False allegations of rape: the true extent remains unknown.
The mental health impact on a man of having a false allegation of rape made against him can be devastating, and it is not uncommon for such men to become suicidal. However, the quest to estimate the rate of false allegations of rape is like hunting the snark: one does not even know what it is one is looking for, let alone where to find it. Unlike the snark, however, we know it exists. I focus here on female victims. Perpetrators of rape in English law are necessarily biological males.
What constitutes an “allegation”? An anonymous claim in a social media post, such as #MeToo? Or an allegation made within a family court submission over contested child contact where bitterness is the ruling passion? Or a report to the police which may not result in a prosecution – or perhaps only those that do?
And what does “false” mean? Certainly an allegation which is entirely fabricated, where no sexual encounter at all ever took place. But what if sexual intercourse did take place and is not disputed, but consent is disputed? How can we ever know the truth about consent, really? Conviction does not necessarily mean the defendant was truly guilty, and exoneration does not necessarily mean he was innocent.
Susan Brownmiller’s 1975 book ‘Against Our Will: Men, Women and Rape’ claimed that just 2% of rape complaints were false. For decades thereafter various sources seemed to confirm that 2% figure, but Edward Greer (2000) traced their sources back to Brownmiller. Brownmiller’s source was a speech by judge Lawrence Cooke on 16 January 1974 in which he said, “according to the Commander of New York City's Rape Analysis Squad, only about 2% of all rape and related sex charges are determined to be false”. Of any evidence of data, reporting, or other provenance behind that claim, there is none.
However, in the sorry tale of misinformation on this topic it is the Crown Prosecution Service (CPS) who take the ignominious prize. Their 2013 report co-authored by Alison Levitt QC carried a Foreword by Keir Starmer, who was then Director of Public Prosecutions, (Levitt, 2013). It is an example of what I call “constructive mendacity”, by which I mean there is no sentence which is an undisputable untruth, but the whole construction is designed to grossly mislead. Starmer wrote, “One such misplaced belief is that false allegations of rape and domestic violence are rife. This report presents a more accurate picture”. But it does not. The report did not even address those thorny issues. Instead it merely looked at the rate of prosecutions for false allegations. But it gave the impression that it had addressed the harder question by stating, “The review has allowed us to examine the suggestion that false allegations of rape and/or domestic violence are rife”.
The report found 0.62% of prosecutions for rape resulted in a prosecution for false allegation. But no one ever disputed that the rate of prosecutions for false allegation was very low. It is beyond credibility that lawyers would fail to appreciate the difference between the underlying false allegation rate and the rate of prosecution for false allegation, hence my claim of constructive mendacity. But ever since, people have been quoting the CPS report as demonstrating that the rate of false rape allegations is only 0.62%.
“The false allegation rates from 19 sources [is estimated to be] anywhere between 1.5% and 90%... not very helpful.”
Evidence on the rate of false rape allegations has been compiled by Rumney (2006) and by Lisak et al (2010). All their sources are based on police reports and so all refer to percentages of allegations made to the police. Let me just flat-footedly give you the false allegation rates from 19 sources which Rumney quotes in his Table 1:
1.5% to 10%; 2%; 3% to 22%; 3% to 31%; 3.8%; 8%; 10.3%; 10.9% to 25%; 11%; 11.8%; 18.2%; 20%; 22.4%; 24%; 38% to 41%; 41%; 45%; 47%; 90%.
You take my point: anywhere between 1.5% and 90% is not very helpful. The main reason for the huge spread is the criterion used to define “false”. Broadly there are two:
A. Categorise an allegation as false only if a thorough investigation based on evidence has established that no crime was committed or attempted.
B. A range of less certain criteria, including the police recording the allegation as a “no crime” but without clear evidence behind that decision.
The advantage of criterion A is that it would establish a firm lower bound to the rate of false allegations. The disadvantage is that this lower bound may be far smaller than the true rate because the police do not usually concentrate on performing a “thorough investigation” to prove that a crime did NOT occur.
