebook as their social network infrastructure, the firm is greatly understaffed: It has about 4,000 employees serving nearly 1 billion users. Facebook staffers — at least those in it for the social good, rather than the bonuses — might even welcome the move to nationalize. Currently, Facebook employees are tasked with discovering marketable trends, selling advertising and doing data mining in the service of profit. Nationalizing Facebook would allow more resources to go into data mining for public health and social research.
Many academics are finding that big social network data sets can generate surprising and valuable information for addressing social problems — for instance, public health and national security. Researchers are working on ways to use social networking patterns to predict the spread of sexually transmitted diseases. We could even use Facebook data to analyze criminal networks in the United States or terrorist networks around the world.
We'd want to be careful about the circumstances under which our security services had access to Facebook data, but if the firm had been nationalized, at least there could be some reasonable public oversight. And while academics have to meet ethical standards for protecting the people they study, Facebook has no such guidelines. Nationalization could allow us to review the ethical implications of their management decisions.
Facebook's data harvesting could be used to improve public policy, yet scholars rarely find the company willing to collaborate on important research questions. Sometimes different scientists unknowingly buy access to the same data, and then duplicate efforts to clean the data for real analysis. A publicly accessible, central way of sharing data would allow better access for social and public health researchers. Or even better, the National Science Foundation and National Institutes of Health could help manage a clearinghouse for data so that we can all be sure it is properly anonymized and research effort is not wasted. There would be economic benefits, too: As the venerable Economist magazine has argued, making large amounts of public data widely available stimulates creative new businesses ventures.
Users in some parts of the world might panic if Facebook becomes an official part of the U.S. government. But there are plenty of examples of good public investment in media and infrastructure. For instance, citizens around the world benefit from the BBC, and many governments use the public purse to support technology innovation and build up information infrastructure. The public policy benefits of scholarship with Facebook's “big data” would spread around the world. Having occasional access to anonymous profiles would help democracy activists living in dictatorships. The high — and globally consistent — privacy standards that could be swiftly implemented after nationalization would be good for everyone.
Even though the benefits outweigh the risks, the idea of nationalizing Facebook is probably a nonstarter. At the very least, though, it is a great thought experiment and way of putting the privacy and data-mining issues front and center. Forcing regulators — and the company itself — to think about the ways in which its organizational behavior might serve the public interest should make it a better company. Facebook is now public infrastructure, and it should be treated as such.
Howard is professor of communication, information, and international studies at the University of Washington. Currently, he is a fellow at Princeton University's Center for Information Technology Policy