The idea of an online social network is simple, even obvious—Kirkpatrick cites an article from 1968 expounding the idea long before it was technologically feasible. Such networks were already up and running—MySpace was the biggest—when Zuckerberg started Facebook. A user has a profile page on which he posts whatever information about himself he thinks might interest people with whom he has—or would like to have—a relationship. (The relationship need not be social, of course: as Facebook has developed, it often is a business relationship.) He gives them access to his page and they give him access to theirs. That is the basic model, but Zuckerberg made a number of decisions at the outset that proved to be prescient. The first (quickly abandoned, but important at Facebook’s beginning) was to make it a network just for Harvard students, because they would be intensely interested in each other and perhaps suspicious or dismissive of outsiders. The second decision was to require every Facebook user to register under his real name (and to provide an e-mail address to assist Facebook in verifying his identity): there would be no pseudonymous role-playing, as in so many online social networks.
Moreover, anyone who wanted access to a user’s Facebook page had to get the user to “friend” him, and the user could “defriend” him at will. In this way the user controlled access to his own site. This, in the universe of Facebook, is a precondition to intimacy, or “radical transparency,” among friends, and it has proved to be a big factor in the success of Facebook abroad. “You only see friends,” Kirkpatrick remarks. “Italy’s Facebook-using hordes, for example, could grow to many millions without often seeing anyone who wasn’t Italian.” To be a “friend”—this was another one of the prescient early decisions—you have to register on Facebook, and having become a Facebook user you may start inviting people to be your “friend”; but for them to become your “friend,” they must join Facebook and thereby become Facebook users as well, and in this way the network can expand exponentially..
Facebook was also determined to prevent interruptions in service. An earlier and rather similar online social network, called Friendster, had failed to gain a significant following largely because of service failures. (Timing is all: an online network could not “scale” until the software and associated technology had evolved to the point at which service failures could be minimized.) And finally, consistent with the presumed dignity of Harvard (though the standard uniform of Zuckerberg and his classmates was Nikes, jeans, t-shirts, and fleece vests), the Facebook format was and is sober, simple, uncluttered. It makes users feel sophisticated.
Facebook filled a void at Harvard. It is a large school and most of the students, especially those newly enrolled, know few other students. Facebook provided a free, rapid, and effective way of getting to know many more of one’s fellow students than would be possible through personal contacts. You got not only their name and picture, but also such personal information as what classes they were taking, where they were from, and what their interests were. And so Facebook spread like wildfire at Harvard.
Students at other elite colleges (some of which already had online social networks, but none so cleverly designed as Facebook) took note and soon were clamoring for Facebook. Its Harvard origin proved serendipitous: an online social network created by and for Harvard students had cachet. So Zuckerberg offered Facebook at the other elite colleges, and it took off at those schools just as it had at Harvard. Students at less elite colleges now began demanding Facebook. The question of whether their students should be allowed to join was a critical issue for the new company, the first of many that Zuckerberg faced. Some of his associates worried that the democratization of Facebook would tarnish the brand. But Zuckerberg decided to take a chance and open Facebook to all college students. It worked. The next question was whether to make Facebook available to high school students. This provoked heightened concern for the brand, but again Zuckerberg overrode the doubters and was vindicated.
Doubts soared when the question then arose whether to open Facebook to everyone. Zuckerberg’s co-venturers were worried not only about diluting the brand but also about where to find the money to buy the large number of servers that would be required to handle such an increase in users. Daily communications across the Facebook network could be expected to climb into the tens of billions. Advertising revenue was not keeping pace with the breakneck expansion in usage, in part because of restrictions that Zuckerberg—remember that he regards advertising as a necessary evil—imposed upon it. Unfazed, he decreed the opening of the network to everyone in the world at least thirteen years old, and the revenues caught up.
At the same time that the network was expanding, the programs running on it were multiplying. Zuckerberg decided to allow anyone to write programs that would run on the network (“applications”). The programmer would not need Facebook’s permission or have to pay a fee, although Facebook reserved the right to remove programs that it did not like. Soon there were many thousands of programs (for making gifts, for video, for this, for that) that a user could add to his or her profile. Facebook profiles could thus be optimized for an immense variety of uses, including business uses. A lawyer of my acquaintance estimates that his Facebook page, which links him to more than two thousand “friends,” has increased his income (at no cost, except the ten or fifteen minutes that he spends each day reading and posting entries) by at least $100,000 since he became a Facebook user. (The number of “friends” of an individual Facebook user is limited to five thousand, apparently to deter spammers.)