The veracity of the estimates resulting from criterion B will depend on the exact procedure. It might give an accurate estimate, or it might be a serious over-estimate or under-estimate. Some of Rumney’s sources claimed that the police had used a no-crime criterion which required the complainant to have admitted there was no crime, but doubt was expressed that the police had truly applied this criterion consistently. There is a suspicion in many cases that the police may have been rather too ready to ‘no-crime’ reports. In the worst cases, and noting that the sources go back as far as the 1970s, the reasons for assigning as a no-crime might have been disconcertingly antediluvian (e.g., “no sign of a struggle”, etc.).
Lisak et al concentrated upon criterion A and found a false rate of sexual assault on this basis to be 5.9%. They also found seven other sources whose data could be analysed in this way and these gave results 2.1%, 2.5%, 3.0%, 6.8%, 8.3%, 10.3%, 10.9%.
Lisak et al conclude that “the prevalence of false allegations is between 2% and 10%”. But this is incorrect (and the referees of the publication were remiss in letting that conclusion go through). These figures are all lower bounds. If they were all applicable to the same population (an issue I have avoided discussing) then 10.9% would be the final lower bound – which still might be far lower than the correct value.
“…lack of a consistent definition of what constitutes a false rape allegation, as well as variations in recording practices by police and others in the CJS, make accurate assessment of the true extent of such allegations very difficult”.
A Ministry of Justice report, (MOJ, 2010), states that “Various studies have estimated that 8–11% of rape allegations in England and Wales are false”, but it notes that “lack of a consistent definition of what constitutes a false rape allegation, as well as variations in recording practices by police and others in the CJS, make accurate assessment of the true extent of such allegations very difficult”.
To drive home how serious these misgivings are, and just how misleading Lisak et al’s conclusion might be, take a closer look at Lisak et al’s own data. They classified the data into (i) “false report” (based on criterion A), (ii) “did not proceed”, (iii) “did proceed”, (iv) “insufficient information”. Out of 136 cases there were 8, 61, 48 and 19 in these categories respectively. Any of the 128 cases which did not meet criterion A might have been a false allegation – but just wasn’t proved to be. Suspicion alights particularly on the 61 cases (45%) classed as “did not proceed”. Quote:
“This classification was applied if the report of a sexual assault did not result in a referral for prosecution or disciplinary action because of insufficient evidence or because the victim withdrew from the process or was unable to identify the perpetrator or because the victim mislabelled the incident (e.g., gave a truthful account of the incident, but the incident did not meet the legal elements of the crime of sexual assault).”
A priori any of the 45% of cases in this category might be a false allegation. Clearly one cannot conclude otherwise when this categorisation included cases where there was “insufficient evidence or the ‘victim’ withdrew”. This will be quite typical of studies based on criterion A.
We are thrown back on the brute fact that, for the large bulk of cases, we do not know if the allegation was false or not.
But there is another approach to the question. I have not attempted to discuss how the false allegation rate might depend upon population (country, demographic), though the pursuit of a single global figure is obviously a fool’s errand. However, there is another issue: the false allegation rate may not be static but might have varied substantially. In the UK, at least, I suspect it has. Here I bring up to date an argument presented in Collins (2019) – and then critique it myself and invite others to do so too.
There are three numerical ingredients to this argument, as follows,
(i) The estimate of the true number of women victims of rape based on the Crime Survey for England and Wales (CSEW), for which I use (CSEW, 2020);
(ii) The CSEW reported percentage of women victims of rape who report to the police (CSEW, 2021);
(iii) The number of police recorded crimes of rape of women (Crime Data, 2022).
I concentrate on adult women (age 16 or over). Item (i) provides the percentage of women aged 16 to 59 who reported rape in the last year, for years 2004/5 to 2019/20. This is based on the sample surveyed by the CSEW and will be subject to a fairly wide error bar. The number of women aged 16 to 59 in England and Wales increased from 16 to 18 million over the period 2005 to 2020. Multiplying this population by the percentage of reported rape victims gives our best estimate of the number of genuine female rape victims in this age range, see Figure 1.
