Investigating and Detecting Rape

by Robert Whiston  FRSA   Oct 3rd 2010

Investigating and detecting recorded offences of rape” is one of numerous reports into rape published by the Home Office (they appear obsessed with the subject !).  

It was as long ago as 2002 that police, in a research paper published by the Home Office, conceded that it would be more than a good idea if society improved its understanding of criminal careers. In particular, it thought that exploring “the previous criminal histories of those who had been convicted of murder or serious sexual assault (SSA)” might prove fruitful.[1]

However, this idea fell instantly onto stony ground and every research paper since has focused purely and relentlessly on putting more rapists in jail and on restricting a defendant’s human right to a fair trial when accused of rape .

The raison d’être for this 2007 report (Investigating and detecting recorded offences of rape), was to understand the “reasons behind the decline in recorded detection rates for rape since 1997” and “why marked variations in detection rates exist in different forces”.

The unspoken assumption here is that there should be more ‘detections’ and more convictions as the reported of rapes increase. No thought is given to the fact that proof has to be first established to get convictions rates to increase.

As regards marked variations there seems to be an ignorance of the different types of people and demographics that different policies forces encounter. We should not expect, nor should there be any compulsion to standardise the rate of ‘reported’ rapes across the country. 

One section entitled ‘Attrition of cases through the criminal justice system’ informs us that in a sample of 676 rape cases 15% were ‘no crimed.’ So broadly speaking the number of false allegations, if the sample is representative, is 15%. 

However, reporting the levels of false allegations was not the purpose of the report and with a little ingenuity and appropriate pressure placed on the police – if they were at all involved – the more appropriate for HO and feminist purposes figure was reduce to 8%.

We are told that the police were happy that this 8% figure represented the number of ‘false allegations’, and tellingly by “the majority through the victim’s [own] admission that this was the case.” One can’t wondering why they wouldn’t be pleased and acquiesce if it got the searchlight off them faster than if they had contended the figures.

Table 4.1, taken from the report is most interesting in displaying no-crimed events. It is reproduced here (right).

Of the false allegations (i.e. 8% or 15%), 86% of them could be verified as having not taken place. This is the portion of the chart opposite coloured Blue. The next largest segment in the no-crimed total is where the ‘victim’ failed to substantiate the allegation.

Added together this 6% of allegation which could not be substantiated brings the total to 92%. A further 5% of alleged rape offences took place in another police forces jurisdiction (making 97%).

As an aside, it is worth noting at this point that by Sept 2010 correspondence showed the Home Office had accepted that being falsely accused created  a second victim. A minor point perhaps but indicative of a shift of thinking even within the Home Office.

The analysis of  Tble 4.1 reveals this breakdown:

  • Verifiable information that no crime took place      (86%)
  • Victim failed to substantiate the allegation                (6%)
  • Offence took place in another force                             (5%)
  • Incident recorded as crime in error                             (1%)
  • Other                                                                                   (2%)

The Numbers

Government press release which then become newspaper stories frequently talk only in terms of percentages (lobby groups do the same). So let’s translate these percentages into real numbers. 

If there were 676 rape cases in this particulate sample, and if false allegations were found in 15% of cases it would mean that 101 out of 676 cases were false, or one in 6½ claims. 

That is obviously far and away above the consensus level of 2%, so what can be done ? Vigorous pruning is the obvious solution. We are not told how this was achieved save to say it was ‘appropriate’. 

When the number of false allegations are cut down in size to 8% (approx 50% of the original 15% figure) the number of false allegations would then number only 55. Suddenly, only one in 12 reported rapes are false allegations. A much more acceptable outcome. 

How the pruning is achieved is not well documented by the methodology section refers extensively to the subjectivity of various researchers since the 1980s who have adopted all manner of means to get at the true figure net of false allegations or vice versa quantify false allegations. 

Before moving on to other techniques used in doctoring the data it is worth maintain the reports other crucial findings. These are that in around 7 out of 10 cases are lost from the system between an offence being ‘crimed’ and charges being brought.

It is astonishing to then learn that of the ‘crimed’ cases that did not result in a charge, the victim withdraw her complaint in 35% of instances and there was sufficient evidence in 40% of cases (these were the two most common reasons identified for not charging). 

Again, let’s translate these percentages into real numbers. If there were 676 cases under review and 7 in 10 (i.e. 70%), are lost from the system then 473 reports of rape resulted in no legal action being taken (103 were actioned, ie 676 – 473).

