Before police condemn an online pursuit of a rapist, they should put their own house in order

Vigilantism's problem is that it's usually quite stupid. You talk about a mob response, and you immediately think of the howling thickos who mistook a paediatrician for a paedophile in south Wales in 2000. Six years later, when this event had been cited many times by commentators concerned about tabloid influence - even "government by tabloid" - Brendan O'Neill, editor of Spiked Online, did a definitive breakdown for the BBC of what actually happened: there was no baying mob, no threat of violence. A doctor came home to find someone ("probably in the 12-17 bracket", police noted) had vandalised her home with the word "paedo". Who's to say that the vandal had confused the two terms? Perhaps he or she was pointing out how amusing it was that paediatrician could be shortened to "paedo". O'Neill makes the more comical point that broadsheet fear of the tabloid mob is usually as unreflective and unfounded as whatever the mob itself is afraid of.

And yet, certain unarguables remain: one's first response to an event is rarely one's most mature or reasonable, and yet this is the response that governs any spontaneous outpouring. So, when the Baby P story was first broadcast, immediately after the Facebook pages urging you to Pray for Baby P came the Facebook page unmasking his killers (it also appeared on Bebo). It was removed, but not before it had served as a good example of the worst kind of social networking guard-doggery. These people were already in prison so there was nothing to be done with the information, beyond co-ordinated, illegal reprisals. It didn't bring anyone back to life, or improve the judicial system; it just jerked a bit more fruitless energy into that savage event.

Naturally, all arguments against people taking the law into their own hands rely on the law being effectively prosecuted in police hands. Even if the court of public opinion scorns the regular court - for inadequate sentencing or parole too readily given - then at least those decisions were arrived at systematically, and can be defended. But nothing complicates the case against vigilantism as much as the current Facebook rape appeal. A rape victim's partner has posted some CCTV pictures of the alleged attacker on the site, along with the name of his regular pub, asking the public to help identify the man. Police are worried that this will make it harder to get a conviction. The poster responded that "if he's not caught, there won't be a court case at all".

It would be a compelling argument anyway, given that an eight-month inquiry has yielded no suspect. But here the police really are beginning to pay for their own inadequacy. The recent cases of John Worboys and Kirk Reid, rapists investigated so lackadaisically that scores of women were needlessly attacked, have changed the temperature. The reputations of the police and the judiciary were already execrable, but the 6% conviction rate at least had room for a bit of ambiguity - perhaps it wasn't always, or often, the fault of the police; perhaps juries were misogynistic.

But since Reid and Worboys there have been rumours that rape units are deliberately underfunded, that traffic offences are taken more seriously than sexual assaults, that conviction rates are being kept low. With nobody taking final responsibility for any of these decisions - and with the debate prematurely cauterised as the police face the more pressing priority of overreacting to climate-change protestors - I am in no way surprised that a regular person, neither baying nor howling, might now accord more faith to a social networking site than they do to their constabulary. Sanctimonious warnings from the police that this might damage the chances of conviction will not stand; they need to look to their own standards if they want to keep hold of law enforcement.

mszoewilliams@yahoo.co.uk