Item (ii) tells us that the CSEW asked what percentage of these rape victims told the police in years 2014, 2017 and 2020. In 2014 they reported 17%, and in 2017 and 2020 (combined) 15.8%. I shall assume that 17% applies from 2005 to 2016 and 15.8% thereafter. By multiplying the above best estimate of the number of genuine rape victims by this percentage gives us our best estimate of the number of reports of rape to police that are genuine (for women aged 16-59), see Figure 1 (above).
Finally, item (iii) provides the number of rape crimes recorded by the police for adult women (16 and over), which, for our purposes, I identify as the number of “allegations”. This is also plotted in Figure 1. Note that this is a very different measure of “allegations” from that used in most of the sources quoted by Rumney because it refers to the numbers of complaints which the police have classed as crimes. Thus, the main source of contention in interpreting the data quoted by Rumney is avoided, namely the inclusion as “allegations” those complaints ultimately designated as “no crime” by the police.
Figure 1 shows there was a large drop in the CSEW-based estimate of rapes between 2012 and 2013, the reasons for which are unknown (to me). The volume of rape crimes recorded by the police has increased since 2012. The reason for this is, at least in part, due to poorer police recording practices in earlier years, and this is acknowledged in CSEW (2017). The same source does, however, also identify the effect of Operation Yewtree (the post-Savile investigations) in increasing the number of reports to police, including encouragement to report historical offences.
Operation Yewtree started in 2012, and Figure 1 shows that, prior to that date, the number of police recorded rape crimes was less than our estimate of the number of genuine reports of rape to the police, perhaps partly because the police were “no criming” too many complaints, and partly simply due to poor record keeping. From 2014 onwards the number of police recorded rape crimes has exceeded our estimate of the number of genuine reports of rape to the police, the difference yielding my estimate of the number of false allegations. Figure 2 shows this as a percentage of the number of police recorded rape crimes, i.e., the apparent false allegation rate. It has become disconcertingly high.
I invite readers to deconstruct the argument and to check the data. My own observations are as follows:
The weak link is the CSEW estimate of the percentage of victims reporting to the police (17% or 15.8%) because this relates to women’s experience over life (since age 16). It is possible, even likely, that reporting rates have been substantially greater in recent years. If so, the implied rate of false allegations from 2014 would be smaller than given in Figure 2.
Another issue is that the CSEW data relates to the number of victims, whereas the police data relates to the number of crimes. This means that the more women who are unfortunate enough to be raped more than once in a year, the more the CSEW-based victim data would under-estimate the number of true reported rapes, and the argument fails.
Readers will note that the police data relate to ages 16 and above, whereas the CSEW data relate to ages 16 to 59. However, I do not believe this is important. The CSEW also provides rape rate data for ages 16 to 74 and these data imply that the number of victims aged 60 and over is not large and does not significantly perturb the argument.
“Claims that the rate of false rape allegations is less than ~10% are actually based on lower bounds. There is no compelling evidence that the true rate might not be far greater, and much highly uncertain evidence that it might be”
Conclusions
[1] The term “rate of false rape allegations” is highly ambiguous; it has no agreed definition.
[2] Reliable quantification of the rate of false rape allegations, even given a specific definition, remains elusive.
[3] Claims that the rate of false rape allegations is less than ~10% are actually based on lower bounds. There is no compelling evidence that the true rate might not be far greater, and much highly uncertain evidence that it might be (Figure 2).
[4] There is some contestable evidence that false rape allegation rates have increased since Operation Yewtree, but quantification is very uncertain.
See here for detailed referencing.
Further information
If you have been affected by false allegations of sexual abuse then contact The Defendant helpline, 0300 124 5098, which is open Mon to Fri 6-10pm.
For general information email info@thedefendant.org.uk
Or visit the website https://thedefendant.org.uk/
Scroll down to join the discussion
Disclaimer: This article is for information purposes only and is not a substitute for therapy, legal advice, or other professional opinion. Never disregard such advice because of this article or anything else you have read from the Centre for Male Psychology. The views expressed here do not necessarily reflect those of, or are endorsed by, The Centre for Male Psychology, and we cannot be held responsible for these views. Read our full disclaimer here.
Like our articles?
Click here to subscribe to our FREE newsletter and be first
to hear about news, events, and publications.
Have you got something to say?
Check out our submissions page to find out how to write for us.
.
