How the government uses dirty data to legislate morality

So what's a standard deviation?

Poppycock problem

Is a trend remotely charitable? Possibly – but only by professionals, who are aware of the underlying factors that affect the figures. The debate is not helped by women’s groups who report far higher incidences of unreported "rape" based on self-identification by "victims" and compound this with assertions that only a very small proportion of victims lie about such matters.

The difficulty here is definitional: It is very likely that many if not most of the women making such claims are correctly asserting that they have suffered a traumatic experience. But, as a crime, whether rape takes place or not depends on the perception of the perpetrator - whether or not he had reasonable cause to believe that consent was present.

Whilst the conviction rate for cases that get to court is considerably higher than 5 per cent (hovering around 50 per cent), it is a long way short of the 97 per cent implied by the claim that women rarely lie about such matters. Are 50 per cent of victims therefore lying? Absolutely not.

Courts are, however, acknowledging a difference in perspective between alleged victim and perpetrator that is almost certainly not resolvable in any straightforward fashion.

This issue of self-reporting is also at the heart of numerous studies that purport to demonstrate attitudes on sexual issues.

One study asked respondents whether violence towards a partner "could ever be" justified – and then used the results from this loaded question to proclaim that a high proportion of young men believed it was ok to hit their girlfriend. The best measure of behaviour is not what people say, but what they do – their "revealed preference" as economists term it – and basing policy on how individuals respond to surveys is particularly dangerous.

A furious debate on the measure used is also at the heart of questions raised in respect of proposed legislation on Trafficking. Provisions introduced in the Police Bill 2009 would make it a criminal offence to purchase sex from any individual who had been trafficked.

The rationale for this law is "research" put together by the Poppy project and other anti-prostitution groups that claims that the number of trafficked women in the UK runs into the tens of thousands. One report put the figure as high as 80,000.

On being questioned with regard to the academic credentials of their research, Poppy rather huffily suggested that this was an attempt to silence them, and they had as much right to publish research as anyone else.

Against that, the numbers of trafficked women that the police have actually managed to find are vanishingly small. A six-month investigation involving over 800 brothels turned up just 167 possible victims

Sponsored: IBM FlashSystem V9000 product guide

Next page: Extreme porn