Washed-up principles on DNA

Adam Rutherford writes on Guardian Comment is Free about the DNA database:

The value of expanding the database is to provide “cold hits”, where unanticipated links between crime scenes and individuals can be made. By retaining the DNA on a database for any length of time, the chances of unanticipated matches increase, because the database serves as a suspect list for crimes not yet committed.

Except it doesn’t appear to work. Analysis in 2006 by GeneWatch UK, which monitors the application of genetic technology for public interest and human rights, showed that the rate of convictions using the DNA database actually fell after the introduction of retention of records at arrest (rather than charge) in 2004. Labour has stated that 23 previously arrested, but not convicted, people were convicted for rape, murder or manslaughter last year based on database entries. The election fact-checker service by the Times gives this a pork-pie rating of four out of five. The data is complex and insufficient.

Labour wants a permanent database. That was ruled an infringement of human rights. So they went for six years, the Tories initially three, and now six. What a farce! The difference between six years, three years and in perpetuity is absurd: either the state keeps data on innocent people or it does not. One position inches towards a police state, one supports the liberty that is in our DNA.

In the linked “Election fact check” at the Times, Tom Whipple points out:

Labour use the [23] figure to argue that Conservative plans to limit the police’s ability to hold DNA samples from those acquitted of minor crimes would hit convictions for considerably more serious ones. Faced with the prospect of defending murderers, the Conservatives have now partially capitulated.

Assume, though, the data are correct. Labour also say that 56 others were similarly linked to rapes, murders or manslaughters, again through the database. 79 rapes and murders is a high number. But so is 1 million – the number of people in the national DNA database who haven’t been convicted of a crime. In fact, if you took a random sample of a million adult Britons last year you would have expected them to have committed 275 serious crimes.

There are three possible explanations for this discrepancy: 1. people on the DNA database are more law-abiding than the general population; 2. DNA isn’t as good at getting convictions as we thought; or, 3. the statistics are nonsense.