Home Office Research Study 293 A gap or a chasm? Attrition in reported rape cases

Full text

Home Office Research Study 293
A gap or a chasm?
Attrition in reported rape cases
Liz Kelly, Jo Lovett and Linda Regan
Child and Woman Abuse Studies Unit,
London Metropolitan University
The views expressed in this report are those of the authors, not necessarily
those of the Home Office (nor do they reflect Government policy).
Home Office Research, Development and Statistics Directorate
February 2005

Home Office Research Studies
The Home Office Research Studies are reports on research undertaken by or on behalf of
the Home Office. They cover the range of subjects for which the Home Secre t a ry has
re s p o n s i b i l i t y. Other publications produced by the Research, Development and Statistics
Directorate include Findings, Statistical Bulletins and Statistical Papers.
The Research, Development and Statistics Directorate
RDS is part of the Home Office. The Home Off i c e ’s purpose is to build a safe, just and tolerant
society in which the rights and responsibilities of individuals, families and communities are
p roperly balanced and the protection and security of the public are maintained.
RDS is also part of National Statistics (NS). One of the aims of NS is to inform Parliament and
the citizen about the state of the nation and provide a window on the work and perf o rm a n c e
of government, allowing the impact of government policies and actions to be assessed.
T h e re f o re –
Research Development and Statistics Directorate exists to improve policy making, decision
taking and practice in support of the Home Office purpose and aims, to provide the public
and Parliament with information necessary for informed debate and to publish information
for future use.
First published 2005
Application for reproduction should be made to the Communications and Development Unit,
Room 201, Home Office, 50 Queen Anne’s Gate, London SW1H 9AT.
© Crown copyright 2005 ISBN 1 84473 555 9
ISSN 0072 6435
A gap or a chasm? Attrition in reported rape cases

i
Foreword
In 1998 the Home Office announced the Crime Reduction Programme (CRP), which aimed
to develop and implement an integrated approach to reducing crime and making
communities safer. As part of this programme the Violence Against Women Initiative (VAWI),
was launched in July 2000, and specifically aimed to find out which approaches and
practices were effective in supporting victims and tackling domestic violence, rape and
sexual assault. Thirty-four multi-agency victim focused projects were funded and aimed to
develop and implement a range of interventions for various population groups in a number
of different settings and contexts. The projects were originally funded until the end of March
2002; however, 24 of these projects had their funding, and in some cases, their evaluations
extended until the end of March 2003. A further 24 ‘Second Round’ projects were funded
in March 2001; however, these were provided with specific service provision funding and
were not evaluated by the Home Office.
For evaluation purposes, the projects were divided into nine packages and projects with
similar solutions or strategies, as well as those operating in the same contexts were grouped
t o g e t h e r. Seven diff e rent independent evaluation teams were commissioned to assess the
p rojects in terms of their development, impact and costs. The findings from all of the
evaluations have been collated and a series of re s e a rch re p o rts and concise practitioner
focused guides have been published.
This re p o rt presents findings from the evaluation of two Sexual Assault Referral Centre s
(SARCs), one non-centre based support service and three comparison areas where there
was no specialist provision for victims. The re p o rt provides a valuable insight into the
p rocess of attrition in re p o rted rape cases in the Criminal Justice System (CJS), with a
specific focus on early withdrawal by complainants. The report provides recommendations
for good practice which will be of interest to a range of professionals who have a role or an
interest in reducing attrition in reported rape cases in the CJS.

Crime Reduction Programme: Violence Against Women
Initiative other reports in the series
Domestic Violence
Douglas, N., Lilley, S.J., Kooper, L. and Diamond A. (2004) Safety and justice: sharing personal
information in the context of domestic violence- an overview.Home Office Development and
Practice Report No. 30. London: Home Office.
Hester, M. and Westmarland, N. (2005) Tackling Domestic Violence: effective interventions and
approaches. Home Office Research Study No. 290. London: Home Office.
M u l l e n d e r, A. (2004) Tackling Domestic Violence: providing support for children who have
witnessed domestic violence. Home Office Development and Practice Report No. 33.
London: Home Office.
Parmar, A., Sampson, A. and Diamond, A. (2005) Tackling Domestic Violence: providing advocacy
and support to survivors of domestic violence. Home Office Development and Practice
Report. No. 34. London: Home Office.
Parmar, A., Sampson, A. and Diamond, A. (2005) Tackling Domestic Violence: providing advocacy
and support to survivors of domestic violence from Black and other minority ethnic
communites. Home Office Development and Practice Report. No. 35. London: Home Office.
Taket, A., Beringer, A., Irvine, A. and Garfield, S. (2004) Tackling Domestic Violence: exploring the
health service contribution. Home Office Online re p o rt 52/04. London: Home Off i c e .
[Available online at: http://www.homeoffice.gov.uk/rds/onlinepubs1.html]
Taket, A. (2004) Tackling Domestic Violence: the role of health professionals. Home Office
Development and Practice Report No. 32. London: Home Office.
A gap or a chasm? Attrition in reported rape cases
ii

Rape and sexual assault
Lovett, J., Regan, L. and Kelly, L. (2004) Sexual Assault Referral Centres: developing good practice
and maximising potentials. Home Office Research Study No 285. London: Home Off i c e .
Regan, L., Lovett, J. and Kelly, L. (2004) Forensic nursing: an option for improving responses to
reported rape and sexual assault. Home Office Online Report 28/04. [Available online at:
http://www.homeoffice.gov.uk/rds/onlinepubs1.html]
Regan, L., Lovett, J. and Kelly, L. (2004) Forensic nursing: an option for improving responses to
reported rape and sexual assault. Home Office Development and Practice Report No. 31.
London: Home Office.
Skinner, T. and Taylor, H. (2005) Providing counselling, support and information to survivors of
rape: an evaluation of the Surviving Rape and Trauma After Rape (STAR) young person’s
project. Home Office Online Report 51/04.
[Available online at: http://www.homeoffice.gov.uk/rds/onlinepubs1.html]
iii

A gap or a chasm? Attrition in reported rape cases
iv
Acknowledgements
We would like to express our thanks to the many people whose co-operation and assistance
was essential to conducting this evaluation. Firstly, to the Crime Reduction Pro g r a m m e :
Violence Against Women Initiative, and particularly Alana Diamond for her input and
support throughout. Secondly, to the staff at St Mary’s, REACH and STAR, police officers in
the Greater Manchester, Northumbria, West Yorkshire, Metropolitan (Brent and Newham)
and Thames Valley force areas, and the many individuals in associated agencies, whose
commitment was critical to the success of this project. Contributions in terms of data
collection and preparation were also made by Dr Tina Skinner, Dr Rosemary Barbare t ,
Helen Taylor, Val Balding and Carolanne Lyme.
F i n a l l y, our sincere gratitude to all the re s e a rch participants who chose to share their
experiences with us. Without them our understanding and this report would have been much
the poorer.

v
Contents
Executive summary ix
1. Introduction 1
Structure of report 2
2. Aims and methodology 3
Data collection 4
Data analysis 7
Defining terms 7
Limitations 8
3. What we know about rape and attrition 13
How common is rape? 13
Incidence and prevalence in the UK 14
Incidence and prevalence research internationally 16
Summary 16
New data on patterns and trends 17
Numbers and age profile of service users 17
Relationship to offender 20
Location of assaults 21
Violence and injuries 22
Time frames 22
Employment status 23
Ethnicity 23
Multiple assailants 23
Multiple attendees 23
Summary 24
Research on attrition 24
UK data and research 25
Attrition internationally 29
Identifying Attrition points 30
The decision to report 31
Expanding the concept of ‘real rape’ 33
Summary 34

A gap or a chasm? Attrition in reported rape cases
vi
4. Understanding attrition: the research findings 35
The attrition process 36
The attrition process revisited 39
Summary 39
Attrition point 1: reporting to the police 41
The decision to report 41
Timing of the report 43
Summary 46
Attrition point 2: no evidence of assault and false allegations 46
No evidence of assault 46
False allegations 47
Reflections 52
Summary 52
Attrition point 3: insufficient evidence 53
Summary 58
Attrition point 4: early withdrawal 59
Victims who decline to complete the initial process 59
Early victim withdrawals 63
Reflections 68
Summary 69
Attrition point 5 – CPS decision making 70
Summary 71
Attrition point 6 – court and trial 71
Factors associated with conviction 72
Reflections 76
Summary 77
5. Challenges, dilemmas and recommendations 79
Attrition in context 81
Key findings 83
Good practice is possible 84
What complainants value and need 87
Recommendations 89
Appendix 1: notifiable offences and numbers of persons prosecuted, tried and
convicted- for rape offences in England and Wales, 2001 and 2002,
including reporting and conviction rates 92
Appendix 2: comparative profile of all six study sites 96
Appendix 3: research tools and data sources 97
Appendix 4: response rates to initial service user questionnaire 104
Appendix 5: attrition process – case outcomes for all reported cases 105
Bibliography 113

vii
Tables and Figures
Table 2.1: Total and projected data collection 1st October
2000 – 31st December 2002 6
Table 3.1: Referrals to St Mary’s 1987-2002 18
Table 3.2: Comparison of prosecution outcomes in adult and child
rape cases 1998-2002 26
Table 3.3: UK research findings on attrition in reported rape cases 28
Table 4.1: Case outcomes for 2,244 cases reported to the police
where police/CPS decision known 37
Table 4.2: Recalculated attrition process: 2,284 cases reported to
police where research categories known 40
Figure 3.1: Numbers of police and self-referrals 1987-2002 19
Figure 3.2: Forensic examinations undertaken 1987-2002 19
Figure 3.3: Age profile of service users 1987-2002 20
Figure 3.4: Relationship between victim and perpetrator 1987-2002 21
Figure 3.5: Reporting of rapes by month of the year and relationship
to offender 1987-2002 22
Figure 3.6: Attrition in rape cases England and Wales 1985-2002 25
Figure 3.7: Attrition in rape cases in Scotland 1977-2001 29
Figure 4.1: Length of time between the assault and reporting to police 43

A gap or a chasm? Attrition in reported rape cases
viii

ix
Executive summary
Introduction
●Home Office figures show an ongoing decline in the conviction rate for reported
rape cases, putting it at an all-time low of 5.6 per cent in 2002. This year-on-year
increase in attrition represents a justice gap that the government has pledged to
address.
●Rape is a unique crime, representing both a physical and psychological violation.
More than with any other crime the victim can experience reporting rape as a
form of revictimisation.
●In no other crime is the victim subject to so much scrutiny at trial, where the most
likely defence is that the victim consented to the crime. Powerful stere o t y p e s
function to limit the definition of what counts as ‘real rape’.
●This report draws on material from two evaluation projects funded by the Crime
Reduction Programme Violence Against Women Initiative (CRP VAWI). It combines
analysis of St Mary ’s Sexual Assault Referral Centre historic database and
prospective case tracking of reported rape/sexual assault cases across six sites
over 17-27 months.
Aims and methodology
●The primary aim of the study was to increase understanding of attrition, with an
emphasis on early withdrawal from the criminal justice system (CJS) process by
complainants.
●Data were collected from three Sexual Assault Referral Centres (SARCs) and three
Comparison areas selected to reflect a combination of metropolitan, inner city
and rural areas.
●A multi-methodological strategy was employed linking quantitative and qualitative
data. The base sample is 3,527 cases, which were tracked prospectively through
the CJS. Sub-samples of victims/survivors opted into a series of questionnaire s
and in-depth interviews (a total of 228 participated). Within this group, police
statements and forensic reports were also accessed, where possible. These data
were supplemented by expert interviews with key informants and police officers
(n=143).

●Whilst this represents the largest data-set in the UK literature on rape and sexual
assault, there are some limitations to the study. It proved extremely difficult to
recruit research participants in the Comparison areas, so more is known about
complainants in areas where there is a SARC. Obtaining information on case
outcomes was also problematic in all areas, and there is only final outcome data
for two-thirds of the sample.
What we know about rape and attrition
●Findings from the 2001 British Crime Survey (BCS) interpersonal violence module
result in a prevalence rate of 0.3 per cent for rape of women over 16 in the year
prior to interviews conducted in 2001, and this equates to an estimated annual
incidence rate of 47,000 adult female victims of rape. Since the age of 16, 7 per
cent of women had suffered a serious sexual assault at least once in their lifetime
(5% had been raped). The same study confirms that women are most likely to be
raped by men they know, and a considerable pro p o rtion re p o r ted re p e a t
incidents by the same perpetrator.
●These findings are lower than a previous BCS-based report, and much lower than
the prevalence rate in a dedicated study on rape from the early 1990s.
●Analysis of the St Mary ’s database (covering the years 1987-2002) re v e a l e d
i n c reasing numbers of young victims aged under 20, and supports the BCS
finding that assaults are most likely to be committed by known men.
●Home Office data on re p o rted rape cases in England and Wales show a continuing
and unbroken increase in re p o rting to the police over the past two decades, but a
relatively static number of convictions, thus the increasing justice gap.
●Prosecuted cases involving children were more likely to result in conviction than
those involving adults.
●All UK studies of attrition in rape cases concur that the highest proportion of cases
is lost at the earliest stages, with between half and two-thirds dropping out at the
investigative stage, and withdrawal by complainants one of the most important
elements.
●A number of studies have found high rates of ‘no criming’, not limited to the
official guidelines for this category.
●R e s e a rch to date has identified four key points at which attrition occurs: the
decision to report; the investigative stage; discontinuance by prosecutors; and the
trial. Victim withdrawals can occur at each stage but the highest pro p o rtion is
evident at the first two points.
A gap or a chasm? Attrition in reported rape cases
x

Executive summary
Understanding attrition
●Analysis of the data for this study revealed that around one-quarter of re p o rted cases
w e re ‘no crimed’; in a pro p o rtion of detected cases no proceedings were bro u g h t .
T h e re was inconsistency in the police classification of case outcomes, part i c u l a r l y
among those that were ‘no crimed’. The vast majority of cases did not proceed beyond
the investigative stage, and the conviction rate for all re p o rted cases was eight per cent.
●T h re e - q u a rters of the overall sample re p o rted to the police. Although this was
more likely among younger complainants, high reporting levels were also evident
among those with disabilities and those involved in prostitution. Cases involving
known perpetrators were least likely to be reported. The majority of reports to the
police were made within 24 hours.
●Nine per cent of reported cases were designated false, with a high proportion of
these involving 16- to 25-year-olds. However, closer analysis of this category
applying Home Office counting rules reduces this to three per cent. Even the
higher figure is considerably lower than the extent of false reporting estimated by
police officers interviewed in this study.
●Evidential issues accounted for over one-third of cases lost at the investigative
stage. This group includes cases where: the complainant had learning difficulties,
mental health issues or was otherwise unable to give a clear account; DNA
testing was not conducted; and an offender was identified but not traced. In a
substantial number of cases in this category the decision not to proceed was
linked to victim cre d i b i l i t y. Consultation between police and Crown Pro s e c u t i o n
Service (CPS) rarely led to enhanced case-building.
●Victims who declined to complete the initial investigative process and victim
withdrawals accounted for over one-third of cases lost at the police stage. The
f o rmer was more common in areas where there was no SARC, while the age gro u p
most associated with the latter was 16- to 25-year-olds. Key factors in not
completing the initial process were being disbelieved and fear of the CJS. Police
o fficers and SARC service user participants also suggested that fear of court was
linked to withdrawal.
●Only a small proportion of reported cases were discontinued by the CPS, but this
does not accurately reflect their involvement in decision-making, since they were
consulted through advice files on many more cases at an earlier point.
●Only 14 per cent of cases reached the trial stage, with a proportion of these not
proceeding due to late withdrawal or discontinuance at court. Around half of all
convictions were the result of guilty pleas, and where a full trial took place, an
acquittal was the more likely outcome. Rates of acquittal were twice as high in
cases involving adults as those involving under-16s. xi

Challenges and dilemmas
●There is an over-estimation of the scale of false allegations by both police officers
and prosecutors which feeds into a culture of scepticism, leading to poor
communication and loss of confidence between complainants and the police.
●Police officers’ early assessments of the difficulties of prosecution and conviction
may be interpreted by complainants as discouragement to continue, and fear of
the court process can also act as a disincentive.
●There is some evidence of poor investigation and understanding of the law, and
in some cases, there has been an emphasis on discrediting features only, by the
police and CPS.
●Categorisation of cases by the police is internally inconsistent within and between
force areas.
●Alcohol consumption was implicated in a far larger number of cases than drugs.
How this contributes to attrition deserves more detailed study.
●Data from service user questionnaires and interviews showed that there are
specific elements that would improve responses to reported rape. These include:
the availability of female practitioners; a culture of belief, support and respect;
access to clear information at appropriate points in the process; being kept
informed about case progress; and courtroom advocacy that does justice to the
complainant’s account.
●F rom the perspective of complainants, the diff e rence in perceptions between
themselves and CJS personnel was too often not just a gap but a chasm. If,
however, each point in the attrition process is examined in detail, what emerges
is a series of smaller gaps, each of which could be bridged by targ e t e d
interventions.
Recommendations
●The most important recommendation from this study is that a shift occurs within the
CJS from a focus on the discreditability of complainants to enhanced evidence-
gathering and case-building.
●The possibility of inter-agency work between police and specialist support
agencies should be further explored with respect to providing support during
statement taking and providing information on case status.
●Home Office counting rules for rape and sexual offences need to be revisited in
light of the findings from this research: whether cases where there is no evidence
A gap or a chasm? Attrition in reported rape cases
xii

Executive summary
of assault should remain classified as assaults needs to be addressed, and
mechanisms should be developed to correct inconsistent classification.
●Guidance should be issued within the police about the timing and content of
p roviding complainants with information about the legal process and the
likelihood of a conviction. At a number of points, from the earliest contact on,
victims/witnesses perceived this as discouragement from proceeding with the
case.
●Development in all areas, through SARCs and/or rape crisis centres, of proactive
follow-up of all complainants reporting rape to the police. Such services would
add the most value if they combined practical support, advocacy and case-
tracking.
●Piloting of early case conferences between police, CPS and counsel, to explore
potential evidential weaknesses, and whether these might be addressed through
additional evidence, expert testimony, research findings or courtroom advocacy.
●An accredited training course for prosecution barristers, and the development of
some form of sanction where briefs are returned or counsel changed just before
trial.
●Mechanisms within CPS to monitor court room prosecution advocacy, and learn
lessons from convictions, acquittals and especially guilty pleas.
●Increased recognition of the significance of alcohol in rape and sexual assault,
including further exploration of the extent to which men target unknown women
who are drinking and the strategies they use to make initial contact.
●In light of the Sex Offences Act 2003, development of a public education
campaign on the meaning of consent and the realities of rape, alongside more
detailed training for CJS personnel, explicitly designed to expand the concept of
‘real rape’.
xiii

A gap or a chasm? Attrition in reported rape cases
xiv

1. Introduction
The most recent Home Office figures on reported rape cases show an ongoing decline in
the conviction rate for England and Wales, putting it at an all time low of 5.6 per cent (see
Appendix 1). The government has expressed its concern at the year-on-year increase in
attrition of reported rape cases, and pledged to address this ‘justice gap’ not just in relation
to rape, but to all reported crime (Home Office, 2002a). There are, however, a number of
ways in which rape presents a set of unique challenges. Rape is not just a physical assault
but also a violation of personal, intimate and psychological boundaries. Most commonly the
offender is known, which also involves a betrayal of trust. It carries particular meanings and
potential consequences, which have specific resonances among many of the diverse
communities within the UK. More than with any other crime, complainants experience
reporting of the rape and the legal process as a form of re-victimisation (Gregory and Lees,
1999; Jordan, 2001a; Temkin, 1997). The reasons for this are historical, with what have
been termed ‘rape myths’ serving to justify:
unique case treatment evidenced in corroboration requirements, consent and
resistance standards, and the admissibility of victim character evidence.
(Caringella-MacDonald, 1985, p66)
Although in most jurisdictions corroboration requirements have been removed from statute, it
remains the case that in no other crime is the credibility of the victim/witness so subject to
scrutiny (Archambault and Lindsay, 2001). Rape is also unique in that the most likely and
also successful defence is to claim that the victim consented to the crime. Commenting on the
legacies of rape laws, two US scholars note:
This concern with protecting men from false accusations of rape went beyond the ‘not
guilty until proven innocent’ standard, and led to arguments of nearly unlimited
admissibility of evidence regarding [the complainant’s] character. This combined with
cultural conceptions of rape and early rape laws, placed serious impediments on the
adjudication of rape cases. (Bachman and Paternoster, 1993, p558)
Unlike other crimes, where the status of victim usually confers a sense of deserving sympathy
and support, declaring that one has been raped frequently invites judgement, and exacts
social and material costs.1
1
1. These may include loss of earnings and even giving up a job, and, where the rape took place in the woman's
home, a need to move house.

The ‘cultural conceptions’ of rape re f e rred to here are often described as ‘myths’, since they
do not reflect the reality of rape. We prefer to discuss powerful stereotypes that function to
limit the definition of what counts as ‘real rape’, in terms of the contexts and re l a t i o n s h i p s
within which sex without consent takes place. As a number of re s e a rchers and legal scholars
have pointed out (see, for example, Estrich, 1987; Myhill and Allen, 2002; Du Mont. et al,
2000), despite extensive legal re f o rm, ‘real rapes’ continue to be understood as those
committed by strangers, involving weapons and documented injury. The failure of criminal
justice systems to address these stereotypes means that the processes involved in re s p o n d i n g
to re p o rted rapes – from early investigation through to court room advocacy – can serve to
re i n f o rce, rather than challenge, narrow understandings of the crime of rape, who it happens
to and who perpetrates it. The attrition process itself reflects, and re p roduces, these patterns.
Attrition presents problems other than low conviction rates. It both originates in and
leads to the perpetuation of myths and differential treatment of rape cases as
compared to all other crimes. (Caringella-MacDonald, 1985, p66)
The comment made in March 2002 by Crown Court Recorder Jonathan Davies (D a i l y
Telegraph, 12th March), that women2are being encouraged to make complaints of ‘date
rape’ that have little chance of conviction, poses the question of whether we face not so
much a ‘justice gap’3but a chasm between the experiences and expectations of
complainants and how the justice system actually responds.
Structure of report
In the next chapter an outline is given of the aims of the research, the methodology and
data upon which this re p o rt is based. Chapter 3 then presents a brief overview of the
c u rrent knowledge-base on rape and attrition, with re f e rence to UK and intern a t i o n a l
research. In this chapter, the authors also present new data exploring trends and patterns
over time, drawing on the St Mary’s Sexual Assault Referral Centre database. Chapter 4
examines the research findings on rape and attrition, the authors present selected findings
that offer the most purchase on expanding understanding of the processes involved. Chapter
5 concludes with reflections on the challenges and dilemmas arising from the findings and
makes a short series of recommendations.
A gap or a chasm? Attrition in reported rape cases
2
2. Throughout this report the authors refer to complainants as women, since the vast majority of those reporting
rape are female. Although, where relevant they report any different patterns for male victims/survivors.
3. Defined specifically as ‘the difference between the number of crimes which are recorded and the number which
result in their perpetrator being brought to justice’ (Home Office, 2002a).

2. Aims and methodology
This research report draws on the evaluation of two projects funded by the Crime Reduction
Programme (CRP) Violence Against Women Initiative: Understanding Attrition, Decreasing
Early Withdrawals and Developing Best Practice in Response to Reported Rape: St Mary’s
Centre; and the linked national evaluation of Sexual Assault Referral Centres – SARCs (see
Lovett et al., 2004). It combines crime analysis of the St Mary’s database with prospective
case-tracking across six sites over a period of 17 to 27 months.4The aim is to increase our
understanding of attrition in reported rape cases, with an emphasis on early withdrawal by
complainants, since withdrawal at later stages have been addressed in recent re s e a rc h
(Harris and Grace, 1999).
SARCs were established as a way of combining the needs of victims/survivors in the
aftermath of rape and criminal justice requirements with respect to the collection of forensic
evidence. All three SARCs involved in this study were the ‘pioneer’ groups established in the
late 1980s and early 1990s – St Mary’s in Manchester, REACH in Northumbria and STAR
in West Yorkshire – and all three also accept self referrals as well as police referrals. The
first two are centre-based and operate according to what the authors have called an
‘integrated’ model, offering: forensic medical examinations; screening for sexually
transmitted infections and HIV counselling; prescription of post-coital contraception and
pregnancy testing; telephone information and support; and one-to-one counselling. STAR is
not centre-based, it offers support and counselling across the West Yorkshire region through
a large group of volunteers and sessional counsellors. Here forensic examinations are ‘out-
sourced’ to a third party, which co-ordinates a group of forensic doctors and uses a number
of rape examination suites (for more details see Lovett et al., 2004). The combination of self
and police referrals to SARCs offers an insight into the reasons for non-reporting, as well as
a non-police route for undertaking research on the process of reporting, investigation and
prosecution of rape cases.
T h ree Comparison areas where no SARC exists were also re c ruited in the south east of
England, where data collection took place through the police (see Appendix 2 for a
comparative profile of all six areas participating in this study). These sites were selected to
c reate a sample that would include a combination of metropolitan, inner city and ru r a l
3
4. Negotiating on protocols for data collection at the six sites took varying times to finalise, meaning that data
gathering began at diff e rent points in time: St Mary ’s from 1st October 2000 to 31st December 2002 (27
months); STAR and REACH from 1st January 2001 to 31st December 2002 (24 months); the Comparison areas
from 1st August 2001 to 31st December 2002 (17 months).

a reas, thus permitting broad comparability with the three SARC sites, although it is
impossible to make direct comparisons since no two areas in the UK are identical. In
addition, as the three SARCs serve large populations, it would have been difficult to
replicate this size for the comparisons without having to engage and negotiate with multiple
police force are a s .5As the authors were dependent on the co-operation of senior police
officers, force areas where they had existing contacts were approached.
I n i t i a l l y, two areas within the Metropolitan Police force area, Brent and Newham, were
engaged. Both are located in densely populated, urban areas with a mixed ethnic profile
and have dedicated teams of officers (Sexual Offences Investigative Techniques Off i c e r s )
trained in dealing with victims of sexual offences based in local police stations. To ensure
greater comparability with the SARC areas a third site, Thames Valley, was also included.
The area covers a wide range of urban and rural locations, with a less mixed ethnic profile
than the two London sites, and also has specialist officers (chaperones). There are very
limited specialist services in all three areas. In Brent and Newham forensic services are
performed by a contractor and in Thames Valley by a locally co-ordinated pool of forensic
examiners. Forensic examinations in all three areas are conducted in rape examination
suites in a range of locations. While officers provide information on Victim Support and
counselling services, there is no co-ordinated provision of support for complainants or a
d i rect re f e rral system at any of the three Comparison sites. More o v e r, re f e rral to Vi c t i m
Support is not automatic but only takes place where police officers ask, complainants say
they wish to receive contact, and the referral processes are followed through.6Police officers
have no record of whether any individual complainant accesses any form of support from
any external agency, including Victim Support.
Data collection
In order to gain both basic and in-depth information, from the perspectives of individual
service users as well as the range of agencies involved, seven strands of data collection
were undertaken.
●Analysis of the historic data in the St Mary’s Centre database.
A gap or a chasm? Attrition in reported rape cases
4
5. Two of the Comparison sites are in the area covered by ‘Project Sapphire’, the Metropolitan Police’s strategic
response to reported rape. However, this was launched in January 2001, and only began serious case-tracking
at the end of 2003, a while after the evaluation had been conceived and the methodology designed.
6. This is in line with current Home Office recommended practice (see Home Office Circular 44/2001,
www.homeoffice.gov.uk/docs/hoc44.html).

