You have no property rights over your Twitter data
In his latest CoinDesk column, my friend Nic Carter wrote about the fallout of the recent Twitter hack and his view that “social media handles, as well as user-contributed content, ought to be understood as property.” I told him I was thinking of writing about why I think that’s wrong and he welcomed a response, so here goes.
Let me stipulate at the outset where Nic and I agree, which is a lot. The Twitter hack highlights the fact that centralized or intermediated identity is bankrupt. What we should want are decentralized systems that allow individuals to have sovereign control over their own identity. I could write a long paper explaining all the ways such systems would be superior to what we have now, but luckily I don’t have to since Peter Van Valkenburgh already has. So, I agree with Nic on the most important thing, which is where we want to get to: a world where you own your own identity credentials (whether that’s a social media handle or your ride-hailing profile or whatever else) and you can take your data with you wherever you want, to whatever service or system will read it.
What I disagree with is Nic’s view (which he outlines in his recent column but more fully articulates in another column from June 9) that
…Facebook, Twitter, et al, did not really create all the content on their platforms, nor do they really own it. Instead, they define a namespace that users occupy, build upon, and in some cases commercialize. The users, not the administrators, create the vast majority of the value, and as such are the rightful owners of their digital property.
Such “digital property,” I gather, includes one’s handle and social graph and users are their “real“ or “rightful” owners. This is incorrect. One doesn’t have a reasonable claim to any such “digital property.”
In support of his argument, Nic cites writings by Balaji Srinivasan, Allen Farrington, and Elaine Ou. On closer inspection, though, we can see that in the linked pieces neither Srinivasan nor Farrington argue that one has a rightful claim to one’s social media identity because one created a lot of value on a social media platform. Instead, their arguments are basically what I stipulated above, that we should build systems that allow for self-sovereign control of one’s identity, like Urbit. Ou, on the other hand, does indeed make a case for an individual to be able to make a claim (against the wishes of a social media company) for the continued use of a service. She cites the property law concepts of “squatter’s rights” and easements, as well as John Locke’s labor theory of property, and these are the arguments that Nic marshals in his opeds. I will address these in turn.
Let’s start with Locke. Putting aside that Locke’s theory is not universally accepted, it’s not applicable here on its own terms. The point of the theory is to morally justify the concept of property by, in turn, justifying initial acquisition through labor. That is, in a “state of nature” in which all resources are unowned, Locke argued, it would be just to acquire property and exclude others from it by virtue of one’s mixing of one’s labor with it. It’s a theory meant to justify legitimacy of property as a concept (as part of the natural law) by making the case that initial acquisition in a state of nature (that never really existed) can be seen to be just.
Arguing for the moral legitimacy of property rights is not the same thing as saying that if, with or without your permission, I put on a comedy show every Saturday at 1 at your theatre, that now that slot rightly belongs to me as my property. Locke’s theory might matter to a discussion of whether digital property is justified as a concept, but it is just plain inapplicable to the question of whether you “own” your Twitter handle because you’ve been tweeting on the platform for ten years and Twitter hasn’t kicked you off.
I’ll pause here and say that I think the root of the confusion at issue is thinking about Twitter use in terms of property rather than contract. Twitter is a company that owns certain physical and intellectual property and it lets people use these subject to a contract that users voluntarily agree to when they sign up for the service. Now, you may (as Nic seems to) believe that the contract is not enforceable because it is poorly specified and capriciously enforced, but if that’s the case then you have to make a contract law argument about why the contract is void, not a property law argument, and that brings me to the “squatter’s rights” and easement arguments.
“Squatter’s rights” is what’s known in property law as adverse possession. As the name implies, the concept applies when the claimant has taken possession of one’s property without one’s permission, usually because one isn’t aware of it. If this goes on for a long enough time, when you go to eject the squatter, a court might find that you have forfeited the property to the squatter, and there’s a long and interesting history about how and why this rule emerged at common law. But here’s the thing: one of the required elements of adverse possession is that “[t]he occupation must be hostile and adverse to the interests of the true owner.” If you have permission to use the property, say by the terms of a contract, then you can’t bring a claim of adverse possession. Again, the concept is just plain inapplicable to the question of whether you “own” your Twitter handle because you’ve long used it to create value. Not only does Twitter know you’re using it, they are giving you permission to use it under the terms of a contract that you agreed to.
It’s a similar story with easements, but let me first point out that even if one accepted that an easement could be enforced against Twitter, what that would get you is not a property right over your Twitter handle and social graph, but a right to continue to use the Twitter service as it exists, which is kinda what we’re trying to avoid in the first place. But again, easements are a property law concept and they’re inapplicable in what is a contract law context. Courts find implied easements precisely when there is no contract between the parties enumerating their rights and obligations, and again use of the property by the claimant must be “hostile”—i.e. without permission, which is definitely not the case with Twitter.
So, it’s just plain wrong to say that one has a property claim over one’s social media handle or social graph or the like. There is no “digital property” at issue here. Instead what we have here is a contract between Twitter and the user, and if you want to argue that the user should have rights that are not contained in the written contract, then you need make contract law arguments like unconscionability, but even then all you may get is the right to get out of the contract.
That all said, it may still be good public policy to force social media companies like Twitter and Facebook to let users download their social graphs and other data. As Farrington says in the article Nic cited, “We could and probably should mandate portability of personal data under the auspices of providing fairness to consumers and encouraging competition among web services.” This is something for which my friends at Union Square Ventures have long been advocating: the law should let users take their data (tweets, followers, likes, everything) and port it to a competing service, including a decentralized one. But while that may be good public policy, such a “right” would be created by and justified by democratic legislation; it could not be justified by appeals to Lockean natural law or the emergent common law of property or contract—indeed it would be an abrogation of these.
 I’m an avid follower of Elaine Ou’s blog and Bloomberg column and have come to the conclusion that her work is meant for a Straussian reading. For this post, though, I’ll take her piece at face value.