Last summer, Facebook founder Mark Zuckerberg and several other tech luminaries announced the creation of Internet.org, an initiative that aims to deliver the Internet to billions of people in the developing world. It would increase access to opportunity, its founders argued—and, implicitly, a fairer future.
But the simple link between social connectivity and justice could be a flawed one, according to a recent analysis from the Massachusetts Institute of Technology. Where Internet access is often touted as a human right, the structure of the Internet itself may actually lend itself to perpetuating inequality.
In January, three MIT data scientists published a paper in Nature Scientific Reports that aimed to determine the relative levels of meritocracy in societies, based on the structure of their social networks. To do this, the researchers sorted the social media world into two types of people: Content creators (a.k.a. "rockstars") and content curators, or middlemen.
The authors were interested in finding the moment where a meritocratic society, one which pays people primarily for the work they produce, shifts into a topocratic society, one in which middlemen are paid more than the creators, simply because of their position in social networks.
"If everyone has more [connections], the system is going to be meritocratic—the guys who contribute more are going to get paid more. With less connections, what’s going to matter most is how connected they are," says César Hidalgo, an assistant professor at the MIT Media Lab. For example, he says, if there is a network of 1 million people, if each person is connected to 1,000 others, you've reached the meritocracy threshold.
The MIT analysis didn’t look at the Internet connectivity specifically. Still, at first, the paper seems like an argument for increasing Internet access. Bring on the connectivity to push us into a new, radical era of fairness augmented by technology. But if we’re applying the analysis to the Internet, Hidalgo also raises an important caveat to consider.
Says Hidalgo: "When you have relatively small systems in which people are able to know each other well—whether it’s a team of people working together in a professional or academic setting—those systems are going to be much more meritocratic, because if everybody knows each other, it implies that there’s a correct way of assigning payoff to each one of the people’s contributions. But when that system becomes large, eventually that information becomes dominated by hubs, and the hubs can get a highly disproportionate amount of the total share of income, based only on their position."
When you extend this model to the Internet, it becomes discouraging. The vast and multitudinous Internet is largely dominated by four hubs: Google, Facebook, Twitter, and Youtube. While none of those sites produce content, all profit well from sharing or curating it. The model they create is one in which the well-connected middleman stands more to gain than content producers.
It’s not all bad. In one sense, megahubs like these are extremely useful and efficient. I found out about the MIT study through Twitter, because Hidalgo and I follow one another, and Twitter creates an immediate proximity between an academic in Boston and a journalist in New York during a snowstorm. But it also means that, to some extent, I’m blinded by well-established connections. Had searching through Google not originally led me to a study by Hidalgo, a professor at a well-connected university, I might never have discovered his work at all.
That said, small meritocracies can persist, even in a larger network dominated by megahubs. Hidalgo cites Etsy or Airbnb as examples of smaller, perhaps more meritocratic networks. In smaller networks, it’s less a popularity contest than it is a peer-to-peer quality contest.
"If you think about something like Airbnb, in that case the institution that Airbnb pushes is one in which you’re developing trust," Hidalgo says. "The person is renting own room, their own house, so the ones that do better might be because it is actually a better unit."
And none of this is to say that technology can’t create more just societies. Hidalgo cites Bitcoin as an example of a decentralized, peer-to-peer currency that could put earning power back in the hands of creators, rather than middlemen like Mastercard, Visa, or Paypal.
But, either way, the MIT analysis may force us to reckon with the fact that while we’d like to believe we live an increasingly meritocratic world, we still drag all of our old, deeply unequal social structures into our new economic inventions and technologies. Income inequality may no longer be dictated by royal blood, or ethnic caste, but Adam Smith’s "invisible hand" doesn’t operate in a fairly connected society, either.
"There’s a difference between inequality and fairness. Inequality can be fair if it reflects a natural difference in people’s talents and contributions to society," Hidalgo says. "The idea that inequality is justified because the market is meritocratic may no longer be valid. The inequality that we have is not justified."