●P rospective case-tracking7using a specially designed database which re c o rd s
demographic data, basic details of the case, reporting of the incident, whether a
f o rensic medical was perf o rmed, pro g ress of the legal case and take-up of
services.
●Pro formas sent to investigating police officers.8
●Q u e s t i o n n a i res to service users in the first, fifth and twelfth months after initial
c o n t a c t / re p o r t. Options of completing the questionnaire by hand, doing a
telephone interview and/or doing a face-to-face interview were offered.
●I n t e r views with SARC staff, police officers and other key informants (Cro w n
Prosecution Service (CPS), Victim/Witness Support, Forensic Examiners).9
●An audit of a sample of forensic medical reports.
●Content analysis of a sample of victim/witness statements.
The total data corpus is presented in Table 2.1 and a full description of the research tools
and data sources can be found in Appendix 3.
St Mary’s has the most extensive database on recent rapes in the UK, comprising almost
6,500 cases over 13 years prior to the start of the research, and 7,931 by the end of the
project. This rich resource had never been analysed, and combined with the authors’ case-
tracking database it forms the basis of analysis which explores whether there are changing
p a t t e rns with respect to: the profile of victims and perpetrators; the relationship between
victims and offenders; the contexts in which the assaults occur; decision-making regarding
reporting; repeat victimisation and offending; and possible correlations with variables such
as gender, age, employment status and ethnicity. Several of these have been proposed as
factors accounting for the decline in convictions, making the analysis directly relevant to the
study of attrition.
The prospective case-tracking across all six sites means that through triangulation with
service user questionnaires and interviews, and police pro formas, it has been possible to
explore the decision-making processes of complainants and the factors associated with early
withdrawals. All data relating to service users were linked using unique reference numbers,
enabling cross-referencing across the entire range of data sources.
5
Aims and methodology
7. The prospective element of the study involved tracking pro g ress through the Criminal Justice System of all
reported rape and serious sexual assault cases in the six areas. Most previous UK research on reported rape
has used a retrospective design, drawing samples from police and/or CPS case files.
8. This was used in all sites except STAR, where a dedicated case-tracker compiles case outcomes on the project
database. Anonymised data from this resource were uploaded into the case-tracking database at six-monthly
intervals.
9. Where these are quoted in the text the date of the interview is also included.

A gap or a chasm? Attrition in reported rape cases
6

Data analysis
Analysis of the prospective case-tracking data consisted primarily of basic frequency counts
and cross-tabulations using the Access database package. Questionnaire data were entered
onto SPSS for Windows and, where appropriate, statistical significance tests (Chi-square)
were conducted. In addition, all responses to open-ended questions were typed into text files
related to each question. These were then grouped thematically and coded for frequency of
responses. Interview data were analysed using a similar process of ‘consolidation’ –
collecting all responses to particular questions in a single text file, undertaking coding and
content analysis to reveal areas of similarity and difference. Where appropriate, the authors
re p o rt on numbers and percentages of interview respondents who expressed part i c u l a r
views and opinions.
Not only is this study based on a much more extensive body of data than previous attrition
studies (see Table 3.44), cases have been tracked prospectively, rather than retrospectively.
In addition, the unique case numbers mean that whilst the entire data-set is anonymised, all
data sources, apart from the interviews with professionals, can be examined by case. The
importance of this was graphically illustrated when analysis of the police pro formas was
u n d e rtaken to examine the reasons for cases dropping out at the early stages. The
explanations provided by police officers created a sense of many deeply problematic cases;
it was only when these were compared with the perceptions of the complainants in these
same cases that a possible alternative narrative emerged.
Defining terms
Attrition is the process by which rape cases drop out of the legal process, thus do not result
in a criminal conviction. Some would argue that the first, and in many ways most
significant, point of attrition is the failure to report the crime (see more detailed discussion in
Chapter 3). As the three SARCs involved in this study support a proportion of service users
who have not reported to the police, the authors have been able to present some findings
on unreported cases (see Chapter 4, Attrition point 1). However, since both public concern
and government pronouncements about the ‘justice gap’ have been articulated in terms of
the falling pro p o rtion of convictions in relation to increasing numbers of re c o rded rape
cases, subsequent findings (see Chapter 4, Attrition points 2-7) and overall analysis of the
attrition process (see Tables 4.1 and 4.2) are based only on reported cases.
7
Aims and methodology

In previous studies of attrition, two alternate conventions have been employed in the
p resentation of conviction rates (cases resulting in a finding or plea of guilt): as a pro p o rtion of
all prosecuted cases or the pro p o rtion of all re p o rted cases. In this study the conviction rate is
defined according to the latter convention, although a figure for the former is also pre s e n t e d .
In terms of the sample, when referring to service user participants the term ‘respondents’ is
used for those completing questionnaires and ‘interviewees’ for the sub-sample who also
undertook in-depth interviews.
Limitations
Table 2.1 summarises the data corpus, comparing the actual data collected with that
projected in the bid to the CRP. In some areas the anticipated numbers were exceeded, such
as: the number of cases in the case-tracking database; and the interviews with staff and key
i n f o rmants. In other cases the targets were not reached, especially with respect to the
involvement of service users through questionnaires and interviews. Below the key issues
that affected the different areas of data collection are outlined.
At the design stage of this project the intention was to use the St Mary’s historic database to
explore changes over time with respect to the declining conviction rate. The database is a
unique resource in the UK containing cases since the project was founded in 1986 (when
the conviction rate was 19%). However, it emerged that not only was the database unstable
and using ‘ancient’ software (for which there was no longer any technical support), but data
entry was inconsistent and especially poor for the early years. A number of these problems
w e re, eventually, overcome: a new database was designed, data imported, and data
checked and inputted for the most problematic years.10 Whilst the historic database does
offer material not previously available, it has limited data on attrition, since case outcomes
have not been routinely recorded. The authors’ experience with the case-tracking database
exposed how time-consuming and frustrating it can be chasing outcome data from a large
number of police divisions. It also highlighted a clear gap in local inter-agency responses to
rape and sexual assault.1 1 Despite these limitations, the database offers insights into the
broader contexts of rape based on a substantial data set.
A gap or a chasm? Attrition in reported rape cases
8
10. This process took years rather than months after the re s e a rchers identified the problems early in the re s e a rch. An
o ffer to provide the already designed prospective case-tracking database was refused by the Health Trust, and a
lengthy tendering process with pre f e rred suppliers undertaken. Installation and transfer of data took months, rather
than the few weeks anticipated, and data from some key fields was not initially transferred. As a consequence,
the re s e a rchers, only had access to the historic data a few weeks before the re s e a rch funding ended.
11. S TAR is the only SARC to have a dedicated case-tracker, whose responsibility is to ensure case pro g ress and
c u rrent status is known and re c o rded. Even here, however, for the data covering this project we had to classify
four per cent (n=32) of STAR cases had to be classified as ‘unknown’ in terms of the authors’ classification system.

Although a higher than anticipated number of cases was re c o rded on the case-tracking
database, considerable time had to be devoted to updating case outcomes (using pro formas to
the police at two points in time, within a couple of months of the initial re p o rt and appro x i m a t e l y
a year afterw a rds, in all areas except STAR, where the data were extracted from their case-
tracking database), and getting re t u rns from the Comparison areas. Whilst outcomes have been
e n t e red for over two-thirds of the whole sample, it proved impossible to access information on a
significant pro p o rtion, despite regular requests for missing data. Where data from the police
p ro formas were used these covered cases from St Mary ’ s, REACH and the Comparison are a s .
Whilst there are classification and outcome data from STAR, these were downloaded from their
database and consist of information re c o rded by the STAR case-tracker rather than by police
o fficers. Outcome data should be taken as indicative rather than absolute for two reasons, firstly,
not all the pro formas were re t u rned (2,643 cases were re p o rted to the police but at the end of
the evaluation both police pro formas were missing for 232). Secondly, the amount of detail in
the section requesting reasons for decisions varied considerably, both between individual
o fficers and in terms of who had taken the decision (where it was a police decision more
i n f o rmation was provided than when it was the complainant’s). This is obviously a limitation of
the study, highlighting the difficulties of prospective re s e a rch across multiple sites and where the
actions of multiple parties are under scru t i n y.
Since the primary focus was on the earliest stages of the process, which has been less
re s e a rched than the CPS and trial stages (see, for example, Harris and Grace, 1999;
HMCPSI, 2002; Lees, 2002), data collection from the police was prioritised. Similar
processes were not developed with the CPS, and when towards the end of the project the
authors attempted to address the large number of unknowns in that category, it was
discovered that the official police numeric identifiers the authors had been using to track
cases were not those used by the CPS. The only way these outcomes could have been
obtained would have involved compromising the confidentiality agreement made with the
SARCs at the outset. Therefore, the data for the final stages were not as detailed, although
the ultimate outcomes, where known, are accurate.
The main reason for the lower than projected level of service user/complainant participation is
the very low take-up in the Comparison areas, where all layers of data collection proved more
p roblematic for two reasons: the police areas had far less investment in the re s e a rch, since they
w e re not part of CRP-funded projects; and repeated changes in staff, for example the key
contact with the research team frequently changed, and the new person usually received no
hand-over information.. Diff e rent timings of first contact were experimented, in respect to the
re s e a rch across the three SARCs. One clear lesson was that delaying this until at least a month
after attending the SARC (as occurred at REACH and STAR) prompted higher take-up (see
Appendix 4 for response rates at all sites). 9
Aims and methodology

Other re s e a rchers have noted the difficulty of engaging women in re s e a rch who have
recently been raped (see, for example, Harris and Grace, 1999; Jordan, 2001a), citing both
a reluctance to discuss the painful issues and difficulties in tracing individuals. Both of these
factors were evident in this project, alongside a level of ‘gate-keeping’, with those who
re s e a rchers rely on for access seeking to ‘protect’ victims/survivors from what they fear may
be perceived as unnecessary intrusion. As part of their re s e a r ch practice with
v i c t i m s / s u rvivors, the authors always included questions on the meaning and impact of
p a rticipation. In every project, and this is with no exception, hardly anyone re p o rt e d
negative impacts and many have noted ways in which they found participation useful. Many
have commented that they hope their involvement might lead to an improvement in re s p o n s e s
by the criminal justice system to other victims (for more details see Lovett et al., 2004).
Despite not achieving the original targets, the participating sample is larger than in any pre v i o u s
UK study of attrition (see Table 3.4 below), and compares favourably to a number of
i n t e rnational studies (see Chapter 3). It also broadly reflects the overall socio-demographic
p rofile of the 3,527 service users and complainants from the SARC and Comparison are a s
re g i s t e red in the case-tracking database.1 2 A total of 228 first questionnaires were re t u rned (66
f rom St Mary ’s, 51 from REACH, 91 from STAR and 20 from the Comparison sites). The 228
respondents were mainly female, white and were aged under 35. However, there were 14
male participants, 12 from Black or other minority ethnic backgrounds and the average age
was slightly older than the overall case-tracking sample at 30 years. Over half (53%, n=121) of
the assaults had been committed by men known to the victims, with just over one-third by re c e n t
acquaintances (13%, n=30) or strangers (24%, n=55). The nature of the relationship with the
perpetrator was unknown in the remaining ten per cent (n=22) of cases. While the majority
w e re police re f e rrals to the service, over one-third were self-re f e rrals (36%, n=81). All those in
the Comparison areas re p o rted directly to the police. In terms of attrition, re s e a rch part i c i p a n t s
( q u e s t i o n n a i re respondents and interviewees) were more likely than the case-tracking sample as
a whole to have cases detected and proceeded with (26%, n=49), and for the case to re a c h
c o u rt (23%, n=38).1 3 Since this was a prospective study, these outcomes could not have been
anticipated at the early point when participants opted into the re s e a r ch. Those who chose to be
i n t e rviewed were all female (n=56), except two, and were noticeably older, with an average
age of 32 years; they were even more likely to have their case proceed to court (37%, n=16 of
43 who re p o rted) and had a higher conviction rate than in the case-tracking sample (21%,
n=9). Revealingly, though, most interviewees, including those where the legal outcome was
positive, expressed dissatisfaction with elements within the criminal justice pro c e s s .
A gap or a chasm? Attrition in reported rape cases
10
12. This is based on analysis using the case-tracking database of all service users in the SARCs areas and
complainants in the Comparison areas (n=3,527), compared with the sub-group of those who participated in
the research (n=228). For a detailed profile of the overall case-tracking sample see Lovett et al., 2004.
13. In the case-tracking sample 20 per cent (n=527) of re p o rted cases proceeded, while 14 per cent (n=322)
reached the court stage.

It could be argued that by drawing the sample from SARCs the majority in this study have
better experiences of re p o rting rape, and there f o re the more negative experiences of
complainants elsewhere are underestimated. Whilst it is to be hoped that complainants
would have better experiences in areas with a SARC than those areas without one, it is
overly simplistic to presume this for every case for a number of reasons: there are
d i ff e rences in the services off e red by the three SARCs, and there is no control over the
location or conduct of forensic examinations (see Lovett et al., 2004); service user data
revealed variability of experience both within and between SARCs; police practice may or
may not be affected by the presence of a SARC, with police and CPS’ decision-making
more likely to be the outcome of national policies and widely held perspectives than local
variations in service provision for victims.
11
Aims and methodology

A gap or a chasm? Attrition in reported rape cases
12

3. What we know about rape and attrition
Understanding attrition re q u i res not only close examination of previous re s e a rch on this topic,
but also a basic grounding in the current knowledge on rape – how common it is, the
contexts in which it occurs and the pro p o rtion of cases that are re p o rted. In this chapter
recent re s e a rch is presented on the incidence and prevalence of rape, as well as findings
f rom the St Mary ’s historic database identifying patterns and trends over time with respect to
the profile of re p o rted rapes. Key UK and international re s e a rch on attrition is also pre s e n t e d .
How common is rape?
M e a s u rements of the prevalence of rape tend to be of two kinds – those assessing its
occurrence over lifetime, and those examining its extent within more limited time periods,
such as a year or before the age of 16. The number of reported rapes is lower than both
incidence and prevalence rates, since rape remains one of the most under-reported crimes
(Walby and Allen, 2004). Official figures for reported assaults often include children, as in
many jurisdictions, including England and Wales, there is no separate crime of child rape.14
Whilst there have been far fewer studies of the prevalence of sexual assault than of domestic
violence (Hagemann-White, 2001; Kelly and Regan, 2001), access to relatively accurate
o fficial statistics for re p o rted rapes has been more straightforw a rd. Researching unre p o rt e d
rape and sexual assault has proved more complex than many other forms of violence against
women, since using the word ‘rape’ in questions greatly decreases the re p o rting of forc e d
sex/sex without consent (Schwartz, 1997). The redesign of questions, for example, in the US
National Crime Victimisation Study in 1992 resulted in findings four times higher than pre v i o u s
versions (Greenfield, 1997). As with all prevalence re s e a rch, inconsistent findings are
attributed to methodological diff e rences with respect to: the sample; the number and content of
questions asked; the format (questionnaire, telephone or face-to-face interview); and the
definition of rape/sexual assault used by the re s e a rchers (see Schwartz, 1997, for a more
detailed discussion). At the same time, comparative data show such significant variations
(Australia, the USA and Sweden re c o rding high prevalence rates per head of population, and
South Africa had the highest rates for both prevalence and re p o rting) that the possibility of
d i ff e rential national levels of sexual violence within diff e rent societies needs to be explore d .
13
14. T h e re are a range of specific sexual offences against children in the Sexual Offences Act, which was
implemented in May 2004. It remains to be seen, however, whether the CPS continue to use the charge of rape
for cases involving under-16s.

Incidence and prevalence in the UK
T h e re has been no dedicated national random sample study of either the incidence or
p revalence of rape in the UK, as distinct from wider surveys such as the national British
Crime Survey (BCS), which focus on the broader issue of ‘crime’ but include a section on
sexual assault. In fact, there has only been a single study designed solely to pro v i d e
information on the extent of unreported rape (Painter, 1991). This survey involved 1,007
women in 11 cities and was primarily an attempt to quantify the extent of marital rape. The
key findings include:
●one in four women had experienced rape or attempted rape in their lifetime;
●the most common perpetrators were current and ex-partners; and
●the vast majority (91%) told no one at the time.
The BCS is a large national victim-focused surv e y, which is used by the Home Office to
p rovide estimates of the prevalence and incidence of crime in England and Wales. The
1998 and 2000 BCS both included a computerised self-completion module, which was
designed to provide an accurate estimate of the extent and nature of sexual victimisation
(Myhill and Allen, 2002). In 2001 the BCS included another computerised self-completion
module, which this time captured data on interpersonal violence, which was defined as
domestic violence, sexual violence and stalking (Walby and Allen, 2004).
Rather unexpectedly, since the ability of the BCS to detect sexual violence has long been
subject to criticism and internal reflection at the Home Office (Myhill and Allen, 2002), the
findings from the 2001 specially designed interpersonal violence module are lower for
sexual violence than in the earlier overview.15 The last-year prevalence rates calculated for
‘sexual victimisation’ – rape and sexual assault of women over 16 – are: 0.9 per cent
(2002) and 0.5 per cent (2001) for sexual assault and 0.4 per cent (2002) and 0.3 per
cent (2001) for rape. The much higher figure for sexual assault in the earlier data (Myhill
and Allen, 2002) might be accounted for by the use of a narrower definition in the later
study, linked to the revisions in the Sexual Offences Act (Walby and Allen, 2004).16 Myhill
and Allen (2002) extrapolate that their figures would produce an annual incidence for rape
of 61,000 in the year before the surv e y, whereas the figures for the more recent study
means the annual incidence rate falls to 47,000 (Walby and Allen, 2004).
A gap or a chasm? Attrition in reported rape cases
14
15. The authors suggest a range of methodological changes that might account for the diff e rences (Walby and
Allen, 2004, p116), and sexual violence still tends to be under-reported in surveys that specify they are about
‘crime’ (Schwartz, 1997).
16. Only incidents that would be defined in law as rape and the new of ‘offence of sexual assault by penetration’
are included.

The earlier study (Myhill and Allen, 2002) recorded a 20 per cent reporting rate, with the
more recent data showing a lower rate of 15 per cent (Walby and Allen, 2004) among
those disclosing rape. Note, the focus in the former was whether the ‘last’ incident was
reported, whereas in the latter it refers to the ‘worst’ incident.
The BCS estimates cannot be simply mapped onto Home Office recorded crime statistics,
firstly because not every crime reported is recorded as a crime at all, or as the same crime
the victim perceives to have taken place. Secondly, to make the data comparable,
disaggregation in terms of gender and age would need to be undertaken.
The prevalence estimates from both studies are considerably lower than in Painter’s study
(1991): one in ten (9.7%) and one in six (16.6%) women respectively had been sexually
assaulted; and one in 20 (4.9%) and one in 27 (3.7%) respectively had suffered at least
one incident of rape since they were 16 (Myhill and Allen, 2002; Walby and Allen, 2004).
Whilst there is considerable disparity here between the actual numbers1 7, there is
c o n v e rgence about the contexts in which rape takes place: women are most likely to be
raped by men they know (intimates, 54%, other known individuals, 29%); and a
considerable proportion involve repeat assaults by the same perpetrator (50% in the last 12
months) (Walby and Allen, 2004). The Myhill and Allen (2002) re p o rt also re f l e c t e d
findings from other jurisdictions (Bergen, 1995; Easteal, 1998) that rapes by current and ex-
partners were the most likely to result in injuries. This profile also accounted for the most
common locations, which were the victim’s home (55%) followed by offender’s home (20%),
public place (13%) and elsewhere (13%).18
Both sets of BCS data confirm an increasing willingness amongst women to tell someone,
but even for those raped in the last five years less than half did so. They also echo findings
from research on rape, domestic violence and child sexual abuse, that the most likely person
to be told is a friend or family member (Kelly, 2000), and that only a small minority access
specialist support organisations like SARCs and rape crisis centres (Walby and Allen, 2004,
p94; Myhill and Allen, 2002, p48).
15
17. Considerable discussion has taken place about the subtle and not so subtle methodological differences which
result in lower prevalence findings for sexual violence in general crime surveys compared to dedicated studies of
either violence against women or rape (see for example, Schwartz, 1997; Hagemann-White, 2001).
18. Here, the authors rely more on Myhill and Allen (2002), which contains extensive contextual data since it is only
concerned with rape and sexual assault.
What we know about rape and attrition

Incidence and prevalence research internationally
International prevalence research sheds further light on some of these issues since several
community-based studies have, through a focus only on violence against women (rather than
‘crime’), enabled more disclosure. The most frequently cited re s e a rch was conducted by
Statistics Canada in 1992, which involved a national random sample of 12,300 women
and used telephone interviews. Whilst most publicity has been given to the domestic
violence findings, considerable data were collected on rape and sexual assault (Johnson
and Sacco, 1995). The BCS findings on known men and repeat assaults were echoed, but
a far higher prevalence rate was found (over one in three reported a sexual assault) and a
lower reporting rate to the police (6% compared to 25% for domestic violence in this study).
The Australian Wo m e n ’s Safety Survey conducted by the Bureau of Statistics in 1996
(Easteal, 1998) involved a random sample 6,300 women aged 18 and over. It produced
incidence finding of 1.9 per cent for sexual assault in the previous 12 months. Known men
accounted for over two-thirds of assailants (68%), and curre n t / e x - p a rtners and dates
comprised more than half of this group (an even higher proportion was found in the most
recent US study, Tjaden and Thoennes, 1998). Over half of the assaulted women in the
sample (59%) had told a friend, and 15 per cent reported to the police.
In her review of studies of ‘date’ rape Koss (2000) notes similar and higher findings in
smaller samples of college students and naval re c ruits (see, for example, Fisher et al. ,
2000). She concludes that the trenchant criticisms of the mid-1990s – that samples were
unrepresentative, female college students were uniquely vulnerable to rape, and prevalence
findings were wildly inconsistent – have all been ‘outgrown’ through the accumulation and
convergence of findings. Archambault and Lindsay (2001) also note the misconception that
non-stranger sexual assault is ‘date rape’. Their analysis of reported crimes in San Diego
(n=894) revealed only 19 per cent took place in the context of dating relationships.
Summary
●There has been no dedicated national random sample UK study of the incidence
or prevalence of rape. However, the national victim-focused BCS has included a
self-completion module to capture data on rape and sexual assault in the 1998,
2000 and 2001 surveys, within the broader issue of ‘crime’.
●A 1991 UK survey found that one in four women had experienced
rape/attempted rape in their lifetime, with current and ex-partners the most
common perpetrators, and the vast majority not disclosing the crime at the time.
A gap or a chasm? Attrition in reported rape cases
16

●Findings from the BCS 2001 revealed the prevalence rate for rape of women
over 16 was 0.3 per cent for the year prior to interv i e w, and an estimated
47,000 adult female victims of rape. The prevalence rate since the age of 16 is
one in 27 women suffering at least one incident of rape.
●The BCS confirms that women are most likely to be raped by men they know, and
that a considerable pro p o rtion had experienced repeat incidents by the same
perpetrator.
●There are international examples of prevalence studies, which through a specific
focus on violence against women have enabled greater disclosure.
New data on patterns and trends
Whilst the limitations of the St Mary’s historic database have been noted in Chapter 2, it
nonetheless represents a unique resource for tracing trends over time and the contexts in
which rapes reported to the police take place. Although the authors cannot present attrition
figures per se, they have been able to test some of the explanations floated in UK research
for the continuing decline in the proportion of convictions.
A large number of calculations and analyses have been undertaken, including cro s s -
tabulations, not all of which are reported here. The key areas investigated were patterns
and changes over time with respect to: overall service use; relationship between assailant
and victim; location of assault; timings of assaults, including days and months; use of
weapons and injuries; and the relevance of demographic factors including gender, age,
employment status and ethnicity. Multiple assailant cases were examined in detail, as were
individuals who reported multiple assaults by different perpetrators. Several re-codings were
developed to make some of these calculations manageable. The most relevant here is a
three-way coding for relationship: well known (current/ex-partners, family members), known
(friends and acquaintances) and unknown (total strangers, those known for less than 24
hours19).
Numbers and age profile of service users
Table 3.1 shows that the number of cases seen at St Mary’s has more than doubled since
1987, with the majority of this increase coming from police referrals.
17
What we know about rape and attrition
19. Archambault and Lindsay (2001) use a similar category of ‘brief encounter’ analysing San Diego’s police data,
noting that from the perspective of victims they do not ‘know’ these individuals.

Table 3.1: Referrals to St Mary’s 1987-2002
Attendance Year Number of % change Cumulative total 5 year
referrals from prior year of referrals % change
1987 301
1988 346 15%
1989 408 18%
1990 399 -2%
1991 400 0%
1992 392 -2% 1,945
1993 463 18%
1994 457 -1%
1995 473 4%
1996 541 14%
1997 593 10% 2,527 30%
1998 556 -6%
1999 621 12%
2000 631 2%
2001 695 10%
2002 642 -8% 3,145 24%
Source: St Mary’s historic database
Interestingly self-referrals are only slightly higher in 2002 than 1987, and have thus become
a decreasing proportion of cases (see Figure 3.1). This is also reflected in the increase in
f o rensic examinations, both overall (from 186 in 1987 to 488 in 2002) and as a
percentage of all cases seen: 62 per cent (n=186 of 301) in 1987, 76 per cent (n=488 of
642) in 2002 (see Figure 3.2).
A gap or a chasm? Attrition in reported rape cases
18

Figure 3.1: Numbers of police and self-referrals 1987-2002
Source: St Mary’s historic database
Figure 3.2: Forensic examinations undertaken 1987-2002
Source: St Mary’s historic database
A large increase in reports by men is also evident in the St Mary’s historic data-set – 280
per cent between 1993-1997 and 87 per cent between 1998-2002 – although they still
only account for 8 per cent of all service users.
19
What we know about rape and attrition

The age profile of service users has altered slightly, with the proportion of those aged under
20 increasing from 25 per cent in 1988 to a high of 43 per cent of all service users in
2002 (see Figure 3.3). The number of over-45s has remained fairly constant at five per cent.
The increase in referrals of under-16s, was the basis of another of the CRP rape projects
(see Skinner and Taylor, 2005). A potential age profile has not been explored as a possible
element in declining rates of conviction.
Figure 3.3: Age profile of service users 1987-2002
Source: St Mary’s historic database
Relationship to offender
Analysis of the relationship between victim and offender for the full data-set (1987-2002)
reveals that the largest category of assailants are acquaintances (33%), followed by
strangers (28%), current and ex-partners (19%), known less than 24 hours (11%) and family
members (10%). Figure 3.4 displays the annual breakdown by relationship type. These
figures are closer to the profile found in the BCS data (see earlier section), with the overall
proportion of assaults by known men at 57 per cent. Nevertheless, it is clear from both the
BCS and this study that stranger assaults are still more likely to be re p o rted to the
p o l i c e / s u p p o rt services, and that assaults by intimates (current and ex-partners) re m a i n
under-reported.
A gap or a chasm? Attrition in reported rape cases
20

Figure 3.4: Relationship between victim and perpetrator 1987-2002
Source: St Mary’s historic database
Under 20-year-olds were more likely to report assaults by acquaintances, with assaults by
c u rre n t / e x - p a r tners most common within the 20- to 45-year-old age group. Looking at
changes over time (see Figure 3.4), whilst there are shifts they are not of the order or nature
that many commentators have suggested: the proportion of stranger assaults has only fallen
by six to seven per cent from 1987 to 2002; and the increasing categories are current and
former partners, and friends – neither of which can be termed ‘date rape’. These data are
different from those reported by Harris and Grace (1999), who argue that the fall in the
conviction rate throughout the 1990s can be accounted for solely in terms of the different
kinds of rape cases being re p o rted. Both the historic data from St Mary ’s and a re c e n t
reassessment of attrition for the same time period as the Harris and Grace study (Lea et al.,
2003) do not fully support this contention.
Location of assaults
The most common location in which rape took place was a public place (32%), followed by
the victim’s home (24%) and perpetrator’s home (19%), with vehicles re p resenting six per cent.
Again, this profile is diff e rent from that in BCS data (see earlier). Almost half (45%) of the
assaults re p o rted by women in prostitution took place in vehicles, followed by public places
(29%). There was minimal variation over time in the locations of assaults, although more subtle
distinctions within public place, which the authors could not track, might reveal some shifts.
21
What we know about rape and attrition

Violence and injuries
Whilst re s e a rch from other jurisdictions suggests that injuries are documented in only a
minority of re p o rted rape cases (Kelly, 2003a), they are documented in the majority (an
average of two-thirds) of those undergoing a forensic medical at St Mary’s. The database
also confirms that rapes by well-known perpetrators are slightly more likely to result in
injuries (70%). These data suggest there is supportive medical evidence in the majority of
cases where forensic medical examinations are conducted in conducive conditions (see
Regan et al., 2004).
Time frames
Aggregating cases from 1987-2002 revealed there are two consistent peaks in reporting,
during the summer (July and August) for stranger assaults and in the New Year (some of
these reports may also relate to the Christmas period) for assaults by known perpetrators
(see Figure 3.5).
Figure 3.5: Reporting of rapes by month of the year and relationship to offender
1987-2002
Source: St Mary’s historic database
In terms of days of the week, peaks in reporting are Friday to Sunday, although an average
of 12 per cent of first attendances at St. Mary’s SARC take place on each of the other four
weekdays.
A gap or a chasm? Attrition in reported rape cases
22