Of those 473 reports of rape that resulted in no action, 165 women withdrew their rape claim, i.e. 35%. There was insufficient evidence in 189 cases

However, 165 + 189 = 354, not 473. And if we apply the same percentages to 676 we arrive at 506 (i.e. 236 +270). 

Perversely, a major concession contained in the report is that stated the rape conviction rate (6%) is found not to be true. The detection rate for the sample as a whole was 30%, meaning in total 72 offences resulted in a conviction. This “equates to a conviction rate of 13%” but as we know rape is the only criminal offence where the attrition rate is used [wrongly] in place of the true conviction rate.

The reason given for this higher figure (i.e. 13%) is that it includes convictions for lesser offences such as indecent assault which the 6% figures does not (presumably the 6%t counts only attrition as it applies to alleged rape offences). 

The technical ways in which raw data can be manipulated are countless. We are given an insight in to one adopted by Liz Kelly – a favoured writer of the Home Office.

In her 2005 paper she urges researchers into rape and false allegations thereof to adopt appropriate decisions when adjusting for ‘no-criming’. She advises against grouping together different types of no crime into a single undifferentiated group. The effect of this is to blur the distinction between cases where 1/. there is a genuine, initial concern that a rape has been committed but ultimately ‘no evidence of an assault’ is found and 2/. those cases that are false allegations. This immediately and drastically lowers the numbers of rapes initialing labeled as ‘false allegations.’ 

The category ‘Doctor/FME’ (forensic medial examination), where the assessment is that no offence took place’ broadly corresponds to Liz Kelly’s ‘no evidence of assault’ category. If no offence took place then common sense should tell us that it is a false allegation. But under the Kelly doctrine it becomes a possible rape and goes into the 1/, above, where it is believed that a genuine rape has taken place initially. 

FME cases account for 42 instances in this sample. In just over half of these cases (51% i.e. 21), the decision to ‘no crime’ was due to the victim admitting that no crime took place.

In a further 10 instances evidence came to light which contradicted the victim’s account of events.  Again, the narrative then tells of an ‘adjusting for offences which were inappropriately classified’, but no premise as to why is stated. All we are told is that they include ‘offences which might be considered to be false allegations of some kind or another’ and that they ‘accounted for 8% of the entire sample (where the sample size was 52 – not the 676 – though even this is not made clear). 

Cynically, one could view all these processes as progressive pruning to arrive at a preordained and acceptable figure.

Characteristics of False Allegations

Having once achieved a more diverse range of allegations under the no crimes label (from plain false allegation to FME and then ‘administrative error’), the authors conclude there “is little value in examining this as a single category.”

Instead the authors decide it would be ‘more fruitful to examine whether false allegation cases have particular characteristics that distinguish them from crimed cases.’

Straight away one can think of attention-seekers, the crazed, fantasists and those bent on revenge or spite.

At first it looks promising; the authors come up with a total of 22 victim and offence variables for  consideration. None of the obvious appear and instead the authors pick out the following as “significant variables” (Table 4.2).

Relationship             Place of initial contact         How offence reported      Extent of injuries      Location of first offence           Employment status            Police Force

These categories were then cross-referenced with stranger, intimate, or acquaintance etc.

Surprise, surprise, these studious academic types found ‘no observable connection’ between any of the variables and people who make false allegations.

The table simply identifies those offence characteristics which are more commonly found in false allegation cases compared to all other offences. The overwhelming majority of reported offences featuring any of these characteristics will still be crimed offences.

Those ‘characteristics which are more commonly found in false allegation cases’ lead to no useful or clear cut answer. For instance, the table (4.2) finds that stranger relationships are over-represented in false allegation cases where they account for 37% of false allegations whereas they account for only 14% of all crimed rape offences.

Something termed the “Initial contact” refers to a ‘public place’ and again false allegation are over-represented at 47% compared to 23% of all crimed cases.

These are so nebulous, ie generic, ill-defined, imprecise etc, as to have little practical relevance  for members of the public.

Worryingly the report describes those who have made previous false allegations or who are ‘vulnerable victims’ as not significantly over or under-represented in the crimed and ‘no-crime’ offences.

Some elements in ther study might prove useful in questioning the truthfulness of claims in future cases. It may alarm some to realise that 87% of false allegation were made direct to the police (as opposed to other methods of contact) compared with 67% for all offences, presumably rape, that were crimed.

