The Crown Prosecution Service are holding a consultation on the prosecution of false allegations of rape. My own response is copied below - Laura Woodhouse at The F-Word has more background information and Women Against Rape have a response to which they are inviting additional signatures.
The deadline for consultation submissions is tomorrow (6 May) so there's not a lot of time if this is the first you heard about the consultation. (Government consultations such as this are not well publicised, and this isn't the first time I've missed or nearly missed something I wanted to comment on)
Here's my response. Most of it covers the 'public interest' criteria for prosecution. Quick summary if you don't want to read the lot - I believe it's basically never in the broader public interest to prosecute, even if it appear to be in the narrow public interest when only looking at a particular case.
I had [stronger trigger warning] this Shakesville post about a recent US case in mind for a lot of what I wrote, and one of the references is taken from comments there.
The description of perverting the course of justice is clear.
In paragraph 15 - for reasons I discuss more fully in my answer to the following question, I believe that for there to be evidence of a false allegation there needs to be far more than simply a lack of evidence of a rape.
To have victims of rape who are considering reporting second-guessing themselves as to whether the evidence they have is strong enough to prevent a charge being placed against them is fundamentally unjust, and as paragraph 13, and the recent Home Office study on attrition  point out, there are many reasons that a victim might later decide that they do not wish a prosecution to proceed.
Since a retracted allegation is not therefore sufficient evidence in this case of the initial allegation being false, I believe that prosecutors would need to look for affirmative evidence that no rape occurred, rather than for a lack of sufficient evidence that a rape occurred - the latter, of course, being all too common for true reports of rape for it to be used as evidence of falsity.
Regarding the public interest factors set out in the consultation document, I believe that there is a major factor missing for consideration, under which a prosecution would only rarely be within the public interest.
Rape and sexual offences, as is known from the British Crime Survey and many other sources, are rarely reported to the police in the first place. Distrust of the police and/or courts is cited as a common reason for not reporting.
Kelly's 2001 literature review for the CPS  refers to research by Jordan into recorded false allegations, which found that:
[...] within the file analysis were three cases that were designated false reports which subsequently turned out be assaults by serial offenders. Additional analysis of one serial rapist case showed that an early report by a young woman who named her attacker had been discounted as a false report; the man was convicted of 24 rapes eight years later.
Prosecuting a suspect who has been deemed to have made a false allegation is therefore extremely risky. Should the allegation be true, and have been retracted under duress, or - as has led to prosecutions recently - been deemed false by the police or CPS without a retraction, then to prosecute that suspect firstly revictimises them again, and secondly leaves a rapist free to rape again.
Furthermore, the idea that a victim can report rape - as victims are encouraged to do by the police - and then end up facing trial herself, is utterly abhorrent, and serves as a severe deterrent to reporting. There are very few other crimes - including those where false reports are more common - where people would find themselves scared away from reporting the crime for this reason.
The public interest test therefore needs to be considered more generally. The ultimate public interest of prosecuting crimes is that future crime is deterred and reduced. Rape is a far more common crime than perverting the course of justice through false allegations of rape. It is also - despite the maximum sentences being identical - a more serious crime.
It might be possible, taking a single case in isolation, to conclude that it is in the public interest to prosecute someone for making a false allegation. However, because of the particular contexts in which rape occurs and in which accusations of rape are investigated and prosecuted, it is not sufficient to consider that case in isolation.
If by prosecuting a case, one successfully punishes the maker of a false report - but by doing so, deters even one other person from reporting their own rape, and so leaves a serial rapist free to commit several further crimes - Weinrott and Saylor  found that convicted rapists on average had committed ten times as many rapes as they were eventually prosecuted for, as well as numerous lesser violent crimes.
It would therefore be very hard indeed to argue that the prosecution was in the interests of the public as a whole.
That prosecutions relating to false allegations of rape (but not false allegations of other crimes) are regularly reported in the press means that the likelihood of such a prosecution having a deterrent effect are extremely high.
Hundreds of thousands of rapes occur in the UK each year. Mere hundreds of false allegations are made, and those in which a suspect is named by the accuser are much rarer - arrest or charge of a suspect based on a false allegation is rarer still. Even a tiny deterrent effect could therefore let a significant number of rapists go undetected.
In addition to this, of course, the high-profile reporting that false allegations receive feeds in to the common belief among many members of the public that false allegations are commonplace. This attitude and culture is something that the CPS already have to work against. By prosecuting a false allegation case, prosecutors may be directly increasing the difficulty of the - already difficult - task their fellow prosecutors have in prosecuting rape cases. Again, this is difficult to justify as being in the broader public interest.
A consideration of whether prosecuting the case could increase the likelihood of other rapes being committed, other rapes remaining undetected, and rapists being difficult to prosecute successfuly - causing great harm to the general public interest - should therefore be weighed against the local public interest served by prosecution of an individual case.
The explanation of "double retraction" is clear. However, under the circumstances described in paragraph 26, I am unable to think of a situation in which a prosecution would be in the public interest, for the reasons of the matter of logic stated within that paragraph.
An example of a case where prosecution might be in the public interest is perhaps needed here.
Given the likelihood of a prosecution having wider repercussions beyond the narrow public interest of prosecuting a particular case, and the difficulties therefore caused in achieving justice for future victims of rape, I believe that the interim measures described in paragraphs 29 and 30 (referral to the Director of Public Prosecutions) should be made permanent.
These cases are extremely rare, but the need for oversight at the highest level is clear.