Employment status
In terms of the status of service users the largest category is unemployed (27%), followed by
students (23%) and those in full-time employment (22%). The high number of students is linked
p a rtly to the preponderance of under-20s in the sample overall, and partly to the number of
universities and colleges in Greater Manchester. Employment status was correlated with
re f e rral type, with a higher pro p o rtion of those in full-time work being self, rather than police,
re f e rrals. There were also patterns within relationship to offenders, with the full-time employed
most likely to be assaulted by unknown assailants (41%), homemakers by assailants known to
them (31%) and students by known and unknown assailants (37% and 38% re s p e c t i v e l y ) .
Ethnicity
The ethnic profile of attendees has become more diverse, especially in the most recent years,
w h e re BME women (there were no BME men in the sample) re p resent eight per cent of all
s e rvice users. At first glance it appears that these are most likely to be self-re f e rrals, however,
analysis by ethnic group reveals a large group of African asylum seekers seeking support for
sexual assaults that took place before they migrated, whilst Pakistani and Black Caribbean
women were more likely to be police re f e rrals than other groups, including white women.
Multiple assailants
Eleven per cent of cases involved more than one assailant (n=863), with the majority of
these cases involving two (60%) or three (23%), but a significant proportion (17%) four or
more – one case involved 14 and another 20 perpetrators. Whilst the number of multiple
assailant cases has increased year-on-year, they have not risen as a proportion of cases, in
fact they have fallen slightly over the 16-year period. Revealingly, this group is not just made
up of strangers either, at least one individual in a third of assaults involving two or three
assailants was known to the victim. Current, and, ex-partners, friends and acquaintances
featured in multiple assailant cases. In cases involving 10 or more perpetrators (n=7), all the
perpetrators were strangers.
Multiple attendees
One issue that has not been addressed in previous studies is the number of service users
who report multiple assaults over time. A total of 287 (9%) service users had attended more
than once about separate offences. The vast majority (77%, n=221) attended twice, with 44
attending three times, two attending six times and one attending eight times. Possible
correlations with either mental health problems or involvement with prostitution were tested,
23
What we know about rape and attrition

but did not account for these patterns. Whilst the database does not enable furt h e r
investigation, these data raise the possibility of a group who may be uniquely vulnerable to
repeat victimisation. Given the high rates of attrition, and that prior unproven allegations
are seen to detract from victim credibility (Jordan, 2004), this group may also, at least for
subsequent assaults, have even less access to justice.
Summary
●The number of cases seen at St Mary’s has doubled since 1987, with the majority
of this increase accounted for by police re f e rrals. The number of fore n s i c
examinations conducted has risen proportionately.
●Reported rapes even to a SARC still differ from the profile in prevalence studies.
●The age profile of service users has altered slightly, with under-20s forming a
significant and growing proportion of all cases.
●Fifty-seven per cent of assaults were committed by known men, but stranger assaults
remain more likely to be re p o rted to the police than those committed by intimates.
●The most common assault locations were public places, followed by victims’ and
perpetrators’ homes.
●In contrast to previous research findings high levels of documented injury were
found in the St Mary ’s sample, and this was more likely in cases where the
perpetrator was well known.
●The two peak reporting times are in the summer (July and August) and over the
New Year period.
●In terms of employment status, service users were most commonly unemployed,
followed by students and full-time employed.
●The ethnic profile of service users has become more diverse.
●Eleven per cent of assaults were perpetrated by more than one assailant.
●Nine per cent of service users re p o rted separate assaults by diff e rent perpetrators.
Research on attrition
It is rather surprising, given the extent of statute and procedural re f o rm in many jurisdictions
over the last two decades (Schulhofer, 1998; Temkin, 2003), that re s e a r ch on attrition in rape
cases is relatively rare. In fact, many European countries cannot even provide basic data on
the numbers of re p o rts, prosecutions and convictions (Kelly and Regan, 2001; Regan and
K e l l y, 2003). In this respect the UK has a stronger track re c o rd than many other countries,
with a number of published studies and yearly statistics compiled by the Home Off i c e .
A gap or a chasm? Attrition in reported rape cases
24

UK data and research
F i g u re 3.6, using official statistics from the Home Office, graphically illustrates the two
p a t t e rns in re p o rted rape cases in England and Wales over the past two decades: a
continuing and unbroken increase in re p o rt i n g2 0 ; and a relatively static number of
convictions. The combination of these two trends, which began in the 1970s (Regan and
Kelly, 2003), means that whilst in 1977 one in three reported rapes resulted in a conviction,
by 2002 this had fallen to one in 18 (32% versus 5.6%). The only change to the overall
pattern is a small increase in the percentage of prosecutions in 2001, which continues into
2002, but with no parallel increase in convictions. This may be an outcome of changes in
local and/or national policy in the police and CPS. However, it is too soon to say whether
this will be a sustained trend, which will also feed through into the conviction rate. In fact,
the attrition rate is even greater, since official statistics exclude reports that are ‘no crimed’
very early in the process (see Gregory and Lees, 1999; Harris and Grace, 1999), not to
mention that some convictions are overturned on appeal.
Figure 3.6: Attrition in rape cases England and Wales 1985-2002
Source: Regan and Kelly, 2003
25
What we know about rape and attrition
20. Changes in the National Crime Reporting Standards (NCRS) introduced in April 2002 may have contributed to
the increased figures for reported rape, since recording and classification of all crimes should now be based on
the perception of the victim that an offence occurred. Simmons and Dodd (2003, p5) suggest that changing this
counting rule may have increased recorded offences against the person by 23 per cent in police statistics. "The
number of recorded sexual offences was thought to be largely unchanged in 2002/03, after accounting for the
effects of the NCRS (reliable estimates for this effect are not available due to the relatively small numbers of
occurrences)" (Simmons and Dodd, 2003, p82). This will only have had a marginal effect on the data in this
report, however, since the majority cases were reported before this change.

Table 3.2 presents disaggregated data for prosecutions and convictions involving adults and
children (under 16-years-olds) for the years 1998-2002, covering all rape and attempted
rape cases. Conviction data include guilty pleas; convictions at trial are lower than this
overall rate. For adults, around one-quarter of prosecuted cases result in a conviction; the
figure was more than one-third for child cases, until a dramatic fall of 12 percentage points
f rom 2000 to 2001, although it re c o v e red very slightly in 2002. Cases considered the
‘weakest’ evidentially are not referred up to the CPS by the police. These data suggest that
even cases deemed relatively strong by police and supported by CPS fare poorly in the later
stages of the criminal justice process.
Table 3.2: Comparison of prosecution outcomes in adult and child rape cases 1998-2002
Rapes and attempted rapes Rapes and attempted rapes of
adult females and males of child (under-16) females and males
Year Prosecutions Convictions Prosecutions Prosecutions Convictions Prosecutions
resulting in resulting in
convictions convictions
1998 1366 382 28% 810 291 36%
1999 1319 317 24% 846 335 40%
2000 1245 329 26% 788 264 34%
2001 1528 326 21% 1112 241 22%
2002 1646 352 21% 1288 292 23%
Source: Home Office data
The 2001 and 2002 Home Office data are revealing in other ways (see Appendix 1). Here
re p o rting, prosecution and conviction data are presented by police force area; for the
purposes of this report, additional calculations of the conviction rate and rates of reporting
by population were undertaken. The conviction rate shows a considerable range between
a reas, varying from one to 14 per cent across the two years. A number of potential
confounding factors were explored, including whether there were higher reporting rates per
head of population, but were not linked to the variations in conviction rates in any obvious
way.21 Nor were higher or lower rates simply a function of either metropolitan versus rural
areas or the number of cases dealt with per year: areas with conviction rates above ten per
cent in 2002 (Cleveland, Cumbria, Gwent, South Wales and South Yo r k s h i re) and those
A gap or a chasm? Attrition in reported rape cases
26
21. There were, however, interesting variations in the number of notifiable offences per 100,000: taking the 2002
data, the majority of force areas had rates of between 15 and 25 (n=25), nine areas had rates of below 15
and eight rates of more than 25. The higher rates were not associated with Metropolitan areas, apart fro m
London that had the highest overall.

with rates below four per cent (Avon & Somerset, Essex, Gloucestershire, Leicestershire ,
Sussex, Wa rw i c k s h i re and West Yo r k s h i re) were not distinguishable on these grounds. In
addition, variations of plus or minus five per cent were evident for a number of are a s
between 2001 and 2002 – a considerable increase or fall given the current low average
rate of convictions. Additional re s e a rch, over longer timescales and drawing on more
detailed case data, is undoubtedly needed if variations both between and within police
force areas are to be understood and accounted for.
Table 3.3 compares five UK research studies (Chambers and Millar, 1983; Grace et al.,
1992; Lees and Gregory, 1993; Jamieson et al., 199822; Harris and Grace, 1999; Lea et
al., 2003) alongside data from St Mary’s Sexual Assault Referral Centre for 1996-199723.
Four of the studies comprise data from England and Wales, and two (Chambers and Millar,
1983; Jamieson et al., 1998) focus on Scotland. It has been less than simple constructing
comparative figures, since the points at which calculations were made vary in each of the
original studies2 4 and two studies included sexual offences other than rape or attempted
rape (Lees and Gregory, 1993; Jamieson et al., 1998).
All the studies demonstrate that the highest proportion of cases is lost at the earliest stages,
with between half and two-thirds dropping out before referral to prosecutors. The rate of ‘no
c r i m i n g ’2 5 remains high, despite repeated guidelines from the Home Office (see, for
example, Home Office, 1983) and internally within the police, specifying that it should be
confined to false allegations26, crimes that occurred in another area and/or were recorded
as a crime in error (Home Office, 2003). No study has found ‘no criming’ to be limited to
these categories. The most significant factors in early loss of cases are designation as false
reports and withdrawals by the victim/complainant.
27
What we know about rape and attrition
22. This unpublished study was a pilot exploring the feasibility of a large-scale tracking project across Scotland. The
sample comprised all sexual offences cases reported in two police areas during 1996-1997, of which a minority
w e re rape. The study concluded that to track all sex offences across the country would prove extre m e l y
expensive.
23. Presented at the Promoting the Model conference, Manchester, in 1997.
24. Several analyse from a base of 100 per cent at each stage of the legal process, meaning that significant re-
calculation has had to be done on the original data.
25. Home Office guidance advises that the police may no-crime a case where following the report of an incident
which has subsequently been re c o rded as a crime, additional verifiable information is available which
determines that no notifiable crime has been committed.
26. Defined as either where the complainant makes a clear retraction or where there is strong evidence that the
report was false.

Table 3.3: UK research findings on attrition in reported rape cases1
Chambers Grace Lees St Jamieson Harris Lea
and et al, and Mary’s et al, and et al,
Millar, 1992 Gregory, 1996-7 19982Grace, 2003
1983* 19932* 1999
Initial sample size 196 464 109 378 47 483 379
Police stage
Cases lost 46% – 57% 44% 36% 67% 61%
No-crimed 22% 45% 43% – – 25% 11%
Unsolved 24% 7% 7% – – 11% –
NFA – 10% 6% – – 31% –
Prosecution stage
Referred to CPS 54% – 43% 30% 64% 31% –
Discontinued 16% – 18% 7% 32% 8% 14%
Court stage
Prosecutions 38% 33% 25% 24% 32% 20% –
Rape conviction 15% 19% 9% 9% 15%3 6% 5%
Conviction other 10% 8% 2% – 4%3 7% 6%
Acquittal 9% 7% 14% – 4%3 7% –
Notes
1 Studies presented findings in diff e rent ways and full data sets were not accessible to perf o rm any re -
calculation. Figures are not available, therefore, for each stage of the attrition process, and some data are
missing.
2 Data limited to the rape and attempted rape cases in the sample.
For some cases the outcomes were still unknown.
* Research conducted in Scotland
These studies suggest higher levels of prosecution and conviction in Scotland. However,
when national data for rape complaints only are examined, a similar pattern to England
and Wales appears, with conviction rates falling year-on-year to six per cent in 2001 (see
Figure 3.8). One difference, however, is that a higher percentage (close to 50%) of cases
that are prosecuted result in convictions in Scotland.
A gap or a chasm? Attrition in reported rape cases
28

Figure 3.7: Attrition in rape cases in Scotland 1977-2001
Source: Regan and Kelly, 2003
Attrition internationally
In studies of attrition rates across Europe (Kelly and Regan, 2001; Regan and Kelly, 2003),
Finland, Ireland and Sweden displayed the same stark trends over time as the UK, in
relation to re p o rting and conviction rates. Whilst a number of other countries re c o rd e d
variations in reporting, a decline in the proportion of prosecutions and convictions was a
common trend for every country that provided data, apart from Germany where the
proportion of prosecutions and convictions has risen since 1997. Much higher prosecution
rates, for example more than 50 per cent of reports in Austria and Denmark, were evident,
alongside a number of countries where the majority of prosecutions resulted in convictions
(Finland, Germ a n y, Hungary, Iceland). Overall these data suggest that, whilst some core
problems link adversarial and investigative legal systems, a number of European countries
have higher prosecution and conviction rates than the UK. In fact, the only country with a
lower conviction rate was Ireland (Regan and Kelly, 2003).
Bachman and Paternoster (1993) note the absence of re s e a r ch in the US to address whether the
policy intent underlying rape law re f o rm has been reflected in increased re p o r ting and
p rosecution, especially of rape by known assailants. They conclude, drawing on a range of
national data sources, that there has been some impact, but it is not substantial, and that
c o m p a red with ro b b e ry and assault offences there is still a large discrepancy between re p o rt e d
o ffences and convictions, and a strong ‘acquaintance discount’, for rape offences. Slightly
29
What we know about rape and attrition

s t ronger impacts were found adjusting the methodology and looking at three states (Bachman
and Smith, 1994) but the higher attrition rate compared to other serious crimes persisted.
The Senate Judiciary Committee issued a report entitled Detours on the Road to Equal Justice
in 1993 that documented attrition in rape cases across the USA. The average conviction
rate in 1990 was 12 per cent. The report demonstrates disparities in how rape and other
violent crimes were prosecuted. The key Attrition points identified were: arrest (62% of
reported rapes do not result in an arrest); dismissal (of the cases that moved into the system,
48% were dismissed before trial); and acquittal at trial. The committee also noted reluctance
amongst prosecutors to bring cases where the parties were known to one another – the
majority of rape cases. However, the review found significantly higher conviction rates in
particular areas of the US, including Washington and New York. A subsequent review notes
that conviction rates across the USA range from 2.5 to 19.9 per cent (Sinclair and Bourne,
1998, p576) but offered little explanation of these wide discrepancies.
B re reton (1993) discusses four Australian studies where patterns similar to those
documented for England and Wales and the USA are re p o rted. The key players are the
victim, in terms of the decision to report and continuing with the case, and the police in
terms of the decision of whether to lay charges and their influence on the victim’s subsequent
decision-making. Several studies also note 20 to 35 per cent of cases being dropped by
prosecutors. Guilty pleas are much lower than in other criminal cases. Two factors had the
most impact on outcomes at trial: evidence of physical injury; and admissions by the
defendant at some point in the process.
Identifying Attrition points
Attrition research identifies a paradox internationally: despite widespread reform of statute
law and, in many jurisdictions, procedural rules, the 1990s witnessed declining or static
conviction rates. The UK has one of the most pronounced patterns. Remarkably little
re s e a rch or legal commentary has, as yet, attempted to explain these common – and
unexpected – international similarities.
R e s e a rch to date in adversarial legal systems has identified four key points at which attrition
occurs. The first point is the decision to re p o rt itself; estimates of the re p o rting rate range from 5
to 25 per cent. Even using the highest re p o rting rate estimate, thre e - q u a rters of cases never
reach the first hurdle within the CJS. The second involves the police investigation stage – the
initial response, forensic examination, statement taking, evidence gathering and arrest and/or
i n t e rviewing of suspects – between half and thre e - q u a r ters of re p o rted cases are lost here. The
A gap or a chasm? Attrition in reported rape cases
30

th i rd point relates to the minority of cases that are re f e rred through to prosecutors, where a
p ro p o rtion are discontinued. The final point is the even smaller number of cases that reach court ,
w h e re between one-third and over one-half of those involving adults result in acquittals. At each of
the points the possibility of withdrawal by the victim exists, although the largest number of these
occur during the re p o rting and investigative stages – what the authors term ‘early withdrawals’.
This is only part of the story, however, since attrition varies according the characteristics of
the case. The most recent studies in England, Wales and Scotland concur that cases involving
c h i l d ren are more likely to be prosecuted and to result in convictions. Adult rape cases have
higher attrition rates, especially if they depart from the ‘real rape’ template. Cases involving
complainants who have mental health problems or some form of learning disability are rare l y
p rosecuted (Harris and Grace, 1999; Lea et al., 2003). In thinking about Attrition points, it
may prove useful to distinguish between cases that are lost and those that are dropped. The
latter involves cases where either a police officer or prosecutor makes a clear and explicit
decision to not proceed with a case. Cases can be lost in a number of ways: through failure
to identify and/or find the assailant; the withdrawal of the victim; and acquittals in court. In
the analysis of attrition the authors have sought to highlight both processes and case
characteristics to illuminate the many factors at play. Since attempts to reduce attrition should
t a rget specific points in the process, it is hoped that a more detailed documentation of how
and why cases are lost or dropped will enhance eff o rts to address the justice gap.
The decision to report
Sexual violence, and rape in part i c u l a r, is considered the most under- re p o r ted crime
(American Medical Association, 1995). Whilst a number of studies have explored the
reasons for not re p o rting, the equally significant decision to re p o r t has received less
attention. In terms of understanding attrition, this is unfortunate, since knowing what women
are seeking when they approach the police may offer insights into their subsequent decision-
making. The findings below draw on international prevalence studies and a series of
‘phone-in’ surveys in Australia designed to explore responses of services to reported rape
(for more details see Kelly, 2002).
A wide range of reasons for not reporting have been documented, the major ones being:
●not naming the event as rape (and/or ‘a crime’) oneself27;
●not thinking the police/others will define the event as rape;
31
What we know about rape and attrition
27. In BCS data, only 43 per cent of women subject since the age of 16 to an act that met the 1994 legal definition
of rape actually thought of it as rape (Walby and Allen, 2004)

●fear of disbelief;
●fear of blame/judgement;
●distrust of the police/courts/legal process;
●fear of family and friends knowing/public disclosure28;
●fear of further attack/intimidation;
●divided loyalty in cases involving current/ex-intimates; and
●language/communication issues for women with disabilities and/or whose first
language is not that of the country where they were assaulted.
For many women it is a combination of factors that militate against re p o rting and some
feature more strongly for certain groups of women than others. A heightened mistrust of the
police has been frequently noted, for example, among women from minority ethnic
communities (Commission on Women and the Criminal Justice System, 2004).29 What has
received less attention are the heightened concerns about others knowing and fear that they
might be blamed (especially, if alcohol or drugs were involved) which dispro p o rt i o n a t e l y
affect young women. Many of the factors that discourage reporting are connected to the
notion of ‘real rape’: the phone-in survey in Victoria, Australia, for example, notes that
almost two-thirds of cases (60%) involving strangers were reported compared to less than
one-quarter (21%) of those involving known men (Brereton, 1993); and the BCS data also
found (Myhill and Allen, 2002), stranger assaults were more than twice as likely than any
other assaults involving any other perpetrator to be reported whereas assaults in the context
of ‘dates’ the least likely.
In the limited studies that ask victims/survivors why they chose to report (see, for example,
Jordan, 2001a) the primary reasons are less varied and include:
●acting automatically/it seeming the ‘right’ thing to do;
●wanting to prevent attacks on others;
●wanting protection for oneself; and
●a desire for justice/redress.
Almost all of the factors identified by re s e a rch as increasing the likelihood of re p o rt i n g
relate to the elements which constitute ‘real rape’: that the offender was a stranger; force
and injury were involved; and it took place either in a public place or in the context of a
break-in. However, there is contradictory research that shows that for England and Wales
the proportion of stranger attacks in rapes reported to the police has decreased from 35 per
A gap or a chasm? Attrition in reported rape cases
32
28. Anonymity is not guaranteed in some of the jurisdictions where research has been conducted.
29. Although a number of projects in the USA note high rates of reporting from Black and ethnic minority women
(see, for example, Russell, 1984).

cent in 1985 to 12 per cent in 1996 (Harris and Grace, 1999).30 The only exception is
when the person whom the woman first tells about the assault strongly communicates that it
is not the woman’s fault and supports the idea of reporting to the police (Schwartz, 1997,
pxiv).
In terms of the criminal justice system, the issue of ‘prompt’ (i.e. immediate) re p o rt i n g
continues to be influential, despite the fact that it is no longer officially a form of
corroboration. Jordan’s (1998) study of women reporting rape in New Zealand shows that
the police were the first to be told in only six per cent of cases, and in almost half of the
re p o rted cases (46%) someone other than the woman made the initial contact with the
police.31 In the majority of instances this took place with her consent and/or co-operation,
but a minority were catapulted into a situation not of their choosing. The BCS data are
similar, with 52 per cent of those reporting doing so themselves, in 35 per cent of cases
someone else doing so, and in 13 per cent the police finding out some other way (Myhill
and Allen, 2002, pvii). Whether having the decision to report made by someone else is a
factor in some subsequent decisions to withdraw complaints has yet to be systematically
studied.32
Expanding the concept of ‘real rape’
The data from UK and international research question the presumptions that rape is a single
event, committed by strangers. Rape is a more frequent and mundane crime than
conventionally believed, with current and ex-partners featuring stro n g l y, and for a
substantial proportion of women rape involves repeat victimisation. The power of the ‘real
rape’ template, however, continues to affect how rape is defined and understood by
e v e ryone, beginning with victims themselves (Myhill and Allen, 2002). An intere s t i n g
example from the most recent BCS findings is that less than half (43%) of women who had
experienced an assault that met the legal definition of rape defined it as such, and this was
even lower where the perpetrator was a current or ex-partner (31%). However, where the
assault had led to additional physical injury, the proportion defining it as ‘rape’ increased
markedly to 62 per cent (Walby and Allen, 2004). The irony of low levels of reporting by
current partners is illustrated by Bergen’s (1995) qualitative study, which demonstrates that
33
What we know about rape and attrition
30. A more recent study (Lea et al., 2003) and this report’s data (see later sections) found a less marked decrease in
the proportion of assaults by strangers.
31. This is similar to a Scottish study (Chambers and Millar, 1983): in 40 per cent of cases someone other than the
victim made the decision to report, and 17 per cent would have preferred the police not to have been told. The
i n t e rviews with victims in the study also revealed that only 29 per cent made the decision to re p o rt entire l y
alone.
32. The ability to explore this issue was compromised by the CPS asking that a series of questions on who the
complainant first told and their response be removed from the questionnaire, in order to prevent the research
tools being deemed disclosable in any cases that went to court.

‘marital rape’ is often extremely brutal, with one-third being termed ‘sadistic’, and the
majority of women being raped ‘frequently’ i.e. more than 20 times. Only one-third defined
these incidents as rape at the time, and only half of this group immediately separated. Re-
definition occurred when: the assaults reached a level of brutality associated with stranger
rape; the woman accessed support; and/or she separated. Overlaps between rape and
domestic violence have received limited attention in either research or policy development
(an exception is Walby and Allen, 2004). This overlap however clearly deserves greater
exploration, not least with respect to the protection issues victims may need addressing in
order to sustain a prosecution, and the finding that women who kill abusive partners are
much more likely to have been subjected to repeated sexual violence (see, for example,
Jones, 1980).
Summary
●Home Office data on re p o rted rape cases in England and Wales show a
continuing and unbroken increase in re p o rting to the police over the past two
decades, but a relatively static number of convictions, thus the increasing justice
gap.
●Prosecuted cases involving children were more likely to result in conviction than
those involving adults.
●All UK studies of attrition in rape cases concur that the highest proportion of cases
are lost at the earliest stages, with between half and two-thirds dropping out at
the investigative stage, and withdrawal by complainants one of the most
important elements.
●A number of studies have found high rates of ‘no criming’, not limited to the
official guidelines for this category.
●R e s e a rch on attrition in Europe shows that a decline in the pro p o rtion of
prosecutions and convictions was a common trend, although only Ireland had a
lower conviction rate than the UK.
●R e s e a rch to date has identified four key points at which attrition occurs: the
decision to re p o rt itself; the police investigative stage; discontinuance by
p rosecutors; and at trial. Victim withdrawals can occur at each stage but the
highest proportion is evident at the first two points.
A gap or a chasm? Attrition in reported rape cases
34

4. Understanding attrition: the research findings
This chapter seeks to strengthen the knowledge base on the processes involved in attrition,
especially with respect to early withdrawal by complainants.
The authors begin by outlining the attrition process in the study sample, firstly (see Table 4.1)
according to the categories used by the police and most previous studies of attrition: ‘no
crime’; undetected; detected no proceedings; and detected. The detected cases that are
proceeded with then appear under CPS prosecution and trial. The range of the research
data and the prospective design means it has been possible to develop a more detailed set
of categories that explicate the complex and varied reasons behind the official designations.
These categories are: no evidence of assault; false allegation; insufficient evidence; no
p rospect of conviction; victim declines to complete initial process; and victim withdrawal
(see Table 4.2). The first four reflect police decision-making (often in liaison with CPS) and
the latter two actions taken by the complainant. The diff e r ence between ‘insuff i c i e n t
evidence’ and ‘no prospect of conviction’ is the extent of CPS involvement. ‘Victim declines
to complete initial process’ covers cases where the complainant chooses not to make an
official complaint, give a statement and/or have a forensic examination, as well as those
w h e r e the complainant did not respond to early police eff o rts to make contact.3 3
Withdrawals are defined as cases where these initial processes take place and support for
the case is withdrawn – sometimes within days of the report being made. Both constitute this
report’s original category of ‘early withdrawals’.
Subsequent sections deal with a series of ‘Attrition points’ revealed by this analysis, which
follow a roughly chronological order of the criminal justice process. In Attrition point 1,
which focuses on the decision to report to the police, findings are based on the full case-
tracking sample of 3,527 cases, i.e. both unre p o rted and re p o rted cases; analysis fro m
Attrition point 2 onwards is based only on cases reported to the police.
For each Attrition point key factors contributing to cases being lost or not proceeding are
identified, bringing to bear a range of relevant project data. Three elements of the data, in
35
33. Originally, the term ‘non co-operation with the prosecution’ was used, but during analysis it was realised that
this represented a police perspective, rather than a more open possibility that it could be personal factors or
system failures that accounted for the actions of the complainant. It was the following passage from an interview,
which led to this reconsideration: “I just was really, really angry. And, somehow it’s not acceptable for you to be
like that. Because even in dealings with the police you’re expected to be courteous and to be co-operative, this
is what I feel is the expectation of you as the victim, to be co-operative” (St Mary’s, Service User Interview 11,
Undetected).

p a rt i c u l a r, provide a window in to the complex decision-making processes of both
complainants and criminal justice professionals in rape cases: the case-tracking database;
p ro formas completed by police officers on case pro g r ess and outcomes; and the
q u e s t i o n n a i res from and interviews with complainants. Extended quotes from the
interviewees have been used in places to convey a sense of the complexities involved and,
either the context in which they are making decisions or how the responses of the CJS are
perceived. Interview material from SARC staff and police officers is also drawn on where
relevant. The primary interest was to explore the early attrition process, which has been less
researched than stages involving the CPS and trial, so whilst data on ultimate outcomes are
presented, the most detailed analysis concentrates on earlier stages. The data which follow
o ffer some support to those who have been critical of the investigative and pro s e c u t i o n
p rocesses, whilst also illuminating the less than simple challenges and dilemmas facing
police, CPS and, ultimately, government about how to investigate, assess and classify cases.
The attrition process
Using police designations and the entire data-set, more than one-fifth of re p o rted cases
(22%, n=575) were ‘no crimed’ and one-third (33%, n=882) were undetected. Just under
o n e - t h i rd (30%, n=787) had been detected, but in 12 per cent of cases (n=320) no
p roceedings were brought. In a further one-fifth of cases (15%, n=399) there was not
complete information on case outcomes, either because the police failed to complete pro
formas or the case was not concluded at the end of the research. Additionally, there were a
small number of self-referrals who had reported at an earlier point but it was not possible to
a s c e rtain precisely when, or what the case outcome was, as well as a small number of
cases that were being investigated in a difference police force area.
Table 4.1 presents these data in summary form, calculated in the same way as the Harris
and Grace (1999) study – looking separately at the investigative and prosecution processes
– for the three SARCs and amalgamated Comparison areas. Cases where the police
designation (n=339) or CPS decision (n=60) is missing in the data have been excluded
here, reducing the sample to a total of 2,244 cases, and hence, changing the percentages
noted above.
A gap or a chasm? Attrition in reported rape cases
36

