A little while back a New York Times article discusses the consequences for┬ácollege admission of saying undiplomatic things in social media. Apparently colleges monitor, or at least check up on, the social media presence of their potential applicants to see whether they’re the right kind of person for the school. Inevitably, students scrub, curate, or simply hide their account in response.

Leaving aside the possible rights and wrongs of this behaviour, we might ask: how does the college identify a social media account as belonging to one of their potential students? The general answer is that social media like Facebook and Google plus have a ‘real names policy’. And the answer to why they have that is, allegedly, that people behave more civilly towards one another when they are not hidden behind an anonymising username. One may doubt that is the only reason, given the value of the personal information thereby acquired. Nevertheless it seems to widely believed that, despite some awkward evidence to the contrary, that this works.

How does it work, if it does?

The logic seems to be that student realises that behaving badly would have reputational consequences, so decides not to. In equilibrium everybody behaves nicely in case the ruin their chances of, say, getting into college. Every now and again a misjudgment happens and the college refuses admission to remind everyone of what the consequences of off-equilibrium play are.

A problem with this particular game is that the internet remembers everything, more or less, so it is the college or future employer or insurance company that must discount past behaviour. It may not. In which case the student is permanently stuck with the inferential consequences. Criminal records work this way in many countries, often even preventing the convicted person from leaving the country to restart their reputation.

The problem is that a global, publicly viewed, personal reputation is usually too big a thing to wield against an individual in these circumstances. Vetting for security clearances might be the only case where the complete story is required.

Fortunately all the logic of reputation costs goes through even when the global, publicly viewed, reputation is not actually yours, but the reputation of a more limited ‘person’ constructed out of just one or two facets of your actual interests and a limited range of your actual behaviour. That is, exactly the sort of enduring but partial identity embodied by a username.

How weird is it to construct and maintain such a partial identity? Not weird at all. We do it offline all the time, every time when we do not present ourselves identically to parents, spouse, children, colleagues, insurance broker, or bank manager. It would, indeed, be weird not to differentiate these roles. Failing to do so in non-family contexts is ‘unprofessional’, and in more personal contexts is just plain rude.

Why might an institution, e.g. a social media platform, want to the whole you rather than just the aspects of your persona relevant them? For two reasons. First, so that you behave better in their domain, and second, so their marketers have the largest chance of selling you things. Doubtless different organisations value these differently, but all value both.

So there are three possible ways to treat an individual user. Allowing them to have complete anonymity, allowing them to maintain a finite number of persistent identities, and enforcing one online identity everywhere.

The behavioural arguments against complete anonymity are probably as good as the real names policy advocates say they are. But against a finite number of persistent identities, they fail. If reputational cost arguments work for one persistent identity, then they work for several.

The arguments against single online identity are also strong. When everything is recorded for ever, institutions are cautious and imperfectly informed, and when there are a new users arriving all the time, then there is no motivation to discount the past. In a world without a concept of bankruptcy, people take fewer risks and fewer goods are generated. A finite number of persistent identities solves this problem by allowing people to start again with a new ‘name’ and a clean slate. This does not subvert the reputational mechanism that encourages better behaviour because the reputation is reset to zero at the same time.

Multiple persistent online identities are therefore the perfect solution. Indeed, they were always the perfect solution. They are built into our social lives and into the earliest online computer systems too. Our task is to get them back.