Cheap Pseudonyms, Privacy, and Sex Offenders

The BBC is reporting on a move by the British government to require convicted sex offenders to register their online identities. Of course, it only takes a minute of thought before you realize that its so easy to get a new email address that registering one doesn't do much good.

There are some scary responses to that, like this one:

If everyone had a single internet identity for life, like a National Insurance number, this would make it far easier to track people, he said. Child internet safety expert John Carr, of children's charity NCH, said: "This is a very welcome move."

"It will mean that we can extend the Sex Offenders Register regime into cyberspace and that will be a great comfort to many people."
From BBC NEWS | Politics | Plan to list paedophile web names
Referenced Tue Feb 06 2007 12:00:20 GMT-0700 (MST)

This kind of legislation isn't limited to Britain. In December, John McCain introduced a bill in the US Senate that would do much the same thing.

The issue here is the pseudonyms are cheap. Eric Friedman and Paul Resnick wrote a paper in 1999 on the social cost of cheap pseudonyms (PDF). Here's the abstract:

On the Internet it is easy for someone to obtain a new identity. This introduces opportunities to misbehave without paying reputational consequences. A large degree of cooperation can still emerge, through a convention in which newcomers ``pay their dues'' by accepting poor treatment from players who have established positive reputations. One might hope for an open society where newcomers are treated well, but there is an inherent social cost in making the spread of reputations optional. We prove that no equilibrium can sustain significantly more cooperation than the dues-paying equilibrium in a repeated random matching game in which players have finite lives and the ability to change their identities, and there is a small but nonvanishing probability of mistakes and a large number of players.

Although one could remove this inefficiency by disallowing anonymity, this is not practical or desirable in a wide variety of transactions. We discuss the use of entry fees, which permit newcomers to be trusted but exclude some players with low payoffs, thus introducing a different inefficiency. We also discuss the use of unchangeable pseudonyms, and describe a mechanism which implements them using standard encryption techniques.

The paper presents a game-theoretic study of various strategies for dealing with cheap pseudonyms. The bottom line is that there aren't many good ones:

It would be nice to create environments where strangers were trusted until proven otherwise. Unfortunately, obvious strategy vectors involving cooperation with strangers are not stable, and we proved that no strategy vector can do substantially better than punishing all newcomers.

Thus, there is an inherent social cost to free name changes. We can mitigate this cost by charging for name changes, but this also requires charging for names in the first place. That may exclude poor people or those who are just exploring and not yet sure whether the payoffs from participation would justify the entry fee. A better solution is to give people the option of committing not to change identifiers.

The paper presents such a mechanism. Would that help here? Probably not. Say, you give someone a "once-in-a-lifetime" identifier. That identifier works within a certain context, so you could commit someone to not change their MySpace identifier, but that doesn't mean they wouldn't have another identifier somewhere else.

But so what? What makes registers work is that they can be easily checked. Send the parole officer by and see of the guy really lives where he said he did. That's hard to do online and once-in-a-lifetime IDs don't help. Confirming that someone is somewhere is a provable thing. Confirming that someone isn't somewhere (or doing something) is nearly impossible.

Legislation to give someone a single identifier for life would likely fail for the same reason. In the end, however, society will decide how much they value the anonymity of the Web when confronted with the attendant costs.

Bonus link: Terrell Russell wrote about this back in December.