37
Understanding attrition: the research findings

The first point to highlight is that the conviction rate is higher for this sample than the
national averages for 2001 and 2002. This is accounted for by the fact that two of the
SARC areas (St Mary’s and STAR) had higher conviction rates, and represent two-thirds of
cases in the sample.34 This provides some limited data suggesting that SARCs can make a
contribution to decreasing attrition.
In order to examine the reasons for cases not proceeding within the diff e r ent police
classifications (‘no crime’, undetected and detected no proceedings), the outcomes for all
2,643 cases re p o rted to the police in the case-tracking database were collated using a
combination of the police and authors’ categories (these are presented separately for each
site in Appendix 5). This analysis reveals the continued inconsistency in police allocations,
with cases for each of the categories appearing under all the police designations.
A salient example is the ‘no crime’ category, which, as documented in previous studies,
continues to be used for a far wider group of cases than counting rules designate. Over one
in five of all re p o rted cases in the sample (22%, n=575 of 2,643) were re c o rded by police
as ‘no crimes’, rising to more than one in four (26%, n=575 of 2,244) of those where the
police classification was known (see Table 4.1). Whilst officially this category should be
limited to a narrow range of circumstances (re c o rded in erro r, where the offence took place
in another jurisdiction, where there is credible evidence that no offence took place), it still
functions as something of a ‘dustbin’. There are a pro p o rtion of re p o rts in the sample that
w e re made in other jurisdictions, both in the UK and abroad, and whilst the latter is not
p roblematic, the former raises the possibility that some re p o rts (n=36 in our sample) are
being counted twice, if official figures include ‘no crimes’. The credible evidence that no
o ffence took place includes what the authors have called ‘no evidence of assault’ and cases
designated as false allegations, the latter re p resenting just under one-third (30%, n=175) of
‘no crimes’. However, false allegations also appear under several other headings. A furt h e r
37 false allegations appear under other classifications at the police stage (29 as undetected
and 8 as detected no proceedings), while there are four where the classification was
u n k n o w n .
Other obvious anomalies in the ‘no crime’ category include: cases designated ‘no crime’
when the reason for no further action is that a) there has been a victim withdrawal, b) the
case is dropped because the victim is ill or especially vulnerable, c) there is insuff i c i e n t
A gap or a chasm? Attrition in reported rape cases
38
34. The least unknown outcomes were for STAR, due to the case-tracking element of their service provision, and for
the Comparison groups, where Project Sapphire provided much appreciated assistance in the final stages of
data analysis in chasing up case outcomes. It is, therefore, possible that the conviction rates in the St Mary’s and
REACH areas are higher still, since the ‘outcome unknown’ category by definition contains some cases that were
proceeding to trial, but for which the eventual outcome is not yet known.

evidence or d) the suspect has not been found or identified. The ‘no crime’ category
comprises a complex layering of different kinds of cases and circumstances, many of which
a re not ‘false’ in the literal meaning of this term. Clearly, neither victim withdrawals nor
cases where there is insufficient evidence should be included as ‘no crimes’. Cases were
also wrongly designated ‘undetected’ when there was a) a named suspect, b) a suspect had
been arrested and was on police bail, and c) when an advice file had been submitted to the
CPS. The ability of the Home Office, or even local police forces, to monitor attrition is
severely undermined by such inconsistency in basic data entry.
The attrition process revisited
E a r l i e r, Table 4.1 presented all cases in the case-tracking sample that were re p o rted to the police
in the SARC and Comparison areas where police and CPS decisions were known (n=2,244). In
Table 4.2 aggregated data for all areas are presented using the categories which aff o rd a more
detailed examination of the attrition process. Cases with missing data at the police (n=299) and
CPS stages (n=60) are, again, removed, making a base sample of 2,284.3 5 Analysis using both
sets of categories in combination is presented for each individual re s e a rch sites in Appendix 5.
In the following sections there are more detailed discussions of the most significant Attrition
points evident in this sample of over 2,000 cases. These ro u g h l y, although not exactly, follow
the points in time at which decisions are made in order to focus on the early stages of the
p rocess, which previous studies, along with this study, confirm are the points at which most
cases are lost/dropped. This section of the re p o rt concludes by combining categories acro s s
the time periods and exploring whether there were any factors associated with a conviction.
The influences at play in terms of attrition at each stage can be multiple and may be specific
to an individual case, however, triangulation of the case-tracking data, service user
q u e s t i o n n a i res and interviews enables identification of some of the most important factors.
Summary
●Around one-quarter of reported cases were ‘no crimed’.
●Just under one-third of reported cases were detected but in a notable proportion
of these no proceedings were brought.
39
Understanding attrition: the research findings
35. There is a slight discrepancy between the sample sizes in Tables 4.1 and 4.2 due to the different categories
reported on at the police stage. In Table 4.1 all cases where the police classification was known were included,
w h e reas in Table 4.2 the authors re p o rt on the reason for the case not proceeding. There are a number of
additional cases here, because while it was known that the case did not proceed and the reason why, the police
classification was not known and therefore could not be included in Table 4.1.

A gap or a chasm? Attrition in reported rape cases
40
Table 4.2: Recalculated Attrition process: 2,284 cases reported to police where
research categories known
n=2,284 No % Overall %
Police 1,817 100% 80%
Insufficient evidence 386 21
Victim withdrawal 318 17
Victim declined to complete initial process 315 17
Offender not identified 239 13
False allegation 216 12
No evidence of assault 83 5
No prospect conviction 37 2
Not in public interest 20 1
Other 67 4
Reason unknown 136 8
CPS 145 100% 6%
Caution/final reprimand 9 6
Discontinued 38 26
Victim withdrawal 25 17
Suspect fled prior to PDH 1 1
Pending trial 72 50
Trial 322 100% 14%
Trial to be rearranged, suspect fled 1 <1
Victim withdrawal 15 5
Discontinued/withdrawn at court 19 6
Acquittal 104 32
Guilty plea 89 28
Not clear if guilty plea/conviction 17 5
Part conviction 11 3
Conviction 66 20
Total convictions 183 8%
Source: case-tracking database
●A conviction rate of eight per cent was found across the research sites collectively,
although this was slightly higher for two of the individual SARC sites.
●This is higher than the national average of 5.6 per cent.

●Inconsistencies were found in the police classification of case outcomes,
particularly in the ‘no crime’ category.
●Eighty per cent of cases did not proceed beyond the police stage, with a further 6
per cent being discontinued by the CPS, and only 14 per cent proceeding to trial.
Attrition point 1: reporting to the police
This section explores the decision-making of complainants at the earliest stage of the pro c e s s
– re p o rting rape/sexual assault to the police. Using the case-tracking database comparative
analysis was conducted between re p o rted and unre p o rted cases across a range of variables,
to see whether there were particular characteristics associated with either group. The serv i c e
user questionnaire and interview data provide additional insights into the range of routes by
which the decision to re p o rt is arrived at and how soon after the attack re p o rts are made.
The decision to report
Of the 3,172 victims/survivors who contacted the three SARCs during the evaluation
period, 2,288 (72% of all cases) reported to the police. In the Comparison areas all cases
(n=355) were, by definition, reported to the police, making a total of 2,643 reported cases
f rom all sites. A pro p o rtion of self-re f e rrals to the SARCs are seeking support and
information about historic abuse, but the majority relate to recent assaults (for discussion of
the differences between police and self-referrals see Lovett et al., 2004). A further factor is a
small number of cases where the assault took place in another country whilst on holiday or
business. Although 13 per cent (n=96) of self-referrals had a forensic medical examination,
only a small proportion of this group (n=13, all from St Mary’s) went on to make an official
police report.36 Among the self-referrals, there is therefore, a discernible group of women
and men making a clear decision to avoid the CJS, whilst seeking the health checks and
other support and advice they need to cope with the recent sexual victimisation.
Analysis of the case-tracking database was undertaken comparing the 2,643 cases who
re p o rted to the police with the 884 cases who did not. The key findings are summarised here .
●R e p o rting decreases with age: 90 per cent (n=355) of under-16s re p o rted, compare d
with 72 per cent (n=581) of those aged 26 to 35 and 69 per cent (n=458) of those
aged 36 and over. However, where details were known, there was far more re p o rt i n g
by a ‘known other’ (56%, n=96) than by the victim/survivor themselves for under- 1 6 s .
41
Understanding attrition: the research findings
36. Another small group of cases re p o rted some time prior to attending the SARC (n=16; St Mary ’s n=5, STAR n=11).

●A slightly higher pro p o rtion of women than men re p o rted to the police (75%,
n=2,474 compared with 69%, n=169).
●The majority of cases where the complainant had a disability (88%, n=170) or
was involved in prostitution (97%, n=66) were reported.37
●Assaults by known perpetrators were the least likely to be re p o r ted (75%,
n=1,369) compared with the highest rate of re p o rting for assaults by re c e n t
acquaintances (88%, n=297).
A larger proportion of the sub-sample of questionnaire respondents and interviewees (82%,
n=188 and 77%, n=43 respectively) had made a report to the police. Analysis of these
data provides insights into additional factors influencing the decision to report to the police
that escape quantitative analysis. The main reasons respondents gave (n=178 who
answered the question, multiple responses possible) were:
●to sanction the perpetrator (69%, n=122);
●to protect others (57%, n=101);
●because they thought they should (50%, n=89); and
●fear of the perpetrator (44%, n=79).
W h e re known (n=179), just under half (46%, n=83) made the re p o rt themselves, most
commonly by telephone, and three-quarters (76%, n=136 of 188) said that they made the
decision to report.
The 40 respondents who did not report to the police provided a range of reasons for their
decision. The most common were:
●the abuse took place some time ago (n=5);
●concern about not being believed (n=5);
●not wanting others to know (n=4);
●lack of faith in the police (n=4);
●confusion (n=4);
●relationship to the offender (n=4); and
●not being able to face the criminal justice process (n=3).
Therefore, for a significant proportion of this group lack of faith in aspects of the criminal
justice system acted as a deterrent to reporting.
A gap or a chasm? Attrition in reported rape cases
42
37. This might indicate limited awareness of the possibility of self-referral amongst these groups.

Timing of the report
Whilst prompt re p o rting no longer functions as a form of legally re q u i red corro b o r a t i o n ,
and is now understood as telling anyone and not just confined to making an official report,
delays in re p o rting continue to influence the perceptions of police and prosecutors. This
study has already cited research showing that: a) police are rarely the first people to be
told; b) the importance of the response of those who are told first in determining whether the
case will be reported and c) that reports immediately following assaults are the exception
rather than the rule. Figure 4.1 presents findings on the length of time between the assault
and the reporting to the police, and shows that around half were reported within six hours
(54%, n=784) and four out of five within 24 hours (79%, n=1150).
Figure 4.1: Length of time between the assault and reporting to police
Source: case-tracking database
n=1,462 cases reported to the police at all sites where dates/times known.
C ross-tabulations were conducted within the case-tracking database to see whether there
was any correlation between these time intervals and relationship with the perpetrator in
cases involving a single assailant. Unsurprisingly, the vast majority of those assaulted by
recent acquaintances and strangers reported within 24 hours (87%, n=155 of 179 recent
acquaintances and 85%, n=347 of 398 strangers). However, high proportions of ex- and
current partner cases were also reported within this period (74%, n=84 of 113 ex-partners
and 66%, n=70 of 106 current partners). These pro p o rtions also held for re p o rts made
within 12 hours.
43
Understanding attrition: the research findings

Questionnaire respondents reflected these reporting patterns, with over half of reports (58%,
n=105 of 180 where details known) being made within 12 hours of the assault and almost
three-quarters (70%, n=126) within 24 hours. Those who had delayed reporting for more
than 12 hours attributed this to some combination of shock, fear, confusion or distress. For a
small number the confusion related to whether, in fact, anything had happened and this was
particularly the case where the respondent thought that they had been drugged.
The following accounts from interviewees illustrate the multiple routes to re p o rting, why
delays of more than 24 hours are not only quite common, but also entirely understandable,
and the role of others – both in making reports and/or encouraging them.
I did, but I did it 24 hours later, ’cause of the two kids, well, not just because of them,
I just didn’t know what to do…They sent a WPC just to talk to me, and then the CID
woman, I still am annoyed by her. She sort of just barged in … “Right, come on,
we’re going to St Mary’s, going to the police station.” And I didn’t want my little girl
to know. She’s only five … she was adamant that I had to get someone to mind them,
which I did ... the WPC was great, you know, she [CID] was talking to me and but
she just sort of took control, because I hadn’t even made up my mind whether I did
want to press charges or even do anything about it really, but she just sort of seemed
to make me do it. (St Mary’s, Service User Interview 3, Undetected)
Well actually, I didn’t report, I thought I was to blame. It’s quite simple. I’d gone out
and stayed out longer than expected, I’d had enough to drink to know that I needed
a little help to get home. I needed somebody to get me a taxi, put me in it, and then I
could have got home. So I was tipsy, but not drunk. And this gentleman said he
would look after me, and I asked him could I trust him? And he said yes. He brought
me home, on the tram, which I’d never been on before ... We got to my house, and
… at the door, I stretched my arm and said, “I’ll get you a taxi.” I didn’t invite him in.
But the next thing, as I’m phoning for the taxi, I realised he was sat on the settee.
Then I was raped, and then he went. I felt a stupid old woman, because I’m fifty-odd
… But I felt as though I could trust him ... I wasn’t going to tell anybody because I felt
so ashamed, and I felt as though it was my fault … Anyway, she [neighbour] came in
the day after, and she just looked at me, and she said, “You look shocking, tell me
what’s the matter.” “If I told you what’s the matter,” I said, “You’ll never speak to me
again” … I felt so dirty, even though I’d showered, showered, showered and washed
and washed – but fortunately I hadn’t washed the clothes or the bedding. Anyway,
she was asking me questions, had I been assaulted, had I been mugged, and I said
“No, no, no,” and then she just turned round and she said, “Have you been raped?”
A gap or a chasm? Attrition in reported rape cases
44

Well then I broke down ... And she said, “You know what you’ve got to do.” I said,
“I can’t.” She said, “Well I am. I will.” And she phoned the police.
(St Mary’s, Service User Interview 10, Acquittal)
I was away from home, on a residential weekend as part of a course, so, you know,
apart from one friend there, my family weren’t with me, and it was people I knew so I
was really shocked, so I didn’t report it to the police for a few days. I find people don’t
understand this unless they’ve been through it, because, especially when it’s people you
k n o w, and also there was drugs involved so I had a couple of hours of the assault
which I couldn’t remember ... So I had trouble believing it myself at first … I really
didn’t know what to do, I didn’t know the centre at St Mary’s existed, the only thing I
thought of at first was I spoke to a friend at home and she said, “What about ringing
one of these Rape Crisis Centres, somewhere like that?” I was just too exhausted
e m o t i o n a l l y, I’d been physically sick and I was, you know, crying most of the time and I
just didn’t have the energy to deal with it or to make any decision. I came home, I was
very strange, I was sat home for the day before I told my husband, and he wanted me
to go to the police and then I got frightened because it became an issue of control and I
couldn’t think straight, I needed to know that I was doing what I needed to do and not
what other people told me to do. I booked an appointment with my doctor … I’ve had
sort of post-natal depression and also clinical depression over the years, and I trusted
her … she said I should see St Mary’s ... St Mary’s said that if they did a forensic
examination, none of the tests would go anywhere unless I involved the police. And I
thought, “Well what’s the point?” … I don’t want to just be examined for no purpose …
I didn’t realise that I wouldn’t get any answers anyway, but that’s beside the point.
(St Mary’s, Service User Interview 11, Undetected)
There were a couple of reasons [why I didn’t report it]: at first, I couldn't face it. I thought
that I would eventually, but because my husband was actually, I think, mentally ill before
this happened, and quite distressed, and he became very difficult afterwards, I got more
sort of concerned with his health than mine. And it seemed that things were diff i c u l t
enough without facing anything even bigger … What I would have liked to have done,
was to give all the information – ’cause I knew who the man was – but not to have to
have an interview. I would have liked to have written it down and handed it over and
said, “Look, there’s the information, I’m sorry but I can’t face being questioned or being
in court or anything else, but just for your information, this was it.” And I did eventually
ring a police station to ask if it was possible to give a name but not be involved, and
what I was told was that they couldn’t guarantee it … So that put me off … I felt all along
my husband wasn’t so much thinking about what was right for me, it was for him.
(REACH, Service User Interview 4, Unreported) 45
Understanding attrition: the research findings

Summary
●72 per cent of re f e rrals to the SARCs and 75 per cent of the sample overall
reported to the police.
●Reports to the police were more likely with respect to younger complainants and
those with disabilities or involved in prostitution.
●Assaults by known perpetrators were the least likely to be reported and those by
recent acquaintances the most likely.
●The most common reasons questionnaire respondents gave for reporting were: to
sanction the perpetrator; to protect others; they thought they should; and fear of
the perpetrator.
●The most common reasons given by those who did not report were: the abuse
took place some time ago; fear of not being believed; and concerns about
elements of the CJS process.
●Where reports to the police were made, around half were within six hours and
the majority within 24 hours of the assault. The nature of the relationship with the
perpetrator had limited effect on the promptness of reporting.
Attrition point 2: no evidence of assault and false allegations
This section addresses two groups of cases that did not proceed beyond a very early stage –
those where, according to the police, there was no evidence of an assault taking place and
those designated as false allegations. Findings on the former are drawn solely from the police
p ro formas, while three perspectives on false allegations are provided by analysis of: the case-
tracking database, the police pro formas, and interviews with police officers and SARC staff .
No evidence of assault
A group of cases not addressed in previous studies are those where there is, what has been term e d
in this study ,‘no evidence of assault’ (n=83). Police pro formas provided sufficient detail for more in-
depth analysis on 56 of these cases (far more were evident in the St Mary ’s data, n=35), of which
t h e re were two distinct types. The first type is when there is no complaint of rape from the individual
(see also Jordan 2001a): in 12 of these cases at the point the police made contact the individual
was either too drunk or distressed to communicate, the presence of injuries or torn clothing led to an
initial coding of ‘suspected rape’, but as soon as the person was able to provide an account it
became clear that this was not what took place; for a further 11 the complaint was made by
someone other than the supposed victim (often a family member or partner) and the initial
investigation revealed that no sexual offence had taken place. There is a larger group of 31 cases
A gap or a chasm? Attrition in reported rape cases
46

(28 women and 3 men) where people regain consciousness – either in a public place or at home –
with no memory of the previous time period. They approach the police because they are worr i e d
that they may have been assaulted whilst asleep, unconscious or affected by alcohol or drugs. In
e v e ry such case police re p o rted that the ensuing forensic examination (and in some cases toxicology
tests) suggested no sex had taken place and the individuals were re p o rted as being relieved by this.
Clearly ‘no crime’ really did take place in many of these situations. It is, however, an open question
whether this group of cases should be combined with false allegations in police re c o rd-keeping. This
is for two reasons: firstly ‘no evidence of assault’ is not a contested category; and secondly, given
the data presented later on how ‘previous allegations’ undermine the credibility of complainants,
t h e re is a danger that such re c o rds might in the future be read, erro n e o u s l y, in this way.3 8
False allegations
False allegations have been one of the most contested areas within law enforcement re s p o n s e s
to rape, with re s e a rch suggesting rates are no higher than for other crimes sitting alongside
p e rceptions of police officers and the media who take the opposite view. Some of the most
f requently cited US studies put the rate as low as two per cent (Katz and Mazur, 1979).
T h e re were 216 cases classified as false allegations: as a pro p o rtion of all 2,643 cases re p o rt e d
to the police this amounts to 8 per cent; as a pro p o rtion of the 1,817 cases not pro c e e d i n g
beyond the police stage it is 12 per cent (see Table 4.2). In only six of these cases was there
evidence of anyone being arrested, and in only two cases were charges laid, although there
w e re at least 39 named suspects. Six advice files were submitted to CPS, with respect to possible
c h a rges being laid against the complainant for perv e rting the course of justice, and two were
c h a rged. Intere s t i n g l y , most cases in this category had a forensic examination (82%, n=178),
whilst far fewer made a formal statement to the police (58%, n=126), suggesting that this is a
critical stage for the admission or designation of allegations as false.
C ross-tabulations using the case-tracking database comparing the group designated false
allegations (n=216) with proceeding cases (n=527) revealed the following findings.
●Cases involving 16- to 25-year-olds accounted for a higher proportion of cases
designated false (52%, n=112) than of cases that proceeded (42%, n=221).
●Those in full-time work formed a smaller proportion of those whose cases were
designated false (11%, n=23 compared with 17%, n=91 of those proceeding),
whilst the opposite was the case for those who were unemployed (37%, n=79
compared with 18%, n=95).
47
Understanding attrition: the research findings
38. The same argument would apply to cases in the previous section, where there was no complaint from the
victim/survivor themselves.

●Whilst small numbers, those with a disability were almost twice as likely to be in
the false allegations group as the non-disabled (51%, n=24 of 47 compared with
28%, n=192 of 695), and in 19 of these cases mental health and learn i n g
difficulties were present.
●Only 2 of the 66 women involved in prostitution who reported were in the false
allegations group.
●A greater degree of acquaintance between victim and perpetrator decreased the
likelihood of cases being designated false.
●Cases were more likely to be designated false where previous allegations had
been made and/or the complainant had attended the SARC or reported to police
in the Comparison areas on a previous occasion.
Exploring the grounds on which cases were deemed to be false allegations is revealing and
120 pro formas contained explanations: in 53 cases the police stated that the complainant
admitted the complaint was false, most commonly within days of the initial accusation; 28
cases involved retractions; three non co-operation and in 56 cases the decision was made by
the police on evidential grounds. Intere s t i n g l y, the majority of cases in which the complainant
themselves admitted the allegation was false could be categorised as the often quoted
motives of ‘revenge’ (n=8) and ‘cover-up’ (n=25). Although, as the explanations provided on
the police pro formas which are summarised in Box A, reveals, the terms ‘revenge’ and
‘cover up’ do not do justice to the complexity of the circumstances involved.
Box A: Examples of false allegations admitted to by complainants
‘Revenge’ ‘Cover up’
Against a difficult neighbour. Eight cases of hiding consensual sex with
Against an ex-partner who the woman another man from husbands/partners.
had sex with hoping this indicated Nine cases of avoiding confrontations with
reconciliation, whilst he had no intention parents.
of leaving his new partner. Four cases where accusations arose where
Against an ex-partner who had forced the complainant was being investigated
sex on previous occasions, although for fraud or theft.
not on this one. One case of an affair with a father-in-law.
To make an ex-partner feel sorry for her.
Against a man who the women had sex
with who ignored her the next day.
A gap or a chasm? Attrition in reported rape cases
48

T h e re were other unusual contexts, including a young Asian woman who made an
accusation as a strategy to prevent her family taking her to Pakistan for an arr a n g e d
marriage, a young woman wanting the morning after pill after having had unprotected sex
and another attempting to regain her mother’s support after being thrown out of the family
home. In most of these cases there was no named assailant.
Whilst not all of the cases where the complainant ‘admitted’ making a false allegation can
be taken at face value, many of the accounts provided on the police pro formas had a
surface credibility. There is, however, considerably less certainty about some of the reasons
underpinning police decision-making: 23 refer to mental health problems, 23 to previous
allegations (some overlap occurs here); and 13 refer to alcohol and drugs. Evidence from
other witnesses that supported the account of the accused was noted in nine cases (two of
these involved an alibi), and the absence of CCTV evidence supporting the complainant’s
account in a further nine cases.39
Inconsistencies in the account given by the complainant feature strongly in 30 cases, and in
only 5 of these did police officers note that this involved deliberate untruths (lies as opposed
to not revealing the whole truth). Previous studies have highlighted the ways in which
withholding, or not remembering, information is the outcome of fearing disbelief, which may
subsequently be interpreted by police officers and prosecutors as ‘lying’.
Fearing disbelief and judgement, victims of rape may try to embellish their accounts,
or conceal wrong-doing, in order to make themselves appear more ‘believable’ to the
police. (Jordan, 2001a, p93)
[P]olice scepticism promoted the narration of the very inaccuracies which, in turn,
consolidated the police view that women fabricate complaints and make false
allegations. (Chambers and Millar, 1983, pp86-7)
Reflecting this, a number (23%, n=14) of police officers interviewed for this study raised the
issue of inconsistencies, and indicated that this had implications for the perc e i v e d
genuineness of complainants.
In theory, I would say that somebody who has been raped is going to stick quite
rigidly to the account that they give, and that might be an account they give to a
uniform police officer and then to us and then perhaps the doctor as well … whereas
49
Understanding attrition: the research findings
39. These were in almost all cases where the complainant claimed to have been in a particular location, often the
city/town centre, at a certain time and the CCTV recordings did not confirm this.

sometimes those that have made a false allegation, the story may well change and
sometimes they might come out and say things that you know couldn’t be possible …
or CCTV might disprove it. (Comparison 1, Police Officer, DC, F14, July 2002)
The data on the pro formas limit the extent to which one can assess the police designations,
but their internal rules on false complaints specify that this category should be limited to
cases where either there is a clear and credible admission by the complainants, or where
there are strong evidential grounds. On this basis, and bearing in mind the data limitations,
for the cases where there is information (n=144) the designation of false complaint could be
said to be probable (primarily those where the account by the complainant is referred to) in
44 cases, possible (primarily where there is some evidential basis) in a further 33 cases,
and uncertain (including where victim characteristics are used to impute that they are
inherently less believable) in 77 cases. If the proportion of false complaints on the basis of
the probable and possible cases are recalculated, rates of three per cent are obtained, both
of all reported cases (n=67 of 2,643), and of those where the outcome is known (n=67 of
2,284). Even if all those designated false by the police were accepted (a figure of
approximately ten per cent), this is still much lower than the rate perceived by police officers
interviewed in this study. A question asked of all of them was how they assessed truth and
falsity in allegations and within this, 50 per cent (n=31) further discussed the issue of false
allegations.
I do deal with a lot of allegations that don’t go anywhere, because it transpires that
it’s not quite a rape, certainly with juveniles. They’re in a difficult position, because
they scream “Rape,” they tell Mother and Father, and before you know it the whole
circus has turned out. And then I think it’s very difficult for them to say, “Well, it didn’t
quite happen like that, actually – it was like this” … It’s the same with some adults as
well, they’ve been out, had a good night, got drunk, done something they shouldn’t
have done, they’re married, and they’ve got to go home and – and then they scream
“Rape” and they ruin it for the genuine ones. I don’t think that people who make false
allegations should be penalised, because clearly there’s something wrong, for them
to make this allegation, I don’t think that’s the right way to deal with it … we spend a
lot of time dealing with allegations of rape that really aren’t rape, that take a lot of
time up. I would say a good half that come through are not genuine ones.
(St Mary’s, Police Officer DC, F8, March 2003)
We have a lot of allegations that are then retracted, we have a lot of allegations that
it comes out in the wash one way or another that it was consensual. He says it’s
consensual and she doesn’t, or they’ve been together for like hours beforehand, she’s
A gap or a chasm? Attrition in reported rape cases
50

gone back to his flat … But stranger rape, you immediately start to think “Oh God,
this could be a real proper sort of drag you in the car,” absolutely nothing
beforehand’s happened. I think subconsciously you would consider it more serious …
I think I’d have more belief in the victim, that was saying it was by a stranger, that …
it was a proper rape, rather than perhaps someone who said “It’s my ex-boyfriend,
he came round”, ’cause then you start to think things like “Oh, she’s just getting back
at him now.” (St Mary’s, Police Officer DC, F2, July 2002)
Well, honestly, it’s because most of them are not telling the truth … I think what
happens to a lot of adults is they may have consensual sex with somebody, they get
found out by their husband, partner, whoever, they then say “Oh but I didn’t consent”
as a way of getting themselves out of that trouble … I mean I have dealt with
hundreds and hundreds of rapes in the last few years, and I can honestly probably
count on both hands the ones that I believe are truly genuine.
(Comparison 1, Police Officer DC, M2, June 2002)
The forensic nurse at St Mary ’s also illustrates how embedded a tendency to disbelieve
complaints is within the police.
One of the things that comes up, time and time again, is their automatic disbelief of a
complainant. I’d be getting details off the officer, and they’d say ‘Ooh, so what did
they say to you in there? What do you think’s happened? Sounds a bit dodgy to me’
… It seems to be a natural cynicism, and I am sure that along the way they have met
complainants who have made false allegations. It seems to me that everybody then
s u ffers because they happen to have met one or two or maybe more who’ve told
lies… I think they disbelieve – apart from what … fits their stereotype of a rape,
perhaps stranger or the injuries or definitely if she was beaten to a pulp … I admit it
is a difficult crime to prove. But I think the way in which people are handled can
always be improved on. They don’t know, they weren’t there, they should treat that
person with respect. I don’t see what’s so difficult about that. I think even if it’s not
happened, they’re more likely to get a complainant being honest about that much
quicker if they’re treated the right way, the way any of us would want to be treated.
(St Mary’s, Forensic nurse, F1, November 2001)
The interviews with police officers and complainants’ responses show that despite the focus
on victim care, a culture of suspicion remains within the police, even amongst some of those
who are specialists in rape investigations. There is also a tendency to conflate false
allegations with retractions and withdrawals, as if in all such cases no sexual assault
51
Understanding attrition: the research findings

o c c u rred. This re p roduces an investigative culture in which elements that might permit a
designation of a false complaint are emphasised (later sections reveal how this also feeds
into withdrawals and designation of ‘insufficient evidence’), at the expense of a care f u l
investigation, in which the evidence collected is evaluated. These perceptions and
orientations are not lost on complainants.
Reflections
I n t e rnational re s e a rch contains salutary lessons about the ease with which cases are
dismissed as ‘false’. In her analysis of 164 police files in New Zealand, Jordan (2001b)
found 3 cases that had been designated false, which subsequently turned out to be early
reports of serial rapists. In one of these, a young woman, who was discounted since she
was on the fringe of a gang and had minor criminal convictions, named a rapist, who
subsequently went on to commit at least 45 stranger rapes over 13 years.4 0 One police
officer involved in this case reflected on what it had taught him.
So the key issue for me is corroboration, to a lesser, a much lesser extent, is
credibility. Because professional women get attacked, prostitutes get attacked, people
who have been abused previously get attacked, people with criminal convictions get
sexually attacked … You’ve got to be so careful because sometimes you get
discrepancies from genuine complaints too. It’s only going through all the evidence,
really, and what supports it and what doesn’t. (Jordan, 2001b, pp268, 270)
Summary
●T h e re is a small group of cases, initially treated as rape where there is no
evidence of an assault: primarily where a third party makes the report and the
victim subsequently denies; or where the victim suspects being assaulted while
asleep, unconscious or affected by alcohol/drugs but the medical/fore n s i c
examination suggests no sex has taken place. How the police should designate
such cases is problematic.
●Eight per cent of re p o rted cases in the sample were designated false by the
police.
●A higher proportion of cases designated false involved 16- to 25-year-olds.
●A greater degree of acquaintance between victim and perpetrator decreased the
likelihood of cases being designated false.
A gap or a chasm? Attrition in reported rape cases
52
40. The cold case review team in Project Sapphire are apparently encountering similar patterns in London (personal
communication, 2004).

