I’m at the Brookings Institution today for an event which will, in part, be the launch of my report for The Hamilton Project on “Enhancing Competition with Data and Identity Portability.” You can download the report here and here is the tl;dr summary. The Hamilton Project have also released a review of the state of competition in the US. There is a livestream of today’s event here.
The basic idea is simple. Facebook and other platforms are insulated from competition because they have successfully appropriated network effects. There is nothing necessarily wrong with that from an anti-trust perspective — that is, it is a way a monopoly can arise through ordinary activities and investments — but from a competition policy perspective it means that it is very difficult for competitors to come in and compete with Facebook for those users who may be dissatisfied with its services.
I note that the same issue arose in telecommunications competition. It was plainly obvious to all that if you wanted to deregulate telecommunications you need to allow calls to be made and received across networks — that is, you needed interconnection. I point out that, while, for social media, we are not talking about deregulation as it was not regulated to prevent competition, the same principle applies. You want people to be able to communicate across networks.
The insight is to understand that social media is just messages — they are richer than voice in some circumstances, have structure (in terms of likes and comments) and they can often be broadcast rather than one-to-one. The critical point is that it should not matter which network you are on — you should be able to receive and send messages to people and others you have given permission to.
The solution to this is to allow individuals to port their identity to other networks, With that identity comes a set of permissions associated with that identity and it is that which creates value here. A competitor to Facebook — call it, for the sake of argument, NoRussiaBook — could come in, offer a different ad policy and you could move there. Your friends and connections need not know the different although I would not recommend that be hidden — just that defaults be set for ubiquity which can then be seen and adjusted at the discretion of individual users. To be sure, privacy is complicated here but that is because privacy is complicated, not because there is anything really special or risky about this proposal.
Zingales and Rolnik were thinking along similar lines last year with a notion called social graph portability designed to address the same problem. I think identity portability is more straightforward and potentially a natural evolution of forces already taking place on the internet.
The report details some initiatives taken both by governments (e.g., Estonia and India) and private organizations (e.g., Microsoft) designed to provide individuals with a digital identity. And just yesterday the Australian government proposed a far-reaching approach.
There are technical challenges to this proposal but I believe that, if we are serious about encouraging competition in the digital space, something like identity portability will be required. There are several start-ups moving to solve this type of problem using blockchain technologies and I have to point out that Facebook has solved this internally for 2 billion people so if they were proactive they could roll this out ahead of any regulation.
All of the other solutions — namely, breaking up Facebook — do nothing to resolve the underlying issue — network effects become barriers to entry. We need to start taking that seriously if we want to do something here.