In false allegation cases the victims did not suffered any injuries in 85% of  instances compared with 66% where the offence is crimed.

One of the last points the report makes in this section is that students are over-represented amongst false allegation cases (23%). This compares to 10% of all other crimed cases.

Reasons for Not Charging

This brings us to the question why and in what circumstances are charges by the police and allegation by the complainant are actually dropped. This is particularly critical as it measures the drop out rate not in the outlandish cases but in cases where the police deem there to be a legitimate cause to believe an offence has occurred.

Overwhelmingly, and in top spot at 40%, is ‘a lack of evidence.’ Given the wonders of DNA and forensic science generally plus the fact that rape is one of the few crimes where the victim knows the alleged criminal in 90% of instances, the 40% is truly astonishing.

The second ranking reason, at 35%, is that the complainant withdraws her allegation. This is sometimes discounted by those seeking to punish anyone accused of rape by implying intimidation on the part of the accused, or by the police being difficult and not supporting the alleged victim, etc, etc . But having looked into this aspect a complainant can only make a withdrawal when she has satisfied strict Home Office criterion and the police are bound by specific regulations on this matter.

The interim results of PAFAA study presently underway into false allegation give the lie to some of those claims. Police forces are shown to be reacting to rape allegations very promptly; to commonly organise a 20 to 30 strong detective teams for the investigation; and spending upwards of £13, 000 to £300,000 on each case (where costs are mentioned).

Shown below is Figure 4.3, a pie chart taken from “Investigating and detecting recorded offences of rape.” It details in percentages, the reasons for not charging a person after a woman has made a rape allegation and where the offence is recorded as a legitimate incident.

         Figure 4.3. Reasons for not charging – all crimed cases not resulting in  charge  (reasons for not charging, i.e. proceeding with the case) 

 The  analysis looks like this:

  • Insufficient evidence                                           40%
  • Complaint withdrawn                                         35%
  • Victim reluctant to assist                                    11%
  • No suspect                                                            10%
  • No formal complaint made                                  2%
  • Suspect identified but not apprehended           2%
  • Not in the public interest                                        1%

Withdrawing Accusations

The number of cases withdrawn by the victim at any of the possible stages (in this study) accounted for 39% of crimed cases.

This percentage is based not on the original 676 sample size but on the 569 cases that were crimed and “where it was known whether or not the victim had withdrawn” their complaint / allegation a difference of 107 (676 – 569).

Nothing is said of the number of allegations where it was not known whether or not the victim had withdrawn their allegation.

At another point in the report it states that of the sample size (676), 593 were crimed cases and 83 no-crimed cases. This is at odds with the 569 and 506 listed above and the 83 doesn’t tally with the 101, 103, 189, 103, mentioned earlier.

Ideally, an explanation as to why these figures differ would be appreciated but the text of the report appears to offer none.

At first sight Fig 4.1 (below), looks to give the answers but on closer examination it  reveals only the attrition rate for offences that have been crimed – not those that have been no-crimed – and we have a new set of numbers not seen before ie 576, 364, 173 etc.

Only the totals 72 and 676 (see Fig 4.1) have appeared before in the report (and they are included earlier in this précis).

            Figure 4.1: Overall attrition within the sample

Of the 6 lesser categories the one that stands out is “No suspect.” It could be argued that the lack of a suspect either a person known to the victim or one that could be categories as ‘a stranger’ underlines the possibility that the true level of false allegation is around 10% if not more.

We hear much (and often) from politicians such as Harriet Harman MP about rapists let loose on our streets or of unapprehended rapists free to roam and strike again but at 2% (for suspect identified but not apprehended) is that stance credible ?

“No formal complaint made” is a little odd, given that the police are obliged to convert any and all reported crimes such as rape into the more serious ‘recorded’ crime (they have for a long time been a ‘notifiable offence’). One can only suppose that the far fetched or hysterical or drunkenness of the would-be complainant has ruled out their complaint from entering deep into the system.

 “Not in the public interest” probably arises when the CPS take the view that the case will most likely fail or be discontinued by the judge due to the flimsiness of the case or unreliability of the witness herself.

That said the latter account for only 5% of cases for not charging even where the offence has been crimed, i.e. recorded as a legitimate incident.