●Cases were most commonly designated false on the grounds of: the complainant
admitting it; retractions; evidential issues; and non co-operation by the
complainant.
●In a number of cases the police also cited mental health problems, pre v i o u s
allegations, use of alcohol/drugs and lack of CCTV evidence.
●The pro formas and the interviews with police officers suggested inconsistencies in
the complainant’s account could be interpreted as ‘lying’.
●The authors’ analysis suggests that the designation of false allegations in a
number of cases was uncertain according to Home Office counting rules, and if
these were excluded, would reduce the proportion of false complaints to three per
cent of reported cases.
●This is considerably lower than the estimates of police officers interviewed.
Attrition point 3: insufficient evidence
In this section the different kinds of cases that fail to move beyond the police investigation
for evidential reasons are looked at. The largest proportion of cases dropped at the police
stage is on the grounds of ‘insufficient evidence’ (21%, n=386). Two additional groups of
cases where lack of evidence is an issue are included in the discussion below: those where
no offender was identified (13%, n=239) and those where there was ‘no prospect of a
conviction’ (2%, n=37). In combination these total 662 cases that did not proceed because
of evidential difficulties. Two key data sources provide perspectives on these groups of
cases: police pro formas, which illustrate the range of issues at play for investigating police
o fficers; and interviews with complainants, which in some cases provide an altern a t i v e
narrative of the conduct of investigation process.
Analysis of the police pro formas enable the identification of a range of reasons why cases
are lost at the investigative stage, including:
●the inability of the complainant to provide a clear account of what happened;
●failure to identify the assailant;
●failure to trace a named assailant;
●assessments by the police (often in consultation through an ‘advice file’ with the
CPS) that cases are ‘weak’ evidentially; and
●the judgement that there is ‘no prospect of a conviction’.
53
Understanding attrition: the research findings

F o r ty-one complainants were unable to give a clear and coherent account of what
happened, but within this group were several women with severe learning diff i c u l t i e s ,
several with mental health problems, and most were individuals who had been so drunk or
drugged that their memory was severely impaired.
In 239 cases the police failed to identify the assailant. From the information provided in the
pro formas it appears that DNA tests were conducted in 19 cases, 5 of which resulted in
hits on the national database. Although it was stated that there were no forensic findings in
30 cases, it is unclear why DNA testing was not conducted in the remainder of cases. More
troubling are the cases (at least 27) where the assailant was not traced – their identity was
known, but they were never formally interviewed. In at least eight instances the suspect
absconded when first contacted. Several of our interviewees (n=7) were in this group, and
were less than impressed by the police investigation.
[H]e contacted the police, and said oh, it was all a big mistake, and could he come
and see them? And instead of saying, “No, we’ll come and see you,” they said,
“Well come up on Monday” … So what did he do? Sunday, he got his brother to go
down to London, to collect his car, and he caught the first train, bus, walked,
whatever, boat, out of the country. Gone.
(St Mary’s Service User Interview 7, Undetected)
They haven’t even tried, because I have seen him a couple of times since, so there’s
a general area where he’s likely to be, sort of car that he drives, I got the beginning
of the registration number, and there’s a strong possibility he’s a builder, and I saw
him come out of a builders’ merchants and they haven’t followed that up or anything.
(Comparison, Interview 5, Undetected)
For the remaining cases, police decision-making turned on evidential issues connected to the
complainant, largely where the victim’s account was either re g a rded with suspicion or not
s u p p o rted by other evidence: in 17 cases the victim’s credibility was explicitly re f e rred to; in
10 cases the issue of previous allegations was seen to cast doubt on the current complaint; in
27 inconsistencies or lies were re f e rred to; and in 26 the police and/or CPS took the view
that sex was consensual (in 5 of these cases it is suggested that the complainant agreed). In a
f u rther seven cases the complainant failed to pick out the suspect in an ID line-up.
The 37 cases designated ‘no prospect of conviction’ were all submitted to the CPS for
advice. In almost all the cases factors related to the complainant were the reason for
dropping the case: three had learning difficulties, three had other forms of disability, three
A gap or a chasm? Attrition in reported rape cases
54

had mental health problems and three had made previous allegations (not mutually
exclusive). The other cases referred to issues of inconsistency and consent.
The following examples from police pro formas gives a sense of the way decisions
regarding cases that are dropped on grounds of insufficient evidence are presented.
Box B: Examples of insufficient evidence cases from police pro formas
AP (Aggrieved Party) lied to the police throughout the enquiry and CPS felt there was no
evidence. AP lied about drug intake and amount of alcohol consumed. Toxicology tests
proved she was extremely drunk and high on cannabis. Advice file submitted to CPS:
Issues were lack of victim credibility. (St Mary’s, Pro forma 11)
No forensic evidence, no CCTV evidence, no description of offenders.
(St Mary’s, Pro forma 77)
Advice file to CPS re g a rding incident and discrepancies with AP/off e n d e r / w i t n e s s
accounts of what actually happened. CPS advised no charge due to fact that unable to
determine whether AP had been raped or gave consent . (St Mary’s, Pro forma 87)
Suffers from learning difficulties … No offences disclosed. Social Services dealing with
persons concerned. 1) AP’s tendency to make false allegations. 2) The alleged suspect’s
inability to have sex due to a medical condition. 3) Lack of forensic, medical or witness
corroboration. (St Mary’s, Pro forma 141)
IP (Injured Party) aware – investigation complete. No further action to be taken against
alleged off e n d e r. This was a case of consensual anal intercourse with consent. Consent which
is said to have been obtained through fear of violence. (REACH, Pro forma 823)
Reason for not proceeding IP found to be unreliable, CPS decision due to pre v i o u s
allegations made by IP. (REACH, Pro forma 598)
Likely to be undetected. Father has been interviewed, denies offence, leaving one word
against other. Anticipated CPS advice NFA due to time lapse between offence and
disclosure. Enquiries still ongoing with mental institutions to establish if earlier disclosures.
Final outcome: undetected. (Comparison, Pro forma 1576)
Two previous studies (Harris and Grace, 1999; Lea et al., 2003) document the failure of
cases involving women with learning difficulties and mental health problems to pro g re s s
through the system. This is especially concerning given the accumulation of data that women
in these groups are targeted for sexual assault. Whilst the new sexual offences legislation
55
Understanding attrition: the research findings

may provide a new route for women with severe learning difficulties41, this would not affect
the majority of cases in this sample. It could be argued that for this group of women the
justice gap is a chasm, which will not be bridged unless specific attention is devoted to
developing prosecution strategies that provide redress for extremely vulnerable victims.
Another concerning issue is the impact of previous allegations on decision-making, the
majority of cases where this issue is mentioned were dropped at an early point (see also
Jordan, 2001b). The data on multiple attendees at St Mary’s suggest that there is a group of
women who are subjected to repeat sexual victimisation by different assailants. Currently,
this vulnerability appears to function as a cue to practitioners to dismiss their allegations
rather swiftly. Indeed, it could be argued that the entire attrition process results in thousands
of people each year entering a group that, should they have the misfortune to be assaulted
in the future, will mean their complaint will be treated with even greater scepticism.
One of the most perplexing aspects of police investigation were a number of cases in which
‘witnesses’ were presented as ‘proving’ that the sexual encounter had been consensual, or
as supporting this interpretation. In no case was there any suggestion that these ‘witnesses’
were present when sex took place, but their evidence was deemed more credible than the
c o m p l a i n a n t ’s account (this was also true in a further nine cases with respect to named
suspects, who in interview claimed consent). One of the interviewees had this experience
and provides insight into the chasm between her account and perception, and how the
investigation was conducted.
I just felt as though they were going through the motions really. The forensic stuff
never even got sent off … I was just frustrated – basically the investigation consisted
of getting them in and them giving their story, sending off questionnaires to students
at the college I was at, because the college didn’t want the police to come in
because they didn’t want them to be a visible presence ... And people wrote really
helpful comments like, “She appeared to be really happy in their company in the
evening,” which of course I was, because I hadn’t just been assaulted by them at the
time! You know, really helpful comments like that which the police sort of put to me
afterwards. I thought “Well what the hell relevance is that to what happened
afterwards?” … I mean the officer in charge of the case never actually set eyes on
me. The impression he gave me over the phone was he didn’t believe me from the
word go, and I thought, “You’ve never even met me, you don’t know me”. I just got
the impression he couldn’t really be bothered to come over here and meet me
because it was a bit of a drive … He basically just said to me, “I’ve got your file in
A gap or a chasm? Attrition in reported rape cases
56
41. The Sexual Offences Act, 2003, contains a strict liability offence for people deemed to not have the capacity to consent.

front of me, I’ve taken it to my senior and he says there’s not a case, it wouldn’t stand
up in court.” And I asked him to go through what was there: “Well, we’ve got
questionnaires and people said things like the fact that you seemed quite happy
chatting to them before”. He kept saying, “I’m not saying anything but you know how
it could look”. I felt he wasn’t just saying how it could look, he was saying, “This is
how it looks to me” … At the moment I’m just left thinking a terrible thing happened
and all I got after it was a load of shit … Nothing is different to how it would have
been if I’d not said anything, apart from the fact that I’d still be on the course. And
I’d have completed. (St Mary’s, Service User Interview 11, Undetected)
In the research data, little indication was found of early consultations between the police
and CPS resulting in further investigation, or that effort was put into building evidence that
might support the complainant’s account. In the case above, for example, the woman
concerned could not understand why the friend she had phoned first and her husband who
had been witness to the impacts of the sexual assault on her had not been interviewed, yet
a lot of emphasis was placed on questionnaire responses from people who hardly knew her.
There were also a number of pro formas where the investigating police officers questioned
CPS advice that there was insufficient evidence to charge. During interviews several police
officers admitted to being confused as to precisely what CPS needed in order to proceed
with a prosecution.
I feel there’s different standards within CPS depending on who’s assessing the papers
… I had a case where … the decision file went into CPS, that came back,
discontinued, before even the toxicology reports had come back from the lab. I was
really quite disheartened … I feel like whenever you have one that comes down to
[an] issue of consent, unless you have areas of bruising, tearing, damage or
whatever, from the medical examination as supporting evidence, it is going to be one
word against another, and very, very few of those cases ever get to court whereby
that IP [Injured Party] is given the opportunity to put his or her side across. I just feel
that the CPS give up too easy, too soon.
(REACH, Police Officer CID/Sexual Offence Liaison Officer (SOLO), F1, May 2003)
I don’t know how they reach their decisions but I find them on the whole very
negative. I think they always want the easy option out so, say you have a victim
under the age of sixteen, they are always going ‘Oh, we can drop this to a USI42’
and they never really want to pursue the rape.
(REACH, Police Officer SOLO, F7, May 2003)
57
Understanding attrition: the research findings
42. USI – unlawful sexual intercourse.

Practice within CO14 – Clubs and Vice, in the Metropolitan Police – who investigate cases of
sexual exploitation, including trafficking, has been to seek early meetings between police, CPS
and the prosecuting council. These help to explore where weaknesses in a case might be and
what additional evidence might support women involved in prostitution who have re p o rt e d
sexual assault. This echoes the eff o rts of some Sex Crime Units in the USA, where a
commitment to finding ways to prosecute what have previously been deemed ‘unpro s e c u t a b l e ’
cases (Vasschs, 1994), has fundamentally altered the stru c t u re and content of investigations.
The movement in 2004 to locating CPS lawyers in police stations and for the CPS to take over
responsibility for charging, could provide a route to such productive inter-agency work, but
only if the investigation and prosecution of rape cases is fundamentally reassessed. Wi t h o u t
this the outcome will simply be a faster route to the large number of advice files (n=330 in this
sample) re t u rned with a recommendation that ‘no further action’ is taken.
Whilst there are undoubtedly a pro p o rtion of cases in this category that could never be
prosecuted effectively, there are equally others where lack of care and commitment in the
investigative process along with preconceived stereotypes and attitudes have contributed to
the justice gap. Jordan (2001b) reached similar conclusions, based on analysis of police
files in New Zealand, noting that police officers (and in the context of this project also
prosecutors) used evidential difficulties that might arise as ways of judging genuineness. The
veracity of victims/survivors was assessed on the basis of how the practitioners thought both
the judge and jury would respond. This creates a discriminatory response to rape compared
with other crimes, since it increases the amount of evidence needed to convince a police
o fficer and/or prosecutor that an offence has taken place, and in some cases such
judgements are being made before a full investigation has taken place (Jordan, 2001b).
Summary
●Evidential issues account for over one-third of cases lost at the police stage (21%
i n s u fficient evidence, 13% offender not identified and 2% no prospect of
conviction).
●Within this group there were a number of complainants, often with learn i n g
difficulties or mental health issues, who were unable to give the police a clear
account of the assault.
●DNA testing appeared to be conducted in a minority of cases.
●There were at least 27 cases where the offender was identified but failed to be
traced by the police.
●In a substantial number of cases where the offender was identified the decision
not to proceed was linked to victim credibility.
A gap or a chasm? Attrition in reported rape cases
58

●The pro formas indicated that consultation between police and CPS rarely led to
further investigation or enhanced case-building.
Attrition point 4: early withdrawal
A significant group of cases (34%, n=633) lost at an early stage is where victims/survivors
opt out of the process at an early stage. In half of these cases (n=315), victims ‘declined to
complete the initial process’. Below complainants’ experiences of the early investigation,
including elements of practice or treatment that contributed to some declining to proceed is
discussed. In addition, there were 318 cases that constitute early withdrawals (a smaller
number – 40 – of late withdrawals are dealt with in subsequent sections), which are also
discussed drawing on analysis of the case-tracking database, the police pro form a s ,
q u e s t i o n n a i res and interviews with service users, and interviews with police officers and
other SARC staff.
Victims who decline to complete the initial process
This group re p resents some of the very earliest lost cases. According to the police pro
formas 315 individuals did this (12% of 2,643 reporting to the police) in relation to one or
more of the following: making a formal complaint; forensic examination; giving a statement;
or withholding information. However, it is worth noting that this outcome was more common
in the areas without an integrated SARC (see Appendix 5). The latter category refers to
cases where the complainant refused to name their attacker (or more rarely where the
assault took place), and whilst in some cases it was clear that they knew the person well, in
others it was unclear whether they could have identified either the assailant or the location
of the assault. Another group includes those who fail to respond to police efforts to contact
them, or who move without leaving a forwarding address.
The contribution of SARCs to responses to re p o r ted rape is addressed in another
publication, as are the varying processes involved in reporting at the six research sites (see
Lovett et al., 2004). Prior to a forensic examination, however, complainants will have had
contact with at least two police officers: one to whom they make the first report, and one
who makes arrangements for the examination and accompanies them. This officer is likely to
take basic details of the offence, so that preliminary phases of the investigation with respect
to the crime scene and suspect can begin, although in some instances a full statement
appears to be taken before the forensic examination.43
59
Understanding attrition: the research findings
43. Internationally there is now increasing recognition that a fuller and more accurate statement is likely if it is taken
a day or two after the initial report (see, Jordan, 2001b).

H e re findings that, from the perspective of complainants, illustrate points where either serv i c e
p rovision or the behaviour of personnel do or do not create a context of support and
confidence in the process are focused on. Responses to the service user questionnaire s
illustrate that the availability of women – police officers, crisis workers and fore n s i c
examiners – was important for the majority (52%, n=93 for the initial police response; 65%,
n=85 for the statement; and 83%, n=90 for the forensic examination). Whilst one of the
i n t e rviewees and Jord a n ’s (2002) re s e a r ch make it clear that this is not a case of ‘any
woman will do’, the vehemence with which almost half of the questionnaire re s p o n d e n t s
a rticulated their feelings suggests that the absence of a female doctor may result in some
complainants refusing a forensic examination, and the lack of a trained female police off i c e r
may at worst mean they do not make a statement, or at best provide a less detailed one.
Reasons for pre f e rring female police officers were expressed either in terms of not being able
to talk to or trust a man (33%, n=44) or it being easier to talk to a female (26%, n=35),
reflecting findings of previous studies (see, for example, Jordan, 2001a and Temkin, 1999).
If possible just female officers. I knew I was safe with the male officers, I just couldn’t
cope at the time. (St Mary’s, Q1, 1051)
It was extremely difficult telling [a] female what had happened – I couldn’t have done
it if it was a male. (REACH, Q1, 3003)
I had just been raped by a male, I didn’t want to have a male to talk to at that time.
(STAR, Q1, 2082)
I felt embarrassed, awkward, didn’t want to tell him the intimate details.
(Comparison, Q1, 4005)
Every woman/man who has been assaulted, please give them female officers, or at
least a choice. (St Mary’s, Q1, 1008)
Although over two-thirds (70%, n=121) of questionnaire respondents evaluated the initial
police response positively, a significant minority (30%, n=53) did not. Intere s t i n g l y, this
varied across the four sites with the highest levels of dissatisfaction registered in the STAR
area (46%, n=28).
Preferences for female practitioners were even more pronounced with respect to the forensic
examination (see Lovett et al., 2004 for more details), and both were echoed in interviews.
A gap or a chasm? Attrition in reported rape cases
60

Alady came down from the local police station, really nice woman. And just took
some sort of basic details at first, you know, related to the assault, and that was fine
... (a) I was glad it was a woman, (b) she believed me … that was my first contact
with the police, so that was really important and that was good. And she didn’t want
to know too much at that stage ... Everybody at this stage believed me, my friends,
my husband, the doctor, the policewoman.
(St Mary’s, Service User Interview 11, Undetected)
They gave me the option of whether I wanted a female or a male doctor as well …
[and that was important] ’cause it’s not really very comfortable at the best of times
having them kind of things done, and I imagine it’s worse at the best of times with a
male than with a female, but at the worst of times, it’s a lot better having a female!
(STAR, Service User Interview 15, Undetected)
The forensic examination itself was experienced, almost universally, as a difficult and
i n t rusive process, made more problematic when one or more of the following factors
applied: lack of a female examiner; delay in arranging an examination; unpleasant
e n v i ronment; and the manner of individual professionals. A small percentage of police
referrals to the SARCs declined to have the forensic examination (2%, n=40).
It could be speculated that this group contains those who are most ambivalent about the
re p o rting process, since they take their decisions at the earliest point possible. However,
responses from the questionnaires suggest that additional factors are at play. Sixteen of the 188
respondents who re p o rted had chosen not to make a formal statement, and police attitudes –
either being disbelieved or discouraged from proceeding – and the CJS process more generally
w e re uppermost for over half (five of eight) of those who provided an explanation.
The police officers at the station after saying it wouldn’t be in the public interest to get
it into court. (St Mary’s, Q1, 1023)
The CID were blunt and unsympathetic. They said if I made a formal complaint they
would have to question the neighbours – but I wanted to keep it quiet out of
embarrassment. (St Mary’s, Q1, 1060)
Police said he had witnesses, so I felt as if they didn’t believe me. (STAR, Q1, 2002)
The DC didn’t believe me, called me a liar. (STAR, Q1, 2080)
61
Understanding attrition: the research findings

[I] was told [by police] because of history [of drug abuse] it could not go further but
[I] could make a statement if [I] still wanted. (Comparison 1, Q1, 4015)
That the behaviour of police officers plays a part in the earliest withdrawals is furt h e r
supported by the fact that more declined to give a statement or make a formal complaint
than refused the forensic medical, an inherently difficult and intrusive pro c e d u re (Kelly,
2003a). In the context of uncertainty and trauma, professionals need to be skilled in order
to build rapport and trust, and police officers have a responsibility to conduct a thorough
investigation. In some of these instances neither appeared to be the case.
The process of statement-taking was also inconsistent, with almost thre e - q u a r ters of
questionnaire respondents (73%, n=114 of 156 where times known) making a statement
immediately or on the day after reporting, 15 per cent (n=24) up to one week later and 12
per cent (n=18) over a week after reporting. Just under one-quarter (23%, n=35) said they
had been under pre s s u re to provide an immediate statement. Experiences of making a
statement among the interviewees revealed that 20 of the 43 who made one described the
p rocess favourably, but 20 described it negatively, while the remaining three did not
express an opinion. Being too tired and confused to provide an accurate account, feeling
judged or disbelieved, insensitive questioning, the length of time taken due to statements
having to be hand-written and not being able to recall details were the predominant issues.
The latter applied particularly to interviewees who suspected drugs had been used in their
assaults, or who had been assaulted when they were drunk.
It is a concern that over a tenth of complainants in the case-tracking sample chose not to
complete the first stages of an investigation is concerning. A number of cases, however,
revealed that the complexity of individual circumstances and potential consequences of an
investigation could be the deciding factor. For example: the re p o rt was made by a health
p rofessional following disclosure to them, but against the wishes of the victim/survivor; the
re p o rt was made by someone else and the victim/survivor was ambivalent; details of the rape
e m e rged in an investigation of domestic violence (in half this was post-separation), in which the
woman was seeking protection and never considered pursuing a rape charg e4 4; finding the
p rospect of an investigation where one’s family, friends and/or work colleagues will not only
find out, but also be questioned, too hard to bear; and finally wanting to forget/deal with the
assault in one’s own way. That some people decide a rape investigation is either not what they
sought, or not what they can sustain, is likely to remain the case however much the services and
responses of professionals become more attuned to the realities of rape and sexual assault.
A gap or a chasm? Attrition in reported rape cases
62
44. In several of these the police officer noted that woman made explicit comments about not wanting to pursue a
rape complaint for the sake of her children.

