Contents

  1. The art of designing markets
  2. Social media are marketplaces
  3. Failures of social media
  4. Designing better social media

Alvin E. Roth is an economics professor at Stanford University, and shared the 2012 Nobel Memorial Prize in Economics. He focuses on designing markets, which he regularly blogs about over here. I recently read his brilliant book Who Gets What - And Why? (2015). It improved my understanding of the mechanisms at play in both successful and failed marketplaces. Since then, I’ve been pondering the topical question who gets what in social media - and why?

The art of designing markets

In the October 2007 issue of Harvard Business Review, Roth published this eponymous article, where he states that “A primary motive for market design is the need to address market failures”. He goes on to say that markets need to do at least the following three things for proper functioning.

  1. They need to provide thickness—that is, to bring together a large enough proportion of potential buyers and sellers to produce satisfactory outcomes for both sides of a transaction.

  2. They need to make it safe for those who have been brought together to reveal or act on confidential information they may hold. When a good market outcome depends on such disclosure, as it often does, the market must offer participants incentives to reveal some of what they know.

  3. They need to overcome the congestion that thickness can bring, by giving market participants enough time—or the means to conduct transactions fast enough—to make satisfactory choices when faced with a variety of alternatives.

In addition, Roth makes a clear distinction between typical commodity markets (such as share markets) and matching markets (such as clearinghouses for newly graduated doctors seeking residencies at hospitals). His focus is on matching markets, where it isn’t enough for you to choose alone, you also have to be chosen in return. These markets comprise transactions based on two-sided preference matching, not simply price alone. His major research contributions have been in such matching markets that he helped design. For example, he worked with surgeons in America to setup kidney exchanges that saved thousands of lives, and continue to save lives today.

One other thing Roth has popularised is that repugnance is a constraint in such matching markets. Consider the example of a kidney exchange based on voluntary participation, as it is illegal to buy or sell human organs in nearly every country. Hence any kind of organ exchange is strictly treated as a mutual gift. However, putting a price on human organs is not something that would automatically harm those who are not involved in the actual transaction, although there are plenty of future possibilities for harm that Roth discusses in detail. So repugnance is attached to a transaction that some people would like to engage in, but which is forbidden by everyone else even if it harms no one (at least not immediately).

Roth doesn’t take sides on the issue of whether repugnant transactions ought to be allowed or not. Instead, he does an excellent job of articulating and explaining the choices that a society has, in the hope that better decisions will be made. I’m finding it very instructive to use his framing of marketplaces and their designs for better understanding a number of current phenomena. The privacy violations by Facebook during 2018 is one such phenomenon that has been on my mind lately. But before focusing on the failures of Facebook (and social media by proxy), the pertinent question is whether or not social media are marketplaces?

Social media are marketplaces

I definitely think that social media are marketplaces, because they bring together users and advertisers. The buzzword platform is now ubiquitous for describing social media, but I think it’s imperative to use the more accurate (and honest) term marketplace. A platform like Facebook does nothing more than mach users to advertisers - it’s not an oversimplification; in fact, it’s the exact truth. So marketplace is the word that I’ll use from now on.

The profitability of Facebook indicates that it has succeeded in doing the three things that Roth pointed out (which I quoted above). There’s definitely thickness, given the large number of users and advertisers. Facebook has convinced users to honestly divulge a huge amount of confidential and sensitive information about themselves, so much so that highly targeted adverts can be displayed to each user. The speed and scale of Facebook’s infrastructure have clearly removed almost all congestion from the marketplace, at least as far as advertisers are concerned. I’ll include LinkedIn and Twitter as same sort of marketplaces, even though they focus on different aspects of each user.

Failures of social media

Now, here’s an interesting question: if social media marketplaces are so highly profitable, they must surely be functioning beneficially for all involved? Not necessarily. I think companies like Facebook have coerced users into honestly divulging their private information in exchange for free cloud storage and free communication services. Most users have no idea that it’s the advertisers in these marketplaces who get to choose exactly what they want. So it superficially appears as though social media marketplaces are safe for both users and advertisers, in terms of the definition provided by Roth earlier. Shockingly, we learned during 2018 that social media marketplaces are evidently unsafe for users, because of the enormous violations of privacy and trust perpetrated by Facebook and co.

Unsafe marketplaces can lead to a counter-intuitive form of congestion that Roth discusses in his book and papers. In the context of social media, I think this manifests as the humongous proliferation of lies and false information, which in turn leads to a distortion of how users interact with each other, and with advertisers. If there’s zero incentive for truth and trust in a marketplace, then participants have to carefully strategise about their choices, presuming that they are capable of being rational. However, cognitive biases in humans mean that we end up making systematic errors. For instance, repetition of a truth can induce people to become desensitised and accept it as a lie instead (see Daniel Gilbert and colleagues’ 2018 paper about “prevalence-induced concept change”). Subsequently, if it becomes a norm that every piece of fact is taken to be a lie by default, any truth will be so much harder to acknowledge; thereby leading to uniform polarisation and distrust in social media.

Designing better social media

Admittedly, I’m speculating above regarding what caused social media to become failed marketplaces. However, the privacy violations are undisputed, and must be corrected. But I think the current spate of suggestions about governments regulating Facebook will end up making a bad situation much worse. The solution that seems most promising to me is the art of designing markets. Applying the framework that Roth has developed might just give us the ideal outcome, where both users and advertisers get what they want. I hope that the Brain Trusts at Facebook, LinkedIn, Twitter, etc. are reading his book.