Zuckerberg has made mistakes along the way, and he has sometimes been slow to correct them. Facebook’s muchpublicized breaches of users’ privacy have hurt it. But although he is no great fan of privacy, Zuckerberg was from the start committed to preventing unauthorized access to a person’s Facebook page. This is easier said than done, however, owing to the Internet’s inherent insecurity, and to the tension between protecting privacy and maximizing advertising revenues. Facebook needs those revenues to finance its increasingly costly network—and what makes Facebook valuable to advertisers is precisely the immensity of the information about its users that is stored on its servers. Advertisers are avid for such information so that they can pinpoint their advertising. Kirkpatrick reports that Interscope Records, for example, wanting to advertise a song about cheerleading, “could be given a guarantee—its ad [on Facebook] would only be seen by college girls who are either cheerleaders or who have mentioned something about cheerleading in their profiles.” Interscope paid Facebook nothing for any “wasted” advertising—advertising to anyone outside the target audience. Advertising on Facebook has turned out to provide potent competition to the mighty Google, because Facebook has more information about its users from their Facebook activities than Google has about its users from their Google clicks.
There is no necessary compromise of personal privacy when Facebook merely aggregates the data that its computers find on Facebook pages into a demographic profile, such as cheerleaders and would-be cheerleaders. But Facebook got into trouble with a program called Beacon, which, unless the Facebook user opted out, allowed “friends” to see his purchase orders from select advertisers. Many Facebook users did not notice the opt-out feature. Kirkpatrick recounts the sad story of the man who bought a ring as a surprise Christmas present for his wife, but the report of the purchase order showed up on her Facebook page and spoiled the surprise. Perhaps worse, the report showed that he had gotten the ring at a 51 percent discount. Facebook canceled Beacon shortly afterward.
The meteoric rise of Facebook raises four general questions (none analyzed by Kirkpatrick, whose focus is on the struggle to make Facebook a viable business). How is it possible for a teenager, however brilliant, to create a multibillion-dollar online business in such a short time? How likely is such a business to flame out? What, if any, legal protection from competition should be given to the ideas that power these businesses? And how far will social networking erode privacy or have other social consequences, good or bad?
That a teenager should have invented Facebook—rather than, for example, the Harvard administration doing so—is not surprising. There have always been teenage prodigies. Teenagers can excel at activities that require only abstract skills, such as math or computer science, and they have experiences and therefore insights that adults lack. Zuckerberg had both the skills and the experiences, and both these assets were important to Facebook’s creation and success. His generation had grown up with computers and online networking (e-mail is an online network service, as are instant messaging and chat rooms), enabling him to understand the demand for a service such as Facebook better than the people running Harvard, who belonged to an earlier generation. He could even understand the latent demand of adults for such a service better, because he knew and they did not what the service might look like, given advances in computer science.
But this does not explain the vertiginous expansion of Facebook’s network. The explanation lies in what are called “network effects.” These effects are nicely illustrated by telephony. Suppose there is a single telephone company, at first with only one subscriber. The company’s service is initially valueless, because the subscriber has no one with whom to talk. So maybe the company pays people to join. Every time a person joins, the service becomes more valuable because subscribers can reach more people. If now a second telephone company starts up, it will have difficulty competing unless it offers a clearly superior service (or persuades the incumbent to interconnect with it), because its network is so much smaller and the value of its service therefore so much less.
It is the same with a social network. Although almost no one actually wants 500 million online “friends” (well, maybe some advertisers do), the bigger the social network, the more valuable it is to the members. Those Harvard kids had friends at other schools, parents, siblings still in high school, prospective employers, and therefore valued the expansion of Facebook to other colleges and then to high schools and then to everyone. Facebook had to be only a little better than other online social networks to leave them trailing in the dust. For the bigger it grew, the more valuable its service became relative to that of the competing networks. And because joining Facebook is free, the cost of switching to it from another online social network is negligible, especially because the time cost of learning to use Facebook and to move one’s social-network files is minimized by Facebook’s user-friendly design.
Zuckerberg was well aware of network effects. His objective, says Kirkpatrick, “was to overwhelm all other social networks wherever they are—to win their users and become the de facto standard. In his view it was either that or disappear.” That was the business reason for his pushing Facebook relentlessly—and riskily—into more and more markets without waiting for his financial base to firm up.
Another reason that Facebook could grow so big so fast is that online network services are, by big-business standards, cheap to create and to operate. Facebook was launched with an investment of $2,000. Online networks do not require expensive plants (they do not manufacture anything) or huge workforces (they do not have retail outlets and most of their work is automated). They also do not require special terminal equipment: everyone who might be interested in joining an online social network already has access to the Internet from his laptop or mobile phone. So Facebook can get by with 1,400 employees, compared to more than 200,000 for General Motors. Owing to its small size, it has not yet been forced to make an experienced business manager its chief executive officer, or to go public (or sell out to venture capitalists) in order to raise needed capital.
But the cheapness of an online network is a source of competitive weakness as well as strength. When it is very costly to duplicate a network, the first one that comes into existence is likely to obtain a natural monopoly; a second would have to invest heavily to build its network while trying to pry enough users away from the first to make its network grow. But when duplicating a network is cheap, an innovator may be able, by charging a zero price for a superior service (as Facebook does), to eat the existing network’s lunch in short order. MySpace, started just a year before Facebook, quickly became the dominant online social network and remained in the top spot until 2007, when Facebook overtook it. (This was a rare victory for refinement: Kirkpatrick remarks upon MySpace’s “distinctive Times Square look—all flashing graphics and ribald images.” And “you didn’t need an invitation from an existing member [to join]. You could use either a real name or a pseudonym.”) Last year, MySpace laid off 30 percent of its employees, and its network is now less than a quarter the size of Facebook’s. But Twitter is surging, and Kirkpatrick reports that Facebook is worried that Twitter might overcome it. Remember the dictum: either overwhelm all other networks by snatching their users away from them or disappear. There are many Zuckerberg wannabes out there, trying to do to Facebook what he did to MySpace.
There is also concern that Facebook may be only a fad, like Beanie Babies. A soaring demand (think how fast online networks expand compared to highinvestment networks such as telephony) for a new good or service may create a bubble phenomenon—people buying because other people (presumed to know what they are doing) are buying. We know what happens to bubbles. And Facebook also has to worry about the aging of its founder and his co-workers. As a start-up company in a rapidly changing economic and technological environment matures, its capacity for radical innovation is bound to diminish. Software programming continues to advance with long strides; but Zuckerberg no longer programs. As the company expands, it will need managers with more experience. They will be older, more cautious, than the current ones. The rising generation will have different tastes and aims; fewer and fewer Facebook employees will belong to it. The best young talent will go to work for new companies.
Network competition is a function in part of the scope of patent protection. Software is patentable, but it is usually possible to invent around patented software—that is, to develop non-infringing software that accomplishes the same task. (Different software can produce the same word-processing capability, for example.) The major legal threat to new online social networks comes from a controversial category of patents called “business method” patents. They enable the patenting of a novel method of doing business provided that the method employs technology, such as computer software. The idea of an online social network is too obvious to be patentable (remember that 1968 article), but Facebook is an example of a business method that specifies a distinctive set of features—the set described earlier that explains much of its commercial success.
Yet neither Facebook nor any other online social network should be permitted to obtain patents and thus impede new entry into the online social network market. The rationale for patent protection is that inventors, or the firms that employ them, will not spend large amounts of money to develop a product that a competitor who incurred no development expense can cheaply copy and produce, and thus sell profitably at a lower price than the inventor. (Pharmaceutical drugs are the canonical example of such a product.) Zuckerberg and his friends spent virtually nothing to develop Facebook, as distinct from the costs incurred to operate and to expand the service once it was up and running. The lure of a patent is not necessary to induce abundant and energetic innovation in online social networking. Patents would have blocked Facebook. As it was, Facebook had to settle a patent-infringement suit for a non-trivial sum.
Privacy scandals involving online services, such as e-mail, Google searches, MySpace, and Facebook, are endemic to this universe. Most of them result from inadvertent disclosures by the provider of the service, but they can also arise from the tension between the users of such services and the advertisers. Yet Americans, if one may judge from their behavior (or “revealed preference,” in the language of economics), are not greatly worried about advertisers’ obtaining personal information about them. Otherwise they would not buy books from Amazon.com or use Gmail. They realize that advertisers are not interested in people’s personal information as such and do not want to embarrass their customers; they merely want to sell more stuff.
But online social networks are different. Information revealed to advertisers may, as in the incident involving the Beacon program, leak out to people whom the network user would prefer not see it. Remember that Facebook is entirely dependent on advertising revenues and has had to struggle to bring those revenues in line with its costs, and that Zuckerberg has a relaxed attitude toward privacy. One former Facebook programmer thinks that “Mark doesn’t believe in privacy that much, or at least ... sees privacy as something Facebook should offer people until they get over their need for it.... Members of Facebook’s radical transparency camp, Zuckerberg included, believe more visibility makes us better people ... more transparency should make for a more tolerant society in which people eventually accept that everybody sometimes does a bad or embarrassing thing.” It is not surprising, therefore, that Facebook has been repeatedly criticized and occasionally sued for carelessly—or maybe even deliberately—compromising its users’ privacy. When a user quits Facebook, for example, the company’s servers retain all the information that had been posted on the user’s Facebook page.
But the more interesting question is whether the reduction in privacy that Facebook users experience—by the very nature of a network that invites people to share personal information, far beyond what is normal or even feasible when friendship arises from personal contact—is a social cost. I have my doubts. I do not share Zuckerberg’s utopian expectations for the creation of community by means of Facebook, but I do think that we tend to exaggerate the value of privacy. We do this by failing to distinguish its private value from its social value. Privacy is a powerful weapon that we wield to advance our personal interests. We reveal to others the information about ourselves that advances our projects (marriage, promotion, disability benefits, purchases, and so forth) and try to conceal the information that would retard them. Like actors, we present a manufactured self to the world. (Erving Goffman brilliantly analyzed the performative dimension of social relationships.) And privacy can also be dangerous. In our new age of terrorism, we are more mindful of this than we once were. We wish Al Qaeda were “radically transparent” to us.
And yet it would be crazy to try to impose a general legal duty of candor—to require everyone to post online, accessible to all, every detail of his private life: the kind of information you would have to disclose in a “lifestyle” polygraph exam if you wanted to work for the CIA. We would be drowned in information, much of it confusing and distracting (arresting, but irrelevant). But that is not where Facebook is going; and despite all the snooping technology of modernity, we have more privacy than our ancestors had (urbanization is a big enabler of privacy)—if we want it. And those ancestors seem not to have suffered disabling psychological damage from their relative lack of privacy.
What is true is that the less privacy, the more conformism. But it is hard to argue that Facebook is making Americans more conformist, composed as it is of millions of separate, more or less private, social networks. Indeed, social networking may also promote a kind of individualism by enabling (with apologies to Wilde) non-conformists to find each other. Kirkpatrick notes that “Facebook has now become one of the first places dissatisfied people worldwide take their gripes, activism, and protests.” But Facebook may be making us more narcissistic. It may do so, as Kirkpatrick rightly observes, by encouraging exhibitionism. “On Facebook we follow the minutiae of our friends’ lives the same way millions follow Britney Spears in People magazine,” Kirkpatrick writes. “The Internet theorist David Weinberger posits that ‘on the Web, everybody is famous to 15 people.’”
There is also a legitimate concern that online social networks are an addictive distraction, even that they are altering our brains in a way that is destroying our ability to concentrate. But all that is really clear is that Facebook is taking up an enormous amount of people’s time. The company reports that its users spend 500 billion minutes online with Facebook every month, though Kirkpatrick reports a figure half as large. Neither figure includes the time spent by users in taking photos and posting them on their Facebook page. (Facebook hosts some 30 billion photos.) Facebook users were certainly doing something with those 250 or 500 billion minutes a month before Facebook came along, and maybe something of greater social value. But who knows? Maybe they were spending that time networking the slow pre-electronic way. Maybe they now have more time for other activities.
We may laugh at Socrates, in the Phaedrus, for denouncing literacy, which he said would create “forgetfulness in the learners’ souls, because they will not use their memories; they will trust to the external written characters and not remember of themselves.... They will appear to be omniscient and will generally know nothing; they will be tiresome company, having the show of wisdom without the reality.” But maybe his anxieties about the cultural consequences of communications technology were just premature. Still, I doubt we need worry too much about the effect of Facebook on the psychology or the cognition of its adult users. They each have their social network created mainly the old-fashioned way, and Facebook will help them maintain it. But what about the teenagers, enabled by Facebook to form immense social networks? They are said to be abandoning “best friends” in favor of having looser relations with more friends, a trend surely accelerated by Facebook—if you spend a lot of “face time” with just one or two of your “friends,” you will have no time for the other 398 or 399. A Facebook network is a social collective, a virtual kibbutz, and studies have found that children brought up in a traditional kibbutz have difficulty forming strong emotional relationships as adults.
The Pied Piper of Hamelin led one hundred children into a cave from which they never emerged. Some 500 million people, of whom about 10 percent are thirteen to eighteen years old and another 25 percent are eighteen to twenty-five years old, are now marching to the digital pipes of Mark Zuckerberg, who is twenty-six years old. I have no idea where they are marching, and whether they will ever return.
Richard A. Posner is a judge on the U.S. Court of Appeals for the Seventh Circuit and a senior lecturer at the University of Chicago Law School.