P rovision of female examiners and trained female police officers at the initial stages is
clearly important for many complainants (including the majority of male respondents), and
essential for a small proportion. Questionnaire and interview data also make clear that a
climate of belief, privacy, feeling safe and reassured by the professionalism of staff creates
a context in which difficult processes can be endured. Equally the absence of these elements
can contribute to a context in which confidence is eroded at the outset, which can either
become a factor in withdrawal or affect the quality and content of evidence-gathering.
Early victim withdrawals
An additional group (n=318), whilst completing the initial processes of a fore n s i c
examination and formal statement, withdrew during the early investigative stage. Cro s s -
tabulations using the case-tracking database, comparing this group with cases that
proceeded (n=527), revealed the following.
●Complainants aged 16 to 35 accounted for a greater pro p o r tion of the
withdrawal group (72%, n=228) compared with the proceeding group (63%,
n=332), whilst for under-16s the opposite was the case (9%, n=29 withdrawn
compared with 20%, n=103 proceeding). The age group most associated with
withdrawal were 16- to 25-year-olds, who constituted almost half of all those who
withdrew (48%, n=153 compared with 42%, n=221 in the proceeding group).
●Those currently in a relationship – either married or living together – were slightly
m o re likely to withdraw (53%, n=63 compared with 47%, n=55 where case
proceeding).
●In terms of the relationship with the perpetrator, the only cases that showed any
distinct variation were those involving current partners, which re p resented a
slightly higher proportion of those in the withdrawal group than the proceeding
g roup (17%, n=48 versus 10%, n=52). In contrast, the pro p o r tion of those
involving family members in the proceeding group was twice that of those in the
withdrawal group (9%, n=45 versus 4%, n=11). Of all cases involving known
perpetrators twice as many appeared to be proceeding than being withdrawn
(65%, n=350 compared with 35%, n=185).
●The involvement of drugs or alcohol appeared to have some impact on rates of
withdrawal. Cases involving both were more likely to be withdrawn (57%, n=31
c o m p a red with 43%, n=23). When taken individually, though, this was more
marked in relation to drugs (54%, n=43 withdrawn versus 46%, n=37
p roceeding) than alcohol (44%, n=127 withdrawn versus 56%, n=159
proceeding).
63
Understanding attrition: the research findings

In only one-fifth (21%, n=66) of cases did the police pro formas provide any details on the
reasons for victim withdrawals, although it is notable that, compared to no crimes and
i n s u fficient evidence, there were far fewer re f e rences to either victim characteristics or
evidential problems here. Where there is information, for half of the complainants (n=33) fear
of the court process was the deciding factor, 15 refer to complainants’ uncertainty about
whether they were raped, including some thinking it may have been consensual, seven state
that the person ‘wants to put the event behind them’, and another seven complainants are
re p o rted to have made the decision after consultation with their family. From the police
i n f o rmation, in at least 13 of these cases the assailant was a curre n t / e x - p a rt n e r, which may
have played a part in the decision (in six cases it clearly did).
There are 74 cases in which the complainant withdrew where the offender had not yet been
identified, which accounted for just under one-quarter (23%) of all victim withdrawals. Since
there is no immediate prospect of a court case, this decision represents, on some level, a
loss of faith in the CJS – either in the ability of the police to detect the offender, or in the
process at a more interactional level with individual officers. The woman speaking below
also illustrates that perceived slackness and failure to maintain contact can pro m p t
withdrawal, prevented in this instance by St Mary’s facilitating discussion.
They didn’t phone me. I phoned them in July to find out what was happening, ’cause
it was just screwing me up, ’cause I didn’t know where he was and I was having
panic attacks, looking round me all the time whenever I went anywhere ... I felt the
police were totally unsupportive and they didn’t understand how I felt ... I came here
[St Mary’s] with the decision that I was dropping the case. X then contacted [police
station], and spoke to them, and then it was decided that I was to talk to the police
again. And lo and behold, two weeks later, I’m sat in here again, with the police
telling me all this information about this guy that they might have told me beforehand.
Like the fact that he’s got a history … he was accused of sexually harassing two
members of staff. And then he was accused after he’d left of actually sexually
abusing a patient. And at that point I thought, “Well I can understand why the police
are so keen to get him” – I was what we call a “date rape”, next time it might be
somebody off the street. So I said, “OK, fine, go ahead, yes, I’m prepared to carry
on with this.” And since then, bugger all. Nothing. And I phoned them the other
week to tell them I’ve moved house: “Right, OK, fine, goodbye” … Hey, you know,
I’m a victim here, and I get absolutely nothing, nothing from the police, at all … I just
feel that the police, after the initial excitement, it’s almost as if you’re forgotten.
You’re a forgotten statistic, and that doesn’t help. I’m surprised more people don’t
drop out. And if they do, that would be why. Because the police don’t maintain
contact with you. (St Mary’s Service User Interview 7, Undetected)
A gap or a chasm? Attrition in reported rape cases
64

There are also nine unusual cases, five where the suspects admitted the crimes (one handed
himself in to the police station) and four where a stranger rapist was identified from the
DNA database. All would have been strong evidentially, but the complainants still withdrew.
Again one has to ask whether in the interim period they had lost confidence in the CJS.
Another pattern involved suspect interviews (n=42) and/or arrests (n=30); in at least one-
t h i rd the offence was either denied or sex was claimed to be consensual. Wi t h d r a w a l
seemed to follow this fairly closely, possibly when police returned to put these versions of
events to the complainant.
Just under half (45%, n=74) of questionnaire respondents who re p o rted had either
withdrawn their complaint or were unsure about pursuing a legal case at this stage. Their
comments further illustrate that fear of court process, and especially anticipation of having to
give public testimony about intimate intrusions, having one’s behaviour scrutinised and
judged and the prospect of acquittal were powerful disincentives.
Reliving it, being questioned in court. Too upsetting to go to court.
(St Mary’s, Q1, 1037)
Why put myself through all that to have a court say that it’s not worth it.
(St Mary’s, Q1, 1023)
I couldn’t bear the cross-examination or being made to feel in the wrong.
(REACH, Q1, 3026)
Giving evidence in front of strangers. I wouldn’t stand a chance in court.
(REACH, Q1, 3047)
Scared I won’t be believed, but I feel guilty ’cause he may hurt others.
(STAR, Q1, 2040)
Because I feel I will not be believed due it being from a partner continuing over
several years. (STAR, Q1, 2052)
T h e re were also several examples of witness intimidation, which had clearly not been
addressed or stopped.
I never took him to court, due to the harassment I have been getting from this person
and his friends. (St Mary’s, Q2, 1023) 65
Understanding attrition: the research findings

For a significant group, thinking beyond reporting to the potential of appearing in court is a
major barrier to continued engagement with the CJS. There were also a small number of
cases where the police officer provided enough detail to make clear that they had doubts
about the account provided by the complainant, which they put to them forcibly – it was
following this that the withdrawal took place. The authors’ data from complainants
themselves underline that the tone, content and timing of providing information about the
low conviction rate can act as a prompt to withdrawal.
I think they should give the benefit of the doubt more … They were acting like they
were there that night and they knew what had already happened … it felt like
“we’ve done all this before, we know what’s going to come out of it” … they were
giving all this negative stroking, not one positive to anything … Everything I said they
questioned back at me. “Why did you go there?”, and then I’d tell them why “Well
he said this or he said that.” Like everything I did they had a question to come back
at me ... It felt like he was the victim and I’d done something wrong ... I’m not a
police officer so I don’t know, but the questioning I thought were disgusting, the way
they treated me. I thought, “If anybody else has been through what I’ve been through
no wonder they don’t press charges” … Before I went in the interview room she was
dead nice and dead posh, but as soon as I got in that room she just went dead stern:
“This is what it will be like in court. This is all the questions, you’re going to have,
tough questions” … If it ever happened again – I hope it never does – I don’t think
I’d even bother going to the police. There’s no point … They were just like talking me
out of it … That negatively influenced me, definitely … I’m not so much angry at him,
I’m angry at the police. They’re supposed to be a public service, and they failed me.
(St Mary’s, Service User Interview 4, Detected no proceedings)
This officer was like “I just don’t believe you.” And, I mean, I could have swung for
her because she said to me “Oh well I’m trained in this,” I’m thinking, “You’re not,”
because from the moment I saw her, it’s like “You’re trying it on.” “You are just
making this up” ... It was her body language and certain things she said … I wasn’t
interested to take it any further. (REACH, Service User Interview 6, Status unknown)
Over one-third (n=26) of the police officers interviewed across all the re s e a rch sites
concurred that the fear of court is one of the main reasons why complainants withdraw from
the process.
A gap or a chasm? Attrition in reported rape cases
66

There may be a perception about getting into court and having to relive it all. Having
everything be brought up and dragged up again and their whole life and lifestyle
examined in a courtroom. That can put somebody off completely. Some people just
can’t face the ordeal, and it is an ordeal going to court.
(St Mary’s, Police Officer DC, M9, March 2003)
I think some people just genuinely don’t want to face the court system, which is
understandable because it’s still quite an adversarial thing for them to go through, it’s
not a pleasant experience, and I think people think that – certainly rape victims think
– they may be questioned about their sexual history. I think that does put a lot of
people off. Once they’ve sat down and thought about it.
(STAR, Police Officer CID, M8, April 2003)
The court case, they know that court cases are and can be traumatic, and that your
character can be called into question and you can be cross-examined over a long
period. I think that definitely puts people off. I suppose their whole sexual life is aired
in public, basically. (Comparison 1, Police Officer DC, M1, June 2002)
Fear of going to court was also cited by the majority of SARC staff and key informants as
the most significant factor in withdrawal from the criminal justice process. Their responses
were typified by the following comment.
Some of them feel that they’ll be torn to pieces.
(STAR, Case Tracker, F1, December 2001)
The interviews with police officers revealed very little consistency in both when complainants
were given details on the process and the type of information provided. Over half of those
who had early contact with complainants were clear that they gave this information at the
earliest possible point, and many thought that being ‘realistic’ was in the complainant’s best
interests. These are examples of the ‘second guessing’ that other researchers have identified
in CJS responses to rape (see Kersetter, 1990 with respect to the police and Fro h m a n n ,
1991 with respect to prosecutors), with each tier basing their decisions on the responses
they expect or predict from the next.
I must admit that one of the things that I do, and it’s perhaps a wrong thing, but I
think you’ve got to empower the victim. She’s got to be the one that’s got to make this
choice and got to know what she’s up against … then it’s her decision. When I talk
about what the police can do, one of the things that I do is go into details about what
court can mean, the things that can happen. I’m not painting a black picture, I’m 67
Understanding attrition: the research findings

painting a truthful picture, so that at no point along the line when they’ve been in
court for three days they can say “You never told me about this.”
(St Mary’s, Police Officer DC, F10, March 2003)
When someone comes into the police station we are very victim-focused, but we also
have to make people aware that in the courts it’s a little bit different and they may get
a harsher time. We’re just being completely honest with these people.
(Comparison 2, Police Officer DS, M1, February 2003)
Nonetheless, a staff member at one of the SARCs identified this as a potential disincentive to
proceeding.
Some of the police … will immediately say to this woman “Do you realise how bad
it’s going to be in court?” I feel that’s far too soon to be making a comment and I
think they shouldn’t say that. It’s almost trying to put her off. It’s like I’ve just been
raped and I’m trying to deal with this and now you’re telling me, “Do you realise
when you go to court they may bring up this, they may bring up that?” … It’s almost
like they’re testing her out, is she going to go ahead with this, is this real, and I really
think this shouldn’t be done. Awoman’s not in a place to be thinking about court
within hours or days of being raped. I think that if the police stopped doing that
maybe she would actually continue. (REACH, Counsellor, F3, January 2002)
Reflections
Early withdrawals, therefore, fall into several categories: decisions made very quickly which
reflect an original doubt/ambivalence about the process which is not allayed (or is possibly
even re i n f o rced) by the re p o rting and early investigative process; those which follow
aggressive questioning or discouragement by the police; those where the complainant feels
t h e re is nothing to support her account/where the suspect makes clear he will plead
consent; and those where the prospect of a trial is too terrifying to contemplate.
Whilst there are many reasons why complainants would be ambivalent about being
involved in a rape prosecution, those who have made an official report to the police have,
at least in part, expressed an interest in achieving justice. That so many lose confidence so
quickly, clearly not believing that the CJS is capable of either recognising their needs and
c o n c e rns or delivering justice, is surely an indication that there are serious gaps in how
re p o rted rape and sexual assault is currently responded to. In this context it is wort h
reflecting on what lies behind the word ‘withdrawal’: for victims/survivors this constitutes a
A gap or a chasm? Attrition in reported rape cases
68

withdrawal of trust and confidence, whereas at the official level it is re g a rded as the
withdrawal of the original complaint.
Summary
●Victims who declined to complete the initial process and early victim withdrawals
accounted for the largest proportion (34%) of cases lost at the police stage.
●M o re victims declined to complete the initial process in the areas without an
integrated SARC.
●The availability of women police officers, crisis workers and forensic examiners
was important for the majority of respondents and interviewees.
●Seventy per cent of respondents were satisfied with the initial police response.
●Twelve per cent of all those who reported declined to complete the initial process
by not either making a formal complaint, not having a forensic examination, not
giving a statement or withholding information.
●Those respondents who did not make a formal statement cited being disbelieved
or discouraged from proceeding by the police and fear of the CJS process as key
factors.
●A proportion chose not to proceed as a way of taking back control and/or for
other personal reasons.
●The age group most associated with early victim withdrawal were 16- to 25-year-
olds.
●Only a small number of police pro formas contained details of the reason for the
withdrawal, but where details were known over half cited fear of the court
process.
●No offender was identified in just under one-quarter of all withdrawals,
suggesting that for these complainants there was a loss of faith in the ability of the
police to catch the offender, or in the CJS more broadly.
●Just under half of first questionnaire respondents who reported to the police had
withdrawn their complaint or were unsure about pursuing a legal case, and cited
fear of court process, the prospect of giving public testimony and being judged as
disincentives.
●More than one-third of police officers interviewed agreed that fear of court is one
of the main reasons why complainants withdraw from the process.
69
Understanding attrition: the research findings

Attrition point 5 – CPS decision making
Only 527 of the original 2,643 cases were referred through to the CPS for prosecution,
meaning that over thre e - q u a rters of cases (80%, n=2,116) were lost or dropped at the
police stage. Of the 527 just under two-thirds (61%, n=322) proceeded to the trial stage,
although in 60 cases the final CPS decision is not known. From the data in Table 4.2, it
appears that discontinuance by the CPS is a relatively small element in attrition of cases that
reach the CPS (26%, n=38). However, the missing data for this stage and the involvement
of the CPS in a significant number of cases where police took decisions after submitting
advice files, mean that these data substantially underestimate the CPS role in attrition.
Late victim withdrawals also feature here, with 40 taking place at these stages in the
process. Four of our questionnaire respondents were in this group. In one case, the police
had advised that the CPS were unlikely to pursue the case, and the complainant also
referred to unhelpful preparation for court. The remaining three all referred to fear of court,
and one also mentioned her relationship to the offender. That these perceptions were, at
least in some cases, linked to information from the police is illustrated by this comment.
I dropped it because I was too scared and the police said it was a hard case. Not
many women win the case. (St Mary’s, Q3, 1025)
CPS advice and decisions are taken on the basis of the Code of Practice, which contains a
two-level test: an evidential test and one with respect to public interest. Only cases where
the evidential test suggests a realistic prospect of conviction are taken forward, and there
may be some circumstances where the public interest test is seen as applying in rape cases,
most commonly where the vulnerability of the complainant leads to a conclusion that a legal
case risks further harm to them.
One issue worth noting here, which might affect decisions on the evidential test, were
findings from the analysis of forensic medical reports and police statements. In a minority of
cases, information which was not relevant to the case was recorded, which could provide a
‘back door’ route to the introduction of sexual history evidence. In a small number of
medical re p o rts re f e r ences were found to prostitution, contraception and previous abuse.
Fifty police statements (31 St Mary’s, 19 STAR) were analysed. Seven of these statements
contained irrelevant material, which in four cases (all from the St Mary’s area) related to
previous sexual history, and was particularly notable in a case involving a gay man, where
extensive details of his sexual life were re c o rded. This was further confirmed by
questionnaire respondents: over half (58%, n=93 of 160 who made formal statements) had
A gap or a chasm? Attrition in reported rape cases
70

been asked about their sexual history. Within this group, one-third (33%, n=30) were asked
about their prior or current relationship with the alleged assailant, 19 per cent (n=17) about
their sex life generally and 48 per cent (n=44) were asked about both these issues. Whilst a
current relationship with a suspect is relevant to the facts of a case, anything beyond this
recorded in a statement may enable the protections of section 41 of the Youth Justice and
Criminal Evidence Act, which limits contexts where sexual history evidence can be
i n t roduced in court, to be evaded (Temkin, 2003). Concerns about credibility of the
victim/witness play a part in assessments at the CPS stage (HMCPSI, 2002), and it will be
especially concerning if irrelevant discrediting material is present in disclosable evidence.
There is a discernable shift within the CPS from independence and separation from other
agencies, and a particular interpretation of ‘objectivity’, to engagement with inter-agency
processes and recognising the needs of victims of crime. In relation to rape, there is also a
policy change with respect to the development of specialisation. These changes are ,
however, very new and their implementation inconsistent. Whilst these moves are welcome,
they need to be linked to a focus on building cases. One recent development worth noting
here is the practice in Newcastle CPS of attempting to learn through formal reports from
counsel from both successful and unsuccessful prosecutions, especially acquittals, in order to
feed back into better case preparation.
Summary
●Only a small proportion of reported cases were discontinued at the CPS stage.
H o w e v e r, this underestimates CPS’ role, since they offer opinions on larg e
numbers of advice files.
●A small number of forensic medical re p o rts and police statements contained
i n a p p ropriate material on complainants’ previous sexual history that may have
affected CPS assessments of their credibility.
●The CPS needs to build on recent moves towards specialisation and place greater
emphasis on building cases.
Attrition point 6 – court and trial
Of the 2,643 reports, 322 (12%) were scheduled for trial. This is equivalent to 14 per cent
of all 2,244 reported cases where the outcome is known. Eleven per cent (n=34) failed to
reach this point due to either late withdrawal, the CPS deciding not to offer evidence or
judge-ordered acquittals. This outcome was most likely for cases involving 16- to 25-years-
71
Understanding attrition: the research findings

olds (n=15), and over two-thirds of these cases involved known offenders (n=23). Around
half of all convictions were due to guilty pleas (49%, n=89 of 183) rather than findings by a
jury. Table 4.2 highlighted that, where a full trial took place (n=181)45, an acquittal was
m o re likely to be the outcome than a conviction (57%, n=104 acquittals compared with
43%, n=77 part or full convictions).
Factors associated with conviction
Cases (n=183) that resulted in a finding of guilt (guilty pleas and whole or part convictions
at trial) have been compared with all other reported cases in the sample (n=2,460), on a
range of factors. Those factors associated with conviction are listed below.
●Cases involving under-16s were more likely to result in a finding of guilt (15%,
n=55), a rate more than twice as high as for those aged 26 to 35 (6%, n=36)
and almost four times higher than for those aged 16 to 25 (4%, n=52).
●Age is an even more pronounced factor where trial outcomes are considere d .
Trials involving under-16s had an acquittal rate of 20 per cent (n=14), whereas
for those involving adults this was twice as high at 41 per cent (n=90). The group
most associated with acquittal was 16- to 25-year-olds, among whom the
acquittal rate was 51 per cent (n=54).
●Cases involving at least one known offender were slightly more likely to result in
conviction (9%, n=117), followed by strangers (5%, n=40), whilst re c e n t
acquaintances had the lowest rate (4%, n=13). However, these findings need to
be contextualised in a number of ways: cases involving known offenders always
involved a named off e n d e r, whereas for those involving strangers the off e n d e r
was identified in only a minority of cases; in one-third (32%, n=37) of the known
offender convictions the complainant was under16.
●The absence of alcohol or drugs was associated with conviction, and was slightly
more likely in the case of alcohol (8%, n=70 versus 5%, n=42) but over three
times more with respect to drugs (7%, n=107 versus 2%, n=6).
●Whilst injury alone had little impact, cases involving the use of a weapon and
resulting external injuries were twice as likely to result in conviction compare d
with those where neither was present (7%, n=12 versus 3%, n=81).
●Convictions were obtained in four per cent (n=7) of re p o rted cases involving
women with disabilities and in 11 per cent (n=7) of those involving women in
prostitution.
A gap or a chasm? Attrition in reported rape cases
72
45. Includes acquittals and part and full convictions, but excludes the 17 cases where the outcome of guilty plea or
conviction was unclear.

●Only three per cent (n=8) of re p o rted cases involving multiple perpetrators
resulted in convictions. Seven of these cases involved two perpetrators and one
involved three.
These data suggest that whilst some aspects of rape cases mean they are more likely to
result in convictions, it is not impossible to prosecute successfully where they are absent.
Nonetheless, the ‘real rape’ template features strongly, and our data suggest that the group
least likely to find their cases proceeding are those in the 16- to 25-year-old age group.
In terms of the research participants, 20 questionnaire respondents had been to trial, and in
another a guilty plea had been offered at the Plea and Directions Hearing; a further nine
were awaiting the trial. Outcomes for the 21 respondents were 11 guilty pleas, 4 findings
of guilt by the jury and 6 acquittals. In the interview sample 15 cases proceeded to trial,
with 6 guilty pleas, 3 findings of guilt at trial and 6 acquittals.
The experience of the lead up to, and actual, trial were draining and disquieting for all
research participants.
I wasn’t sleeping, not eating … I didn’t realise you could exist on such a lack of
sleep. Because I was going through it all in my head. What if he says this? It was
bringing all the feelings I didn’t want to feel back ... I knew it would be like being
raped all over again. But there would be an audience listening this time. ’Cause of
all the people in the court. The jury, and the people that sit there just to listen, all the
clerks and the barrister, the judge, the one – the usher that takes the oath – they’re
like the audience, aren’t they? The first rape was just between me and ‘it’. T h e
second rape is in front of a lot of people. All listening. And you’re stood there. And
you’re trying to explain to these people all the pain and suffering and degradation
and violation and how can they possibly comprehend? Because they won’t have
experienced it. Could I be convincing enough by telling the truth, that they would
believe me, or would he con them like he conned me?
(St Mary’s, Service User Interview 10, Acquittal)
I do think that there should be something whereby these kinds of cases are seen
within six months. They shouldn’t be allowed to drag on. They really shouldn’t. It’s
emotionally wearing on anybody. But to actually drag on for over twelve months, it
really is soul-destroying ... Because you just can’t get on with your life. You cannot do
anything. (St Mary’s, Service User Interview 1, Acquittal)
73
Understanding attrition: the research findings

I think ’cause he was in the courtroom, I think having to see him. And also what were
they going to ask me, I don’t know, ’cause you just read about … previous sexual
history and they kind of try and make you look bad. I was concerned about that,
really, like what I was going to be asked.
(STAR, Service User Interview 14, Conviction)
Well, they made me out to be obviously the town tart, which I wasn’t, and they had
nothing to prove it, and once in my life I had to have an abortion and they made out
that I was such a bad person to have that … I don’t know how they found that [out],
and then they tried to make out that I’d had lots of abortions, which was rubbish … I
think it is really hard, because they keep saying in your face “You’re lying, you’re
lying, you’re making it up,” and you just have to keep saying “I’m not lying,” and
you just have to stick to it … but I’m glad I did now because I got to see him give
evidence, the lies that he was coming out with, so I could go to my barrister and say
“He’s lying about this, and I know he’s lying about this because I’ve got proof.”
(Comparison, Interview 2, Conviction)
One SARC counsellor highlighted the extent of complainants’ fear of what court
proceedings might entail.
[They are] absolutely terrified of going to court, and what they always say is, “His
defence is going to tear me apart, how am I going to deal with that?” Te r r i f i e d .
Terrified of the brief, terrified of the culture of gowns and wigs and the judge,
terrified of being able to see the offender, but most of all what they seem terrified of
is bringing up all their past history and him getting off.
(St Mary’s, Counsellor, F1, January 2001)
Guilty pleas are undoubtedly the outcome of careful investigation and case pre p a r a t i o n ,
resulting in strong evidence against the accused.4 6 Often they are the outcome of plea
negotiations, where prosecutors are faced with having a finding of guilt for an offence versus
taking the case through the lottery of a criminal trial. At least 30 of the 89 guilty pleas in this
sample were to lower charges; 17 involved under-16s, of which 11 were pleas to Unlawful
Sexual Intercourse (USI) and four to indecent assault. A guilty plea frequently results in a
sentence discount, in part recognising that this saved the victim/witness from having to give
evidence. In one case the plea was entered five minutes before the complainant was due to
give evidence and she expressed strong re s e rvations about what happened next.
A gap or a chasm? Attrition in reported rape cases
74
46. In a very few cases there was also some suggestion of a ‘guilty conscience’ on the part of the perpetrator, in two
instances to the extent of giving themselves up to the police at an early stage. These were invariably cases
where the rapist was known.

The judge praised him. I was angry. He only did it because all my witnesses turned
up. Men who leave the victim till the last minute before going into court should not be
congratulated. “Even at the last minute he saved the victim,” said the judge. No, he
saved himself. They played down the violence, the violence was worse than the rape.
(St Mary’s, Q3, 1010)
Better case preparation, and ensuring that no disclosable evidence contains unnecessary
material that can be used against the credit of the complainant are undoubtedly factors that
increase guilty pleas, and there may well be important lessons to be learnt from cases with
this outcome.
The issue of lack of contact between the prosecution – both CPS and the pro s e c u t i o n
barrister – was a source of frustration and confusion to many of those whose cases reached
court. Five respondents whose cases resulted in acquittals attributed this to the prosecution
counsel being poorly prepared, and appearing unfamiliar with the facts of the case.
I know they can’t tell me exactly what questions are going to be asked, but I do think
it unfair that I can’t even meet the person who is going to represent me. Even if just to
give a personal impression, even if it was just a 20 minute chat … I just feel it’s so
unjust when he has had twelve months of support and legal assistance. It’s his word
against mine, and he’s had twelve months to get his speech rehearsed, whereas I
haven’t … I found it terrible, I really did, I thought I’m just as entitled to speak to
somebody. They said, “You’ll get to see the statement before you go in.” I just thought
– so that’s it! The statement was made in a state of shock, for one thing, and I do
know there’s things I haven’t said that probably I should have … My version of events
didn’t come out at all. He didn’t ask anything. He [prosecution barrister] just made a
statement. “Oh you went this way, and you had a drink here”, and I said, “Yes.”
“And he attacked you on this particular spot,” and I’m thinking “Well, surely I’ve got
to give more than this” … He didn’t ask questions, he was reading from a piece of
paper all the time. And I thought, “He’s not even read the evidence, he’s not seen it
at all” … I’m stood there thinking, “You should be saying something, and he didn’t
say a thing … He should have intervened more … I think if I’d just been able to stand
there and say what was on my statement, my version of events, things would’ve been
so different … After [giving evidence] I said, “I can’t believe it, this is going to go
against me, I know it is. At one stage he didn’t even know my name.” And the police
said, “Oh, he’s really quite good.” That was on the Monday.And I found out on the
Thursday that he’d got away with it. (St Mary’s, Service User Interview 1, Acquittal)
75
Understanding attrition: the research findings

Two main witness statements weren’t at court, vital information details weren’t
brought to the court. Why is it [that] the defendant has a solicitor on hand able to see
all that is being brought to the trial yet the victim has nothing? (STAR, Q2, 2092)
I had no idea what I was going to be asked by my barrister, and I remember his first
opening question and I’d never been asked this, either, nobody had said to me, you
know, “What was your first sexual experience?” or, you know, “Have you had sex
before?” And, funnily enough, it wasn’t actually my boyfriend at the time that I’d first
slept with, so the barrister asked the question, and inevitably the answer came out,
“Oh, it was a friend of mine,” you know, which obviously starts off the wrong
impression in the minds of the jury, and it sort of went downhill from there, really… I
mean, looking back now, he was obviously trying to paint a picture of somebody
who, you know, had a sort of impeccable sexual history and that had only ever slept
with one person and was in a relationship etc., I had never been told about that, they
didn’t know that I’d already slept with somebody else. The look on the barrister’s
face when I answered was absolute horror. And, yeah, they’d never asked the
question. Unfortunately for me … and yet I never met the solicitor preparing the case.
(STAR, Service User Interview 20, Acquittal)
It was not part of this project to observe rape trials, but all studies of adversarial systems
re p o rt that cross-examination is uniformly described by complainants as savage and
g ruelling (see, for example, Department of Women, 1996; Lees, 2002), focusing on
whatever aspect might undermine the credibility of the complainant. The extent to which
defence barristers are pre p a red to deploy outmoded gender stereotypes and legacies of
victim/woman blame has also been documented (Temkin, 2000b). Within this, the spectre
of ‘real rape’ recurs and is reproduced. These factors play a part in the acquittal rate of
more than 50 per cent in rape trials.
Reflections
Two American law professors (Schulhofer, 1998; Taslitz, 1999) conclude that how rape
narratives are, and are not, constructed as ‘believable’ in the courtroom has more influence
than statute changes and even evidential requirements. Moreover, it is the failure to address
this that accounts for the failure of legal reform.
A gap or a chasm? Attrition in reported rape cases
76

The inability to engage with cultural narratives and macho adversarialism explains
rape law reform’s failure. These primary mechanisms by which rape jurors determine
credibility are unchanged. Consequently, unjustified acquittals mount.
(Taslitz, 1999, pp154-5)
The most powerful and persistent cultural narratives are the notion of ‘real rape’ and that
women frequently lie about rape.
Such a perception outweighs, in many cases, evidence of resistance and visible
injuries sustained by the victim, or even gross inconsistencies, changed stories and
‘lies’ in the defendant’s account. (Jordan, 2001b, p72)
Summary
●Twelve per cent of all reported cases, or 14 per cent of those where the outcome
is known, reached the trial stage.
●A proportion of these did not proceed to trial due to late withdrawal or the case
being discontinued at court.
●Around half of all convictions were due to guilty pleas rather than verdicts.
●In cases where a full trial took place an acquittal was more likely to be the
outcome than a conviction and the acquittal rate in trials involving adults was
twice as high as in those involving under-16s.
77
Understanding attrition: the research findings

A gap or a chasm? Attrition in reported rape cases
78

5. Challenges, dilemmas and recommendations
The combined analysis of the St Mary ’s database and the prospective attrition re s e a rc h
p rovide a more detailed picture of the processes underlying attrition than has pre v i o u s l y
been possible. Some of the findings support the claim of police officers that they are faced
with a complex mix of cases, some of which are false allegations and some of which do not
involve a sexual offence. Even within these categories, however, one cannot take all police
designations at face value, since a proportion are contested by some complainants, and
re g a rded with scepticism by re s e a rchers. It is, however, clear that around six per cent
( n = 1 6 0 )4 7 of initial re p o rts are either false complaints or cases where early investigation
reveals no evidence of an assault. These are cases where ‘no crime’ has taken place, and
the data suggest that the majority of such cases were identified at an early point.
There is also a proportion of cases where contact with the CJS is tentative, or not chosen by
the complainant themselves. The St Mary ’s police liaison officer reflected on these
complexities early in the project.
When people ring the police or involve the police, it’s at that critical moment, when
something has happened or is happening and they want help. I’m not always
convinced that the help they want is a rape investigation. The help they want might
be “Just get me out of this situation, you know, get me home, get me safe”, and that’s
why they ring the police. Very often people haven’t fully thought through the
consequences of what ringing the police actually means. And so perhaps what does
happen is that people want a service at that time, and that then gets them into the
sausage machine of the investigation and perhaps they never wanted to be in that
sausage machine anyway.And it’s very difficult to get out of it, other than at some
point saying “Well, look, hang on, I’m withdrawing from this.”
(St Mary’s, Police Liaison, January 2001)
Whilst acknowledging that this may well be the case for some, this study’s data suggest that
the ambivalence many complainants feel relates to the processes of investigation and
p rosecution: the practical challenges of a forensic examination; making a statement; the
prospect of courtroom testimony; and their experience and perception of treatment at each
stage by the professionals involved. Decisions to report and to continue with the case are
79
47. Included here are all no evidence of assault cases (n=83), and those false allegations deemed probable and
possible (n=77).

made when an interest in justice and protection for themselves and others outweighs the
potential costs, and where victims/survivors feel supported in this decision. This is a fine
balance, which CJS professionals and support services, through their actions and inactions,
can tip one way or the other.
The belief amongst some CJS professionals that many complaints are false, that victims are to
blame for ‘risk-taking’, places unreasonable re q u i rements on complainants to demonstrate
that that they are ‘real’ and deserving victims. Aw a re of these implicit standards, many
v i c t i m s / s u rvivors adjust their initial account in order to appear believable (see also, Jord a n ,
2001b). The danger is that these omissions and/or evasions are understood through the lens
of a false complaint at worst, and as creating evidential problems at best. Both re s p o n d e n t s
and interviewees stressed the importance of encountering a climate of belief at the initial
stages, which enabled them to be more relaxed, and in turn created a space in which tru s t
could develop. In such a context police officers would be in a stronger position to encourage
h o n e s t y, whilst acknowledging that there may be aspects of what took place which
v i c t i m s / s u rvivors feel uneasy about disclosing. There were a few examples of this occurr i n g
in the study sample, including one case where the complainant was involved in pro s t i t u t i o n
and had a drugs problem. The initial response of the police officers to this information was
respectful and open – making clear that they had to tell the CPS, but did not think either fact
was relevant to the context of her assault. This case resulted in a guilty plea.
Part of the attrition process is a feedback loop where, given that 80 per cent of cases drop
out or are lost in the earliest stages (see Table 4.2), everyone – including complainants – is
second guessing what the likely outcome at the final stage will be. Given this re p o rt ’s
finding that trials involving adult victim/witnesses are more likely to result in an acquittal,
these concerns are anything but groundless. Whilst the police, and to some extent the CPS,
have begun to address ways in which they can improve their response to rape, these efforts
will not have a major impact unless the complex interactions between each stage of the
process are understood.
An unexpected finding, which deserves more attention than can be given here (see also
F i n n e y, 2004), was that alcohol was involved in over one-third of cases re p o rted to the
police (34%, n=891). Drugs were suspected in over one in ten (12%, n=312): in a
significant pro p o rtion of these (36%, n=111), the presence of drugs was either not
confirmed by toxicology tests or limited to recreational use, often combined with alcohol.48
A gap or a chasm? Attrition in reported rape cases
80
48. Higher rates of involvement of alcohol and drugs were revealed in the St Mary’s data, where there was wider
access to accurate details in a greater number of cases. Here, alcohol was involved in over half of cases
reported to the police where a forensic examination was carried out (56%, n=595), while drugs were suspected
in around one-fifth (18%, n=198), with recreational drug use apparent in 40 per cent (n=79) of these.

At minimum alcohol is three times more associated with rape than drugs, and probably closer
to four times, but little, if any, work has addressed this correlation. Two aspects that deserv e
closer examination are the additional correlation of alcohol with rapes by re c e n t
acquaintances (55%, n=162), and that the presence of alcohol increased the levels of
re c o rded injuries (51%, n=450). In addition, analysis of statements and data from serv i c e
users suggest there is a group of pre d a t o ry men who target women when they are drunk, so
d runk in a number of cases that their capacity to consent had to be impaired. There were
indications of a range of potential targeting strategies involved, which deserve further study.
For example, two of the interviewees were women in their fifties, re t u rning home from a night
out with friends. In both instances a man presented himself as tru s t w o rt h y, and in one case he
o ff e red to ensure she got home safely. She was subsequently raped in her own home. Lawyer
and author Andrew Vasschs terms this pattern ‘targeted predation’, and comments:
Rapists see two different forms of weakness: (1) she’s drunk and (2) she’s drunk.
Meaning, she’s physically impaired because of the alcohol, but also no one is going
to believe her because they disapprove of her drinking.
(Personal communication, September 2003)
Attrition in context
Attrition is a process within which institutional rules, previous and predicted experience and
gendered expectations of behaviour interact. Police, CPS, judges (and presumably juries) all
work with ‘gender schemes’ which set the boundaries of what is considered appro p r i a t e
and inappropriate behaviour for men and women. Jeff n e r ’s (2000) re s e a rch with young
people in Sweden, suggests that the same factors provide men with more of what she terms
‘space for action’, whilst simultaneously narrowing and restricting the space available for
women. In relation to rape these factors include: alcohol; previous consensual sex;
reputation; and mental health issues. Where these factors are present – whether for men or
women – sex without consent is normalised, as being something other than ‘rape’. She
concludes that despite young people in Sweden defining rape on a general philosophical
level as any sex that happens after a woman has said no, when invited to explore this in
relation to specific contexts:
Only a sober young woman, who does not have a bad reputation, who has not
behaved sexually provocatively and who has said no in the right way can be raped,
and only by a young man who is sober and ‘deviant’ and with whom she is not in
love. (Jeffner, 2000)
81
Challenges, dilemmas and recommendations

The malleability of rape, and the meaning of consent, were also evident in research with
young people in the UK and Ireland (Burton et al., 1998; Regan and Kelly, 2001),
illustrating that notions of ‘real rape’ not only reinforce the normative stereotype referred to
earlier, but also designate conditions under which non-consensual sex is ‘not really rape’.
It is the awareness of these realities, and the fact that their credibility will be questioned in
court (and, potentially, their sexual history raised), that was the most potent reason behind
w o m e n ’s decisions to withdraw their complaint. Ve ry early decisions to not complete the
initial elements of an investigation are influenced by careless and insensitive treatment by
police officers, as well as an existing perception of not being able to face any potential
c o u rt process. Examples were uncovered of miscommunications, where police officers, in
good faith4 9, honestly conveyed the unlikelihood of a conviction and the difficulty of the
court experience. These messages, whilst accurate, especially when given at a very early
point (for example during or after taking an initial complaint or statement), are perceived by
the victim as indicating disbelief at worst and as discouragement from engaging in the legal
process at best.
There was little evidence of inter-agency links between police and CPS that sought to build
cases; to find evidence that might go to the credit and believability of the complainant.
Rather, all too often, the focus appeared to be on what was discrediting. There is evidence
f rom other jurisdictions that a focus on building cases, especially when applied to those
where consent is the likely defence, can increase the level of prosecutions and convictions.50
For generations, sexual assault or rape was and continues to be defined in our minds
as a violent and unprovoked attack by a stranger on defenceless, vulnerable, and
unsuspecting victim. Our understanding of this crime … influences (even without
conscious intent) our attempts to investigate and evaluate sexual assaults … We
found officers were looking for serious injury, weapons and immediate cries for help
to establish the elements of force or fear and to judge the ‘believability’ of the victim
... and missing the opportunity to collect, describe, or emphasize the very necessary
evidence that would support the lack of consent (Archambault and Lindsay, 2001).
In terms of understanding early withdrawal, there is no doubt that there has been a shift in
policing to victim care/satisfaction and the impact is palpable in the more positive
assessments from our questionnaire respondents, especially with respect to the earliest
stages of intervention, compared to studies conducted in the late 1980s and early 1990s.
A gap or a chasm? Attrition in reported rape cases
82
49. Good faith is presumed here, although previous research in the US has suggested that some police officers use
this as a deliberate strategy to encourage withdrawal (Kersetter, 1990).
50. See for example, Archambault and Faugno, 2001; Archambault and Lindsay, 2001 with respect to San Diego.

Far less attention has been given to improving investigation51, and it is at this point in the
p rocess that considerably more dissatisfaction is expressed (see also Jordan, 2001b for
similar findings with respect to New Zealand). The lack of knowledge transfer within the
domestic violence field, especially with respect to enhanced evidence-gathering and
supporting complainants to remain in the system, is also worth noting. This study confirmed
that there are two major points of intense disappointment and loss of confidence that
contribute to the decision of victims/survivors to withdraw their trust and co-operation:
where the early investigative processes explicitly or implicitly suggest that women are not
telling the truth and/or that the case is not worth pursuing; and later when a failure to keep
them informed, to show that the CJS cares and/or ambivalence in the investigative officers
leads to a loss of faith.
Perhaps the greatest challenge to the CJS is to re-think what rape is, and from this to then
develop new understandings of how to approach, investigate and prosecute it.
Key findings
●There are false allegations, and possibly slightly more than some researchers and
support agencies have suggested. However, at maximum they constitute nine per
cent and probably closer to three per cent of all re p o rted cases. An over-
estimation of the scale by police officers and prosecutors feeds into a culture of
scepticism, which in turn leads to poor communication and loss of confidence
between complainants and the police.
●In areas with integrated SARCs fewer complainants declined to complete the
initial process, and two SARC areas had slightly higher conviction rates.
●When given at an early point, honest assessments by police officers of the
likelihood of conviction and the difficulties of a prosecution are interpreted by
complainants as discouragement to continue with the case.
●There was little evidence of attempts to build cases, and some evidence of poor
investigation and understanding of the law, with clear emphasis in some cases
entirely on what was discrediting.
●Whilst there are many reasons why victims/survivors withdraw their co-operation,
fear of the trial process and discouragement by the police featured strongly.
●Police categorisation of cases is internally inconsistent, making monitoring and
evaluation extremely problematic.
83
Challenges, dilemmas and recommendations
51. Project Sapphire in the Metropolitan police is beginning to move in this direction, but in the first years prioritised
ensuring consistent and sensitive responses and developing SARCs.

●Few cases involved women with learning difficulties or women with mental health
problems (see also Harris and Grace, 1999).
Intellectual disability and psychiatric instability ... tend to be viewed as diminishing
the victim’s credibility, rather than enhancing her vulnerability.
(Jordan, 2001b, p349)
●None of the gang rapes with more than three assailants resulted in convictions.
●Women whose cases went to trial expressed disquiet and a strong sense of
injustice that they had no contact with the prosecution barrister52, and the majority
of those where the outcome was an acquittal complained of weak pro s e c u t i o n
advocacy in the courtroom.
●All stages of the attrition process appear more marked for young women,
particularly those aged 16 to 25, who are highly represented in cases involving
false allegations, early and late withdrawal of complaints and acquittals. This
re q u i res strategies to address the way this age group is responded to and
supported through the CJS.
●Alcohol consumption was present in a much larger number of cases than drugs –
either voluntarily consumed or administered to facilitate rape. One would not,
h o w e v e r, know this from recent media coverage or police aware n e s s - r a i s i n g
campaigns. The ways in which consumption of alcohol contributes to all points of
attrition deserves more detailed study.
Good practice is possible
Given that the focus of this study has been attrition, most attention has been on the factors
that account for cases being lost or dropped. The evaluation of SARCs (see Lovett et al.,
2004) also contains lessons about elements of practice that make a positive difference.
Poor practice erodes confidence, not just that of complainants with the CJS, but also that of
agencies with each other and of the general public. It should be axiomatic that a
complainant who is treated in a respectful and compassionate manner, by skilled
professionals, is not only less likely to withdraw support from the legal process at a later
date, but also more likely to feel able to tell what she knows, and include at the outset any
issues which could be seen as discrediting. These are gestures of trust, and unless those
A gap or a chasm? Attrition in reported rape cases
84
52. Recent Bar Council guidelines do now allow for barristers to introduce themselves, but this does not appear to
have become routine practice.

responding demonstrate that they are trustworthy, they will not be forthcoming. The police
are often the first official point of contact and as such set the tone for subsequent processes.
A negative police response can compound the trauma suffered by a rape victim,
making it less likely that she will decide to proceed with legal action and a strong
possibility that her experience will deter others from even making the initial police
contact in similar circumstances. If for no other reason, it is in the overall interests of
law enforcement for the police to act in ways that are consistent with promoting the
victim’s emotional well-being. (Jordan, 2001b, p75)
The experience of feeling believed was particularly vital ... Feeling believed was of
significant assistance both in dealing with the criminal justice system and also in
coping with the trauma of the rape itself. (Temkin, 1997, p519)
The research data contain many examples of initial police responses being thoughtful and
respectful (70% of questionnaire respondents were satisfied with the initial police response),
but the investigative process was experienced much more negatively. The following positive
example reveals the significance for complainants of being kept informed and of continued
contact.
The police rang me and kept me in touch and said, “We’ve not heard anything yet,
how are you keeping?” and X [female police officer] she rang every fortnight, just to
see how I was. I explained my fears to her about actually going to trial and him
getting away with it ... I said, “X, I love to hear from you in some ways, because it’s
nice to know that you’re still bothered and you still care.”
(St Mary’s, Service User Interview 1, Acquittal)
C o m p a re this with the far greater number of questionnaire respondents and interv i e w e e s
(n=23) who had the opposite experience of communication from the police.
My argument is with the police. And the way they’ve treated me ... If they’d have
kept me informed then maybe I wouldn’t have been as annoyed with them … And
they probably might have made me feel better if I still thought that they cared.
(St Mary’s, Service User Interview 3, Undetected)
And then they kind of just disappeared for ages, I mean, about three weeks after that
I called X [SOLO officer] and she said she had nothing to do with it anymore but
she’d try and get them to ring me. And they didn’t. And nobody was telling me
85
Challenges, dilemmas and recommendations

anything, I didn’t speak to them for like four weeks before they came to work, and
then I don’t think I ever spoke to them again after that ... That really was it for ages
... I was always really scared of phoning, because when I phoned I would sort of say
“Can I speak to so-and-so please?” and it would be like “Oh, there’s nobody in, and
what’s it about?” and I wouldn’t really know what to say … It just seemed to me to
be all wrong, that they were a service provider and I was almost having to, I don’t
know … I almost felt like I was putting them out, because every time I spoke to them
they were telling me how busy they were.
(REACH, Service User Interview 8, Undetected)
I am a bit annoyed because the only way I have found out about the courts is
through the STAR helpline. The police have not been in touch, only once has
someone come to see me and that was to take another statement for the CPS about
how I have been after the attack. I thank STAR for the case tracking – at least I know
when the court dates are and what is happening. (STAR, Q1, 2085, Conviction)
I think the police handled my case very badly, at first when they found out about the
rape they were very eager for me to press charges. Then after a few weeks they
didn’t seem at all interested. When the case went to court the DC didn’t even turn up
for the trial. I was furious the police did not telephone me to tell me the outcome. This
was very important to me as I needed to obtain an emergency injunction against my
attacker. Because the police failed to contact me this injunction was delayed for two
days, which was very upsetting and frightening for me.
(Comparison 3, Q2, 4016, Detected no proceedings)
Some of the participants were insightful about why these problems persist, and how they
could be overcome.
The crime of rape is still very much a difficult process to go through, the police are
still in the dark ages and you are still very much made to feel that somehow you are
to blame, that you could have prevented it. The police really need to look into the
training of officers and how as a unit of trust they can improve on this. It’s hard
enough to admit to being raped without someone doubting you. It’s no wonder so
many people don’t report it … STAR at least listens in a non-judgemental way, and
listens to you, never doubting. They are really understanding and helpful and just
having someone believe you can help you live again. Without STAR, I would have
completely fallen apart. (STAR, Q1, 2065, Status unknown)
A gap or a chasm? Attrition in reported rape cases
86

The comments about SARCs reveal that it is the tone and content of human interaction, the
sense of being respected and treated as someone who is more than ‘just a victim’ which
was not only appreciated, but enabled women to cope with lengthy and difficult demands,
such as a forensic examination.
And the respect I was shown! I didn’t respect myself, you see, so for somebody to
show you that much respect, and kindness, you start thinking, “I’m not a bad person,
I’m not this dirty person, I thought I was, tainted, violated. They’re treating me as
normal.” Something terrible has happened but they’re treating me normally! And you
get a little bit – you feel confident, and more in control of things.
(St Mary’s, Service User Interview 11, Acquittal)
What complainants value and need
The information from questionnaires and interviews about aspects of current serv i c e
provision that were strongly valued has been summarised here, along with suggestions on
how responses to reported rape could be improved.
●Availability of female police officers and forensic examiners.
●A private, calming environment.
●A culture of belief and support.
●Being treated as a whole person, and “as a person not a case”.
●Respect – strongly distinguished from being patronised, or condescended to.
●Being ‘in control’ of the forensic examination.
●Technological innovation to make the statement taking process less protracted, i.e.
the police officer not having to write everything down in longhand, and to be in
the words of the complainant.
●Access to a simple card or flyer with a phone number where support and
information would be easily accessible, including out of hours.
●Awareness of the meaning of information, and the time it is most appropriate to
give it, especially where police officers are providing a realistic assessment of the
prospects of prosecution and conviction.
●Services keeping proactive contact and being kept informed about case progress
(see Lovett et al., 2004).
●Continuity of police officers, and meeting investigative officers in person.
●Being able to re-read statements the week before trial or at least not in the hour
before giving evidence.
●Contact with prosecution barristers.
87
Challenges, dilemmas and recommendations

●P rosecution barristers being familiar with the facts of the case and court ro o m
advocacy that does ‘justice’ to the complainant’s account.
●Building cases, pursuing supporting evidence and presenting a believable
narrative that explains the behaviour of the complainant.
Many of these elements constitute what has been termed ‘procedural justice’, and this
research confirms the significance of a sense of procedural fairness to victims’ perceptions
of the CJS. The summary below, from a study of women’s satisfaction with how domestic
violence cases were dealt with in Canberra, Australia echoes the perceptions of those in this
study who reported rape.
Victim/witnesses seek n o t i f i c a t i o n on a range of things – dates, decision-making
points, decisions (charge, prosecution, bail, court outcome), and identities of key
participants. Linked to notification is the need for information about the system, how it
works, the role of the prosecution and of the victim/witness … In the main,
victim/witnesses also seek opportunities for participation through making the initial
complaint to police, the provision of a full statement, offering (or not) background,
consultation on their views and opinions, input at bail, input at post conviction
through the Victim Impact Statement and (occasionally) the Pre-Sentence Report …
Respectful acknowledgement relates to the inter-personal communication between the
victim/witness and justice practitioner.This idea stresses the style of the interaction –
for example, the extent to which the communication is empathetic and builds rapport.
(Holder and Mayo, 2003, original emphasis, p18)
The fact that these may be more, or at least as, important as eventual outcomes is evidenced
by the responses of those women whose cases resulted in acquittals, but who were
s u p p o rted throughout by the Support Worker at St Mary ’s. Whilst all were disappointed
with the outcome, none re g r etted pursuing the case. This contrasts sharply with re s e a rc h
participants whose cases were dropped or lost at an earlier point, and who encountered
limited support: many of them were profoundly dissatisfied with the responses they
e n c o u n t e red, re p o rting that if they were ever sexually assaulted again they would not
report, and that they would discourage any friend from doing so. The research interviews
also highlighted the extent to which women are aware of the low conviction rate. If the
attrition rate continues to increase, and if the layers of the criminal justice system fail to
a d d ress their role in it, erosion of confidence will expand, as will the numbers of those
whose designation as having made a ‘previous allegation’ means any subsequent report is
treated with scepticism.
A gap or a chasm? Attrition in reported rape cases
88

The attrition process is currently reproducing itself and in the process is creating a justice
gap that is experienced by many rape complainants as more like a chasm. From the
perspective of practitioners and policy-makers, however, the authors explication of the layers
involved in attrition suggests that this distance is made up of a series of distinct, whilst
connected, gaps. Addressing each of these systematically, rather than attempting to bridge
the chasm in one leap, is more likely to produce improvements. The re c o m m e n d a t i o n s
below are offered as initial steps to reverse the trend of falling conviction rates.
Recommendations
1. The most important recommendation from this study in terms of the investigation of
rape complaints is that a shift occurs from an exercise in scepticism focusing on
discreditability to enhanced evidence-gathering and case-building. Within this three
working groups are suggested to develop guidelines and resources to support this
change of emphasis.
●To rethink the entire process of rape investigation and prosecution through the
lens of known offenders and consent defences. This group should include police,
CPS, barristers, forensic scientists, academics and the voluntary sector.
●To develop, drawing on international experience and the disability sector, policy
and advice on investigating and prosecuting cases where complainants have
learning disabilities and/or mental health problems.
●The development of new rape examination kits, which can be used flexibly
depending on the facts of the case and likely defence. Elements should include
a g reed national forms for forensic examiners, which should accompany any
samples sent to the forensic science service (see Kelly, 2003b for more detailed
discussion).
2. The possibility of inter-agency work between police and specialist support agencies
should be further explored with respect to providing support during statement-taking and
providing information on case status (see also Lovett et al., 2004).
3. Home Office counting rules for rape and sexual offences need to be revisited in light of
the findings from this re s e a r ch: whether cases where there is no evidence of assault
should remain needs to be addressed, and mechanisms should be developed to correct
inconsistent classification.
89
Challenges, dilemmas and recommendations

4. Reflection on, and guidance issued, within the police about the timing and content of
providing complainants with information about the legal process and the likelihood of a
conviction. At a number of points, from the earliest contact on, this was perceived by
victims/witnesses as discouragement from proceeding with the case.
5. Development in all areas, through SARCs and/or Rape Crisis Centres, of pro a c t i v e
follow-up of all complainants reporting rape to the police. Such services would add the
most value if they combined practical support, advocacy and case tracking (see also
Lovett et al., 2004).
6. Piloting of early case conferences between police, CPS and counsel, to explore
potential evidential weaknesses, and whether these might be addressed thro u g h
additional evidence, expert testimony, research findings or courtroom advocacy.
7. An accredited training course for prosecution barristers, and the development of some
form of sanction where briefs are returned or counsel changed just before trial.
8. Mechanisms within CPS to both monitor and evaluate courtroom prosecution advocacy,
and to learn lessons from convictions and acquittals, and especially guilty pleas.
9. I n c reased recognition of the significance of alcohol in rape and sexual assault,
including within the govern m e n t ’s proposed alcohol policy. This should also include
further exploration of the extent to which men target unknown women who are drinking,
and the strategies they use to make initial contact.
1 0 . Last, but certainly not least, in light of the Sex Offences Act, development of a public
education campaign on the meaning of consent and realities of rape, alongside more
detailed training for CJS personnel, explicitly designed to expand the concept of ‘real
rape’.
A gap or a chasm? Attrition in reported rape cases
90

Appendices
91

A gap or a chasm? Attrition in reported rape cases
92
Appendix 1

93
Appendices

A gap or a chasm? Attrition in reported rape cases
94

95

Appendix 2
A gap or a chasm? Attrition in reported rape cases
96

97
Appendices
Appendix 3: Research tools and data sources
St Mary’s historic database
St Mary’s has their own database containing details of all cases who have been referred to
the Centre since it opened in late 1986. The original database ran on a FoxPro system,
which is now obsolete. In 2003 this was replaced with a new Access database, designed
by an external consultant, and all original data were imported. The new database contains
a similar range of fields to those in the case-tracking database, which are organised around
the following four main areas:
●Attendance and personal details – SARC number, referral type, attendance date
and time, age, sex, ethnic origin, disability, relationship profile, employment
status and area of residence.
●Assault details – date and time it occurred, type of assault, area it occurred in,
type of location, number of assailants, relationship between assailant and victim,
and use of weapon and force.
●Police details – date of report, crime number, Force Wide Incident number, type
of offence re c o rded, statement details, crime classification, and court and
sentencing details.
●Forensic details – examination date and time, nature of incident, and details on
the presence of alcohol, drugs and injuries.
Case-tracking database
The case-tracking database was created specifically for this project using Access software to
include details on all cases being referred to the three SARCs and being reported in the
Comparison areas. Every individual entered on the database has a record consisting of a
header and a series of linked forms. The header contains: the unique re f e rence number;
research site; referral type; and details of participation in the evaluation. The linked forms
and fields within them focus on:
●Details of the assault – date and time it occurred, location, type and additional
details.
●Details of the victim – age, sex, ethnic origin, disability, relationship profile and
employment status.
HORS293.2.2.qxd 2/14/05 10:36 AM Page 97

●Details of the perpetrator(s) – age, sex, ethnic origin, disability, employment status,
relationship to the victim, length of acquaintance and whether a weapon was used.
●Forensic examination – date and time of, examiner type, whether injuries were
sustained, whether drugs/alcohol were involved and which samples were taken.
●Service use – type of service accessed at each site.
●Police report – date and time re p o rted, who re p o rted, whether a statement was
taken, whether a suspect was identified, arrested and held in custody, CPS advice,
reasons for cases not proceeding and additional details on case pro g re s s .
●Legal process – final police classification, whether the case went to court, plea,
trial outcome and sentence.
The level and type of data re c o rded, as well as the re c o rding systems in place, varied at
each site. Consequently, diff e rent processes for obtaining the information re q u i red for the
case-tracking database were employed at the six sites. STAR has its own database that
contains many fields comparable to those in the case-tracking database. STAR has a case
tracker who informs service users of case pro g ress, there f o re, providing the service users
have re p o rted the assault, the database contains detailed information on case outcomes.
Copies of the STAR database were anonymised by removing the names and contact details
of the service users and periodically downloaded for the re s e a rch team. In the thre e
Comparison areas all data were supplied by the police. At St Mary ’s and REACH, there were
v a rying degrees of reliance on the police for elements of data, such as details of the police
re p o rt, perpetrator details and legal outcomes. This information was obtained by distributing
two pro formas to police officers, one within a month of the initial police re p o rt and the
second up to a year later. In a number of cases multiple follow-ups were needed, both to
recover outstanding pro formas and to obtain complete information on cases that had not
been completed at the time of re t u rning the second pro forma. The remaining details on
victims, forensic examinations and service use were provided by the SARCs completing client
details sheets. The database has been used mainly for quantitative analysis. However, the
f ree text fields containing qualitative data on reasons for cases not proceeding and
additional information on case pro g ress have been subjected to content analysis, which has
been used as a supplementary data element in the discussion of re s e a rch findings relating to
the series of Attrition points.
Service user questionnaires
All those included on the case-tracking database were potential participants in this element
of the research. Three phased questionnaires were distributed to those service users who
a g reed to take part at the first, fifth and twelfth month after their initial contact with the
A gap or a chasm? Attrition in reported rape cases
98

SARCs or report to the police in the Comparison areas. The questionnaires were adapted to
reflect the processes and available services at each site, whilst retaining a common core to
ensure comparability. They were also designed to be applicable to those who had reported
to the police and those who had not. The questionnaires were referenced using the unique
re f e rence number to enable them to be linked to cases entered on the case-tracking
database. The areas covered in each of the questionnaires were:
Questionnaire 1
●About you – age, ethnic origin, disability, relationship profile, employment status.
●After the assault – did you tell anyone about the assault before reporting to the
police and what was their response.
●Reporting to the police – did you report to the police, reasons for reporting, how
long after assault was the report made, how was the report made, your state of
mind when reporting, how would you rate the response of first police officer you
spoke to, satisfaction with the police at this stage, any suggestions for improving
the initial police response.
●Crisis worker (St Mary’s only) – how you understood their role, sex of the crisis
w o r k e r, did this make a diff e rence, how would you rate their re s p o n s e ,
satisfaction with crisis worker, how much information did they give you about the
next stage of the process.
●Specialist police officer (REACH, STAR and comparison areas only) – did you see
a specialist officer, when did you see one, how you understood their role, sex of
the off i c e r, did this make a diff e rence, how would you rate their re s p o n s e ,
satisfaction with specialist officer, any suggestions for improving the response of
specialist officers.
●Forensic medical examination – did you have one, did you have to wait to see an
examiner, feelings before the examination, sex of the examiner, did this make a
difference, was the process of the examination explained, how you experienced
the examination, how would you rate the response of the examiner, satisfaction
with examiner, any suggestions for improving examinations.
●After the forensic medical examination – contact with specialist off i c e r / c r i s i s
worker after the examination, how would you rate their response.
●Making a statement to the police – did you make a statement, when did you
make one, where did you make one, were you given a choice about the timing
and location of the statement taking, feelings before making a statement, sex of
the officers present, did this make a difference, how you experienced making a
statement, how would you rate the response of the officers taking the statement,
99
Appendices

were you given information about the next stage of the process, have you had
any contact with the police since the statement, any suggestions for improving the
statement-taking process.
●Use of services (SARCs only) – which services have you used so far, how would
you rate the response of the service provider, any suggestions for improving the
service, were there any services you needed that were not provided.
●Your decisions and the future – are you intending to pursue the criminal case, if
not, why, how are you coping at the moment, what are you finding difficult.
●About the questionnaire – are there any additional questions we should have included,
how have you found completing the questionnaire, any additional comments.
Questionnaire 2
●About you – how are you coping at the moment, what are you finding difficult,
are there any areas where you need support.
●Take-up of services – have you been in contact with the SARC since completing
initial questionnaire, if yes, has this been helpful, which services have you
accessed, how would you rate the response of the service pro v i d e r, were you
re f e rred to any other services by the SARC/police, how would you rate their
response, have you received any support from family or friends, who has
provided the best support so far.
●Contact with police – have the police been in contact since completing the initial
q u e s t i o n n a i re, what about, how would you rate their response, how well have you been
kept informed about their case, is there anything more the police could have done.
●Criminal case – do you know the current status of the case, if yes, what is it, what
are your views on the outcome.
●About the questionnaire – are there any additional questions we should have included,
how have you found completing the questionnaire, any additional comments.
Questionnaire 3
●About you – how are you coping at the moment, what have you found most
difficult over the past year, are there any areas where you need support.
●Take-up of services – have you been in contact with the SARC since completing the
second questionnaire, has this been helpful, have you been in contact with any
other services since completing the second questionnaire, what was their re s p o n s e ,
how would you describe your overall experience of support services over the past
y e a r, could anything have been done diff e rently to support you more .
A gap or a chasm? Attrition in reported rape cases
100