Waiting to Know

For those at the receiving end of a false allegation it might be helpful to know at what stage one can expect the charges, or threat of being charged, to be dropped.

Around 1 in 10 allegations are withdrawals  by the victim after a suspected offender had been charged; the vast majority of withdrawals take place during the police investigation stage of the process (Fig 5.1).

Figure 5.1: Victim withdrawal by point within the investigative process. 

 The analysis at this stage looks like this:

  • Withdrawal after the arrest but before being charged      33%
  • Withdrawal before the suspect is apprehended               58%
  • Withdrawal after the suspect is charged                              9%

National Picture

Given the above information does it transform our perception of the national picture when the ratios and percentages are fed into the equation ?

The number of ‘reported rapes’ in England & Wales were 13,991 in 2009 – 10. We can ignore for this exercise how many of these claims resulted in proceedings being taken, or how many individual were charged, or how many convictions resulted

  • If there were, indeed, 13,991 allegations of rape in England & Wales and the rate of false allegations is 15% then 2,099 of the alleged rapes are in fact false allegations.
  • Compare this with the preferred 2% figure which gives rise to only 280 – a figure so small it can be confidently dismissed in official circles as trifling.
  • Even the substantially reduced figure of 8% results in 1,120 occasions of false rape allegations – a number that dwarfs the official number of 280.

To Recap

Of the original sample of offences, i.e. 676 rape offence examples –

  • 85% were actually crimed
  • 15% were no crimed

Of the original sample of offences, i.e. 676 examples, 576 cases were crimed by the various police forces. Of these:

  • 32% were classified as detected
  • 166 of these offences resulted in an offender (or offenders) being charged;
  • 7 cases resulted in a caution
  • 9 cases resulted in a non-sanction detection

The report goes on to state this “corresponds to a ‘sanction detection rate’, that is, the number of charges plus cautions divided by crimed offences of 30%.”

Of those offences that resulted in a charge:

  • 130 went for trial
  • 72 offences ended up with the offender being convicted
  • 7 cautions were issued
  • In 32 offences a suspect was convicted for rape
  • In a further 40 cases the suspect was convicted for a lesser offence.

After the attrition process (see Figure 4.1), the proportion of crimed offences resulting in a rape conviction (alone) was 6%. The proportion of crimed offences resulting in a conviction for rape or another lesser offence was 13% (in 5 cases the lesser offence was not a sex offence).

In total there were 39 occasions when the offender was also convicted of an additional but lesser offence(s).

As the report admits;

A conviction rate of 13% is considerably higher than the frequently quoted statistics on the conviction rate for rape.

The only problem with the 13% (above) is that it is used as a substitute for the more common 6% conviction ratio. Arguably, both are actually ‘attrition’ ratios but the Report, in its 118 pages, fails to clarify the matter.

It then gives the reasons; the 6% is due apparently to an ‘administrative’ method where the figures for convictions for rape alone are used. Where an offender charged with rape is convicted of a lesser crime the offender is excluded form this ‘administrative’ method but not the offence.

Where the offenders is found guilty of either rape or a lesser crime, say, indecent assault or violence, the total conviction rate becomes 13% (see reference to ‘pruning’ above).

In this 2007 report, convictions for lesser offences represent a sizable proportion of total outcomes and accounted for 56% of all convictions.

We also have to recall that that when Labour came to power in 1997 there were 6,281 reported rapes. By 2001 it had risen to 8,990 reported rapes per annum and hit 11,441 in 2002 – an increase of 82%.

This must surely be unparalleled in any other sphere of social interchange or social science. And it is no comfort whatsoever to hear the increase explained away as shooting up “partly because of new ways of collecting statistics.”  If anything, that should make us more wary that we are not being told, the whole truth, and nothing but the truth.

NB. The risk of rape threat and the threat of rapists to the general public will be discussed in a future article.


[1] “Murder and Serious Sexual Assault: What criminal histories can reveal about future serious offending”. (2002). Police Research Series Paper 144,


“Investigating and detecting recorded offences of rape” (2007)

“Guidelines for medico-legal care for victims of sexual violence” (UN)



2 Responses to Investigating and Detecting Rape

  1. Verity Justice says:

    Thank you for this article Robert. The truth is out there, we just have to force it out into the open.

  2. Verity Justice says:

    Thank you for this article Robert.

    The truth it seems will have to be forced out of the authorities involved.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: