Crying rape falsely: rare or common?

The founder of Wikileaks, Julian Assange, is claiming that a dirty tricks campaign lay behind the charges of rape and sexual assault laid against him by two Swedish women last week.
The rape charge has been dismissed as groundless by the Swedish authorities, who are still investigating the charge of sexual assault, The Times reports today (p 25). Mr Assange met the two women ten days ago. He denies the charges, but fears that they have damaged him and his organisation, which recently published 70,000 confidential frontline reports on the Afghan war, causing fury in the Pentagon.
Sweden has the highest rate of reported rape in the world, 46.5 cases per 100,000 people, almost twice that of England & Wales. But it also has a low conviction rate: around 10 per cent of cases reported in Sweden end in convictions. In general, there is a roughly inverse relationship between the reported rape in individual countries and the proportion of these reports that end in conviction.
The difference in reported rapes has almost nothing to do with the actual level of these crimes, unless you believe that Swedish and British men are 10 to 20 times more likely to commit rape than are men in, say, Portugal (reported rate 3.2 per 100,000) or Hungary (2.2 per 100,000). They are related, rather, to the willingness of women to report rape.
Some of these reports are false, as in Mr Assange’s case, but searching the data to tease out how many is a thankless task. Official publications in this country claim levels of false reporting that range from a low of 2 per cent to as high as 10-12 per cent. Some of those involved in investigating rape and in defending those accused of it believe the level may be much higher.
The issue is live because, to the surprise of many, the coalition Government has put the issue of anonymity for those accused of rape back on the agenda. That is hard to justify without any evidence that men’s names are being blackened by false accusations. Before the recess, Justin Blunt MP, junior minister at the Department of Justice, faced a barrage of criticism in Parliament from Labour MPs arguing that there is no higher level of fraudulent claims in rape than in any other offence.  
In the debate on 8 July, he promised to publish an independent assessment of the current research and statistics on rape, commissioned from the director of analytical services at the department. It would be published in the final week of July, he promised, but it wasn’t. It is now scheduled to come out in the autumn.
What might it say? Mr Blunt’s remarks in Parliament suggested, curiously, that he did not believe the granting of anonymity to defendants had anything to do with the level of false accusations. That may be because he also accepts the long-held Home Office view that false accusations of rape are proportionately no higher than those of any other crime.
Others disagree. In a recent letter to The Times (19 July 2010), two barristers, David Wolchover and Anthony Heaton-Armstrong, say that they believe concoction is much commoner in rape trials than in other offences.
They sought, under the Freedom of Information Act, to get a breakdown of false allegations by offence type, in order to see if the official figures were justified. They were turned down on grounds of cost, and encouraged Baroness Stern, who was conducting an inquiry on behalf of the Government Equalities Office, to seek the same information. She, too, was rebuffed, they say.
Her report quotes police officers, Crown Prosecution lawyers and judges as saying that false accusations are very rare. But the two barristers say they listed a “huge number of established cases of concoction” in Criminal Law and Justice Weekly (April 24, 2010).
Official documents offer a range of figures. The Crown Prosecution Service’s Rape Manual, in a section called Societal Myths, states that “studies have indicated that only 2 per cent of all reported rapes are false, which is slightly less than false reporting in all other crimes”.
It gives no references to these studies, but the 2 per cent figure originates in the US and has been frequently cited. An attempt to trace it to its source by a US lawyer, Edward Greer, found that it originated in the feminist writer Susan Brownmiller’s 1976 book Against our Will, using data quoted by a judge that in turn came from the Commander of the New York City Rape Analysis Squad in the mid-1970s. There appears to be no published report to substantiate the claim, nor any evidence of how the statistics were collected.   
In fact, the US literature on the subject provides almost any estimate for false accusations of rape you care to choose. One study, published in 1994 by Eugene Kanin, found that in a small unnamed community in the Mid-West where every reported rape was carefully investigated, 41 per cent turned out to be false. The study covered 109 reported rapes over nine years and false claims were only classified as such when the complainant admitted fabrication.
More recent British studies come up with figures of 8-12 per cent. Liz Kelly and colleagues from London Metropolitan University in a 2005 report for the Home Office (A gap or a chasm: understanding attrition in reported rape cases)  found that of 2,643 cases in their data set, 216 were classified as false allegations (8 per cent). But as a proportion of the cases not proceeding beyond the police stage (1,817) this represented 12 per cent.
These were judgements made by the police, which the authors of the report were reluctant to accept. In addition, there were 318 cases where the victim withdrew the claim (17 per cent) and a similar number (315, also 17 per cent) where the victim declined to complete the initial process, which the report attributes largely to poor handling of the initial complaint by the police, or the fear the women had of court proceedings and of being judged.
Such fears are entirely understandable, but it is also possible that some of these withdrawals represented false allegations which the complainants were reluctant to acknowledge. Of all the rapes reported, 12 per cent failed to make any progress because the complainant declined to make a formal complaint, refused to have a forensic examination, failed to give a statement or withheld information.  
 Pity the poor civil servant who has been asked by Mr Blunt to produce “an independent assessment of the current research and statistics on defendant anonymity in rape cases”. The statistics are so open to interpretation that what you believe they show depends very much on the preconceptions you start out with.