●Contact with police – have the police been in contact since completing the second
questionnaire, what about, how would you rate their overall response since you
reported the assault, could the police have done anything differently to support
you more.
●Criminal case – has your case been heard in court, if no, what is the current status,
if yes, what was the outcome, did you meet the prosecution barr i s t e r, did you give
evidence, what was your experience of giving evidence, what are your views on
the outcome, how would you describe the overall experience of going to court, is
t h e re anything that might have made the experience of going to court easier.
●About the questionnaire – are there any additional questions we should have
included, what has it been like taking part in this re s e a rch, any additional comments.
Service user interviews
All service users participating in the questionnaire element of the research were also invited
to take part in an interview, either face-to-face or over the telephone. This was to gain more
detailed qualitative information on their experiences of service use and the criminal justice
process. For face-to-face interviews participants were given a choice of locations, including
the SARC, their own home or counselling premises. A letter explaining what taking part in
the interview would involve and a return form were distributed with the second of the three
s e r vice user questionnaires. The interviews were semi-stru c t u red and slightly adapted in
o rder to be relevant to service users at the diff e rent sites. Interview transcripts were
anonymised and referenced using the unique reference number, so they could be linked to
completed questionnaires and the case-tracking database. All interviews covered the
following broad themes.
●Decision-making about reporting to the police.
●Experience of initial contact with the police.
●Experience of contact with the SARC.
●Response of crisis worker/specialist police officer.
●Experience of giving a statement to the police.
●Follow-up contact/support from the SARC, police and elsewhere.
●Decision-making about the legal process.
●Current status of the case.
●How they are feeling/coping.
●Comments on the process so far.
●What victims of rape/sexual assault need. How did their experience compare with this.
●How the process could be improved.
101
Appendices

Witness statements
A sample of witness statements made by victims to the police in the SARC areas was
obtained. It was not possible to acquire a random sample of statements for two reasons.
Firstly, to avoid disclosure that might affect any proceeding case, the research team was
only allowed access to statements in completed cases. In addition, SARC managers
requested that consent was obtained from service users to access this material, which limited
cases to where the victim/survivor had opted into the evaluation and/or gave permission
for the statement to be disclosed. All statements received were anonymised, with the
complainant’s name removed and were coded with the unique identifier so they could be
linked to the other data sources.
Forensic medical reports
D i ff e rent protocols were negotiated at each SARC site for accessing forensic medical re p o rt s ,
l a rgely reflecting the varying systems for conducting examinations. At St Mary ’s, copies of all
f o rensic medical re p o rts are stored centrally. Here, all cases where a forensic examination
was conducted during the evaluation period were grouped according to examiner type
(female doctor or forensic nurse) and then according to final police classification (‘no crime’,
undetected, ‘detected no proceedings’ and detected). Random sampling was then conducted
within these groups and a roughly equal number of re p o rts selected from each. At REACH
and STAR, consent to access these documents was sought from service users directly in the
same way as for the witness statements (see above), there f o re the sample size was more
limited. Reports relating to participating service users at REACH were sought from the
individual forensic examiners, as they retain all documentation relating to the examinations
they have conducted. In the STAR area, where the SARC does not control the conduct of
f o rensic examinations, lengthy negotiations were held with Heathcall, the contractor used to
supply forensic examinations in West Yo r k s h i re, re g a rding access to forensic medical re p o rt s .
U n f o rt u n a t e l y, no re p o rts were forthcoming by the end of the evaluation from this area. All
f o rensic medical re p o rts received were anonymised, removing both the service user and
f o rensic examiner’s names, and re f e r enced with the unique identifier to enable cro s s -
re f e rencing with the case-tracking database and other data sourc e s .
Interviews with service providers
Phased semi-stru c t u r ed interviews were conducted with SARC staff (management,
counsellors, crisis workers, case tracker and support worker and forensic medical
examiners) and key players (police officers, Crown Prosecutors and Victim Support). At St
Mary’s interviews were conducted at three stages during the evaluation – at the beginning,
A gap or a chasm? Attrition in reported rape cases
102

at a mid-point and at the end. In the REACH and STAR areas interviews were conducted at
a mid-point and at the end of the evaluation. In the Comparison areas senior officers were
interviewed at a mid-point and at the end of the evaluation, whilst investigative officers and
police chaperones/specialist officers were interviewed once.
For SARC staff the interview questions covered the following themes.
●Role in and length of involvement with SARC.
●Contribution of SARCs.
●Any noticeable changes while they have been in post in the type of rape/sexual
assault cases they see.
●System for providing forensic examinations.
●Reasons why rape/sexual assault is under-reported.
●Reasons why complaints are withdrawn.
●Reasons why cases are lost later in the criminal justice process.
●Progress and impact of CRP-funded interventions.
●Inter-agency links.
The views of police and other key informants were sought on the following areas.
●Outline of their role.
●Connection with SARC.
●Contribution of SARC and services provided.
●Service offered by police to rape/sexual assault victims in area.
●System for providing forensic examinations.
●Reasons why rape/sexual assault is under-reported.
●Reasons why complaints are withdrawn.
●Reasons why cases are lost later in the criminal justice process.
●Progress and impact of CRP-funded interventions.
●Inter-agency links.
103
Appendices

Appendix 4
A gap or a chasm? Attrition in reported rape cases
104

Appendix 5: Attrition process –
case outcomes for all reported cases
Table 1: St Mary’s
n=1,071 N % Overall %
Police 843 79%
No crime 233 100% 22%
False allegation 82 35
No evidence of assault 54 23
Insufficient evidence 36 15
Victim withdrawal 35 15
Victim declined to complete initial process 20 9
Other 3 1
Reason unknown 3 1
Undetected 294 100% 27%
Victim withdrawal 74 25
Offender not identified 72 24
Insufficient evidence 59 20
Victim declined to complete initial process 39 13
False allegation 9 3
No prospect conviction 8 3
No evidence of assault 3 1
Reason unknown 30 10
Detected no proceedings 125 100% 12%
Victim withdrawal 64 51
Insufficient evidence 22 18
Victim declined to complete initial process 22 18
False allegation 3 2
No prospect conviction 3 2
Other 1 1
Reason unknown 10 8
Detected 21 100% 2%
Victim withdrawal 3 14
Insufficient evidence 3 14
105
Appendices

Victim declined to complete initial process 2 10
Not in public interest 1 5
Other 6 29
Reason unknown 6 29
Classification unknown 170 100% 16%
Victim withdrawal 4 2
Victim declined to complete initial process 3 2
No prospect conviction 1 1
Reason unknown 162 95
CPS 83 100% 7%
Decision unknown 31 37
Caution/final reprimand 2 2
Discontinued 8 10
Victim withdrawal 13 16
Pending trial 29 35
Trial 145 100% 14%
Trial to be rearranged, suspect fled 1 1
Victim withdrawal 10 7
Discontinued/withdrawn at court 8 5
Acquittal 49 34
Guilty plea 46 32
Not clear if guilty plea/conviction 9 6
Part conviction 5 3
Conviction 17 12
Total convictions 77 7%
A gap or a chasm? Attrition in reported rape cases
106

Table 2: REACH
n=380 N % Overall %
Police 321 84%
No crime 82 100% 22%
False allegation 46 56
No evidence of assault 14 17
Insufficient evidence 8 10
Victim withdrawal 7 9
Victim declined to complete initial process 5 6
Not in public interest 1 1
Reason unknown 1 1
Undetected 90 100% 24%
Insufficient evidence 35 39
Offender not identified 22 24
Victim withdrawal 10 11
Victim declined to complete initial process 9 10
False allegation 6 7
No prospect conviction 3 3
No evidence of assault 1 1
Reason unknown 4 4
Detected no proceedings 22 100% 6%
Insufficient evidence 9 41
Victim withdrawal 7 32
False allegation 2 9
Victim declined to complete initial process 2 9
Not in public interest 1 5
Reason unknown 1 5
Detected 13 100% 3%
Insufficient evidence 4 31
Victim withdrawal 4 31
Not in public interest 2 15
Reason unknown 3 23
Classification unknown 114 100% 30%
False allegation 3 3
Victim withdrawal 2 2
Insufficient evidence 2 2
No evidence of assault 1 1
107
Appendices

No prospect conviction 1 1
Victim declined to complete initial process 1 1
Other 1 1
Reason unknown 103 90
CPS 18 100% 5%
Decision unknown 3 17
Discontinued 5 28
Victim withdrawal 3 17
Pending trial 7 39
Trial 41 100% 11%
Discontinued/withdrawn at court 3 7
Acquittal 18 44
Guilty plea 12 29
Part conviction 3 7
Conviction 5 12
Total convictions 20 5%
A gap or a chasm? Attrition in reported rape cases
108

Table 3: STAR
n=837 N % Overall %
Police 653 78%
No crime 207 100% 25%
Victim declined to complete initial process 66 32
False allegation 29 14
Insufficient evidence 17 8
Victim withdrawal 10 5
Not in public interest 3 1
Offender not identified 3 1
No evidence of assault 2 1
No prospect conviction 1 0
Other 44 21
Reason unknown 32 16
Undetected 324 100% 39%
Insufficient evidence 112 35
Offender not identified 89 27
Victim declined to complete initial process 53 16
Victim withdrawal 12 4
False allegation 9 3
No prospect conviction 9 3
Not in public interest 5 2
Other 2 1
Reason unknown 33 10
Detected 90 100% 11%
Insufficient evidence 40 44
Victim withdrawal 20 22
Victim declined to complete initial process 13 14
No prospect conviction 7 8
Not in public interest 5 6
False allegation 1 1
Other 3 3
Reason unknown 1 1
Classification unknown 32 100% 4%
Insufficient evidence 6 19
Victim declined to complete initial process 6 19
Not in public interest 2 6
109
Appendices

Victim withdrawal 2 6
False allegation 1 3
No evidence of assault 1 3
No prospect conviction 1 3
Offender not identified 1 3
Other 1 3
Reason unknown 11 34
CPS 57 100% 7%
Suspect fled prior to PDH 1 2
Caution/final reprimand 5 9
Discontinued 25 44
Victim withdrawal 7 12
Pending trial 19 33
Trial 127 100% 15%
Victim withdrawal 5 4
Discontinued/withdrawn at court 8 6
Acquittal 36 28
Guilty plea 29 23
Not clear if guilty plea/conviction 8 6
Part conviction 3 2
Conviction 38 30
Total convictions 78 9%
A gap or a chasm? Attrition in reported rape cases
110

Table 4: Amalgamated Comparison area
n=355 N % Overall %
Police 299 84%
No crime 53 100% 15%
False allegation 18 34
Victim declined to complete initial process 12 23
Insufficient evidence 6 11
No evidence of assault 6 11
Victim withdrawal 6 11
Other 2 4
Reason unknown 3 6
Undetected 174 100% 49%
Offender not identified 51 29
Victim declined to complete initial process 42 24
Victim withdrawal 40 23
Insufficient evidence 23 13
False allegation 5 3
No prospect conviction 3 2
No evidence of assault 1 1
Other 2 1
Reason unknown 7 4
Detected no proceedings 41 100% 12%
Victim declined to complete initial process 20 49
Victim withdrawal 13 32
Insufficient evidence 3 7
False allegation 2 5
Other 1 2
Reason unknown 2 5
Detected 8 100% 2%
Victim withdrawal 5 8
Insufficient evidence 1 2
Offender not identified 1 2
Other 1 2
Classification unknown 23 100% 6%
Reason unknown 23 100
111
Appendices

CPS 47 100% 13%
Decision unknown 26 55
Caution/final reprimand 2 4
Discontinued 0 0
Victim withdrawal 2 4
Pending trial 17 36
Trial 9 100% 3%
Acquittal 1 11
Guilty plea 2 22
Conviction 6 67
Total convictions 8 2%
A gap or a chasm? Attrition in reported rape cases
112

Bibliography
American Medical Association (1995) Sexual Assault in America. AMA.
A rchambault, J. and Lindsay, S. (2001) ‘Responding to non-stranger sexual assault’. In M.
Reuland, C. Sole Brito and L. Carroll (eds.), Solving Crime and Disorder Problems Current
Issues, Police Strategies and Organizational Ta c t i c s . P E R F. [Available online at:
http://www.mysati.com/resources.htm].
Bachman, R. and Paternoster, R. (1993) A Contemporary Look at the Effects of Rape Law Reform:
How Far Have We Really Come? Journal of Criminal Law and Criminology, 84: 554-574.
Bachman, R. and Smith, P. (1994) The Adjudication of Rape Since Reforms: Examining the
Probability of Conviction and Incarceration at the National and Three State Levels. Criminal
Justice Policy Review, 6: 342-358.
B e rgen, R. (1995) ‘Surviving wife rape: How women define and cope with the violence’.
Violence Against Women, 1(2) : 117-138.
B re reton, D. (1993) ‘Rape prosecutions in Victoria’. In P. Easteal (ed.) Without Consent:
Confronting Adult Sexual Violence. Canberra: Australian Institute of Criminology.
Burton, S., Kitzinger, J., Kelly, L. and Regan, L. (1998) Young People’s Attitudes towards Violence,
Sex and Relationships: A Survey and Focus Group Study. Edinburgh: Zero Tolerance Trust.
Burton, S., Regan, L. and Kelly, L. (1998) Supporting Women and Challenging Men: Lessons from
the Domestic Violence Intervention Project. Bristol:Policy Press.
Caringella-MacDonald, S. (1985) ‘Sexual assault prosecution: An examination of model rape
legislation in Michigan’. In C. Schweber and C. Feinman (eds.), Criminal Justice Politics and
Women: The Aftermath of Legally Mandated Change (pp 65-82). New York: Haworth Press.
Chambers, G. and Millar A. (1983) Investigating Rape. Edinburgh: HMSO.
Commission on Women and the Criminal Justice System (2004) Women and the Criminal Justice
System. London: Fawcett Society.
113

Crown Prosecution Service (2000) The Code for Crown Prosecutors. London: CPS.
D e p a rtment of Women (1996) Heroines of Fortitude: the Experiences of Women in Court as
Victims of Sexual Assault. Sydney: New South Wales Department of Women.
Du Mont, J. and Myhr, T.(2000) ‘So few convictions: The role of client-related characteristics in
the legal processing of sexual assaults’. Violence Against Women, 6(10): 1109-1136.
Durston, G. (1998) ‘Cross-examination of rape complainants: ongoing tensions between
conflicting priorities in the criminal justice system’. Journal of Criminal Law, 62(1): 91-104.
Esteal, P. (1998) ‘Rape in marriage: has the licence lapsed?’ In P. Easteal (ed.), Balancing the
Scales: Rape, Law Reform and Australian Culture. Sydney: Federation Press.
Estrich, S. (1987) Real Rape: How the Legal System Victimizes Women Who Say No. Boston:
Harvard University Press.
Estrich, S. (1995) ‘Is it rape?’ In P. Searles and R. Berger (eds.), Rape and Society: Readings
on the Problem of Sexual Assault (pp 183-193). Boston: Westview Press.
Finney, A. (2004) Alcohol and Sexual Violence: Key Findings from the Research. Home Office
Findings 215. London: Home Office.
F i s h e r, B., Cullen, F. and Tu rn e r, M. (2000) The Sexual Victimization of College Wo m e n .
Washington DC: Department of Justice.
Frazier, P. and Haney, B. (1996) ‘Sexual assault cases in the legal system: police, prosecutor and
victim perspectives’. Law and Human Behaviour, 20: 607-628.
F r a z i e r, P. and Seales, L. (1997) ‘Acquaintance rape is real rape’. In M. Schwartz (ed.),
Researching Sexual Violence Against Women (pp 54-65). Newbury Park: Sage.
Frohmann, L. (1991) ‘Discrediting victims’ allegations of sexual assault: prosecutorial accounts
of case rejections’. Social Problems, 38(2): 213-226.
Grace, S., Lloyd, C. and Smith, L. (1992) Rape: From Recording to Conviction. Home Off i c e
Research & Planning Unit Paper No 71. London: Home Office.
A gap or a chasm? Attrition in reported rape cases
114

G reenfield, L. (1997) Sex Offences and Offenders: An Analysis of Date Rape and Sexual
Assault. US Department of Justice Bureau of Justice Statistics, NCJ-163392.
Gregory, J. and Lees, S. (1999) Policing Sexual Assault. London: Routledge.
Hagemann-White, C. (2001) ‘European re s e a r ch into the prevalence of violence against
women’. Violence Against Women, 7(7): 732-759.
Harris, J. and Grace, S. (1999) AQuestion of Evidence?: Investigating and Prosecuting Rape in
the 1990s. Home Office Research Study 196. London: Home Office.
HM Crown Prosecution Service Inspectorate and HM Inspectorate of Constabulary(2002) A Report on the
Joint Inspection into the Investigation and Prosecution of Cases involving Allegations of
Rape. London: HMCPSI.
Holder, R. and Mayo, N. (2003) What Do Women Want? Prosecuting Family Violence in the
ACT. Current Issues in Criminal Justice. 15(1): 5-25.
Home Office (2003) Home Office Counting Rules for Recorded Crime. London: Home Office.
Home Office (2002a) Narrowing the Justice Gap – Framework Document. London: Home Off i c e .
Home Office (2002b) Action Plan to implement the recommendations of the HMCPSI/HMIC
joint investigation into the investigation and prosecution of cases involving allegations of
rape. London: Home Office
Home Off i c e ( 1 9 8 3 ) Investigation of the Offences of Rape. Home Office Circ u l a r
25/1983.London: HMSO.
H o rn e y, J. and Spohn, C. (1991) ‘Rape law re f o rm and instrumental change in six urban
jurisdictions’. Law and Society Review, 25(1): 117-153.
Jamieson, L., Burman M., Gru n d y, S. and Dyer, F. (1998, unpublished) The ‘Attrition’ of Sexual
Offences in the Criminal Justice System: AReport of a Pilot Study Monitoring Cases from
First Report to the Police Final Outcome. Report submitted to the Scottish Office.
J e ff n e r, S. (2000) ‘ D i fferent space for action: The everyday meaning of young people’s
perception of rape’. Paper at ESS Faculty Seminar, University of North London, May.
115
References

Jewkes, R. and Abrahams, N. (2000) Violence Against Women in South Africa: Rape and Sexual
Coercion – A Review Study. Crime Prevention Research Resources Centre.
Johnson, H. and Sacco, V. (1995) ‘Researching violence against women: Statistics Canada’s
national study’. Canadian Journal of Criminology: Special Issue: Focus on the Vi o l e n c e
Against Women Survey, 37(3): 281-304.
Jones, A. (1980) Women who Kill. New York: Fawcett, Columbine.
Jordan, J. (1998) Reporting Rape: Women’s Experiences with the Police, Doctors and Support
Agencies. Wellington: Institute of Criminology, Victoria University.
Jordan, J. (2001a) ‘Worlds Apart? Women, rape and the police reporting process’. British
Journal of Criminology, 41: 679-706.
Jordan, J. (2001b) True ‘Lies’ and False ‘Truths’: Women, Rape and the Police. PhD thesis,
Victoria University of Wellington, New Zealand.
J o rdan, J. (2002) ‘Will any woman do?: Police, gender and rape victims’. Policing: A n
International Journal of Police Strategies & Management, 25(2), 319-344.
J o rdan, J. (2004) The Word of a Woman? Police, Rape and Belief. London: Palgrave
Macmillan.
Katz, S. and Mazur, M. (1979) Understanding the Rape Victim: Synthesis of Research Findings.
New York: John Wiley.
K e l l y, L. (2000) ‘Ending the silence, challenging the tolerance: developing community
responses in the prevention of domestic violence’. Keynote paper at Intern a t i o n a l
Conference on Violence in the Family: Plan of Action for the 21st Century, Nicosia, Cyprus,
November.
Kelly, L. (2002) AResearch Review on the Reporting, Investigation and Prosecution of Rape
Cases. London: HMCPSI.
Kelly, L. (2003a) Specialisation, Integration and Innovation: Review of Health Service Models
for the Provision of Care to Persons who have Suffered Sexual Vi o l e n c e . Final Report to
World Health Organisation.
A gap or a chasm? Attrition in reported rape cases
116

Kelly, L. (2003b) Good Practice in Medical Responses to Recently Reported Rape, Especially
Forensic Examinations. Briefing paper for the Daphne Strengthening the Linkages Pro j e c t
conference, Dublin, October.
Kelly, K., Moon G., Bradshaw, Y. and Savage, S. (1998) ‘Insult to injury? The medical investigation
of rape in England and Wales’. Journal of Social Welfare and Family Law, 20(4): 409-20.
Kelly, L. and Regan, L. (2001) Rape: The Forgotten Issue? A European Attrition and Networking
Study. London: Child and Woman Abuse Studies Unit.
Kersetter, W. (1990) ‘Gateway to justice: police and prosecutorial response to sexual assaults
against women’. Journal of Criminal Law and Criminology, 81: 267-313.
Konradi, A. (1996) ‘Preparing to testify: rape survivors negotiating the criminal justice
process’. Gender and Society, 10(4): 404-432.
Koss, M. (2000) ‘Acquaintance rape: a critical update on recent findings with application
to advocacy’. [Available online at:
http://www.vip.msu.edu/theCAT/CATAuthor/MPK/colorado.html].
Lea, S., Lanvers, U. and Shaw, S. (2003) ‘Attrition in rape cases: developing a profile and
identifying relevant factors’. British Journal of Criminology, 43: 583-599.
Lees, S. (2002) Carnal Knowledge: Rape on Trial. (2nd edition). London: Women’s Press.
Lees, S. and Gregory J. (1993) Rape and Sexual Assault: A Study of Attrition. London: Islington
Council Police and Crime Prevention Unit.
Lovett, J., Regan, L. and Kelly, L. (2004) Sexual Assault Referral Centres: Developing Good
Practice and Maximising Potentials. Home Office Research Study 285. London: Home
Office.
M a rtin, P. and Powell, R. (1994) ‘Accounting for the ‘second assault’: Legal org a n i z a t i o n s ’
framing of rape victims’. Law and Social Inquiry, 19: 853-890.
McGregor, M., Du Mont, J. and Myhr, T.(2002) ‘Sexual assault forensic medical examination: is
evidence related to successful prosecution?’ Annals of Emergency Medicine, 39(6): 639-647.
117
References

M e t ropolitan Police Serv i c e (2002) MPS Project Sapphire Strategy Action Plan 2002-2003.
London: Metropolitan Police.
Matoesian, G. (1993) Reproducing Rape: Domination through Talk in the Courtroom. Chicago:
University of Chicago Press.
Myhill, A. and Allen, J. (2002) Rape and Sexual Assault of Women: the Extent and Nature of
the Problem – Findings from the British Crime Survey.Home Office Research Study 237.
London: Home Office.
P a i n t e r, K. (1991) Wife Rape, Marriage and Law: Survey Report, Key Findings and
Recommendations. Manchester University, Department of Social Policy and Social Work.
Regan, L., Lovett, J. and Kelly, L. (2004) Forensic Nursing: An option for improving responses to
reported rape and sexual assault. Home Office Online Research Report 28/04. [Available
online at: http://www.homeoffice.gov.uk/rds/onlinepubs1.html]
Regan, L. and Kelly, L. (2003) Rape: Still a Forgotten Issue. Briefing Paper for the Daphne
Strengthening the Linkages Project. [Available online at: http://www.rcne.com].
Regan, L. and Kelly, L. (2001) Teenage Tolerance: The Hidden Lives of Irish Young People.
Dublin: Dublin Women’s Aid.
Russell, D. (1984) Sexual Exploitation. Beverley Hills, CA: Sage.
S c h u l h o f e r, S. (1998) Unwanted Sex: The Culture of Intimidation and the Failure of Law.
Boston: Harvard University Press.
Schwartz, M. (1997) Researching Sexual Violence Against Women. Newbury Park: Sage.
Senate Judiciary Committee (1993) The Response to Rape: Detours on the Road to Equal Justice.
New York: US Senate. [Available online at: http://www.mith2.umd.edu/WomensStudies/
GenderIssues/Violence+Women/ResponsetoRape/full-text].
Simmons, J. and Dodd, T. (2003), Crime in England and Wales 2002/2003. Home Off i c e
Statistical Bulletin 07/03. London: Home Office.
A gap or a chasm? Attrition in reported rape cases
118

S i n c l a i r, H. and Bourne, L. (1998) Cycle of Blame or Just World: Effects of Legal Ve rdicts on
Gender Patterns in Rape-Myth Acceptance. Psychology of Women Quarterly, 22: 575-588.
Skinner, T. and Taylor, H. (2005) Providing counselling, support and information to survivors of
rape: an evaluation of the Surviving Rape and Trauma After Rape (STAR) Young Person’s
project. Home Office Online Research Report. 51/04. [Available online at:
http://www.homeoffice.gov.uk/rds/onlinepubs1.html].
Smith, L. (1989) Concerns about Rape. Home Office Research Study 106. London: HMSO.
Spohn, C. and Horney, J. (1996) ‘The impact of rape law reform on the processing of simple and
aggravated rape cases. Journal of Criminal Law and Criminology, 86(3): 861-884.
Taslitz, A. (1999) Rape and the Culture of the Courtroom. New York: New York University Pre s s .
Temkin, J. (2003) ‘Sexual history evidence – beware the backlash’. Criminal Law Review,
April, 217-242.
Temkin, J. (2002) Rape and the Legal Process. Oxford: Oxford University Press.
Temkin, J. (2000a) ‘Rape and other sexual assaults: a literature review’. In Home Off i c e ,
Setting the Boundaries, Volume 2. London: HMSO.
Temkin, J. (2000b) ‘Prosecuting and defending rape: perspectives from the bar’. Journal of
Law and Society, 27(2): 219-248.
Temkin, J. (1999) ‘Reporting rape in London: a qualitative study’. Howard Journal of Criminal
Justice, 38(1) 17-41.
Temkin, J. (1998) ‘Medical evidence in rape cases: a continuing problem for criminal justice’.
Modern Law Review, 61: 821-848.
Temkin, J. (1997) ‘Plus ca change?: re p o r ting rape in the 1990s’. British Journal of
Criminology, 37: 507-528.
Tjaden, P. and Thoennes, N. (1998) ‘Prevalence, incidence and consequences of violence against
women: findings from the national Violence Against Women Survey’. National Institute of
Justice Centers for Disease Control and Prevention: Research in Brief, pp 1-16.
119
References

Ullman, S., Karabatsos, G. and Koss, M. (1999) ‘Alcohol and sexual aggression in a national
sample of college men’. Psychology of Women Quarterly, 23: 673-689.
Vasschs, A. (1994) Sex Crimes. New York: Arrow Books.
Walby, S. and Allen, J. (2004) Domestic violence, sexual assault and stalking: Findings from the
British Crime Survey. Home Office Research Study 276. London: Home Office.
Victim Support (1996) Women, Rape and the Criminal Justice System. London: Victim Support.
Wo m e n ’s Unit (1999) Living Without Fear: An Integrated Approach for Tackling Vi o l e n c e
Against Women. London: Cabinet Office.
A gap or a chasm? Attrition in reported rape cases
120