VULNERABLE AND INTIMIDATED WITNESSES: A REVIEW OF THE LITERATURE

Robin Elliott Home Office Research and Statistics Directorate

99 100 The views expressed in this report are not necessarily the views of the Review Group or the Home Office.

101 Acknowledgements

I would like co thank the staff of the Home Office Library for their assistance in collecting the material on which this reporr is based. Thanks are also due to the many others, both within the Review team and within the Home Office, who provided material and commented on early drafts of this report. Thanks also to Natalie Wood and Margaret Allday for their help in producing the charts and tables.

ROBIN ELLIOTT

102 Summary

Defining vulnerable andinrimidarcd witnesses • victim intimidation is more common than will be an important: factor in influencing whether non-victim witness intimidation; those in most need of special assistance receive it. This is nor straightforward, but some jurisdictions • women are at greater risk than men; have legislation enabling extra help to be given to "special witnesses" which suggest possible criteria. • risk of intimidation seems to vary according These include: to the nature of the initial orrence;

• the witness's personal characteristics (such as • intimidation is more likely when the physical or mental condition, age and offender is known to the witness; and cultural background); • in most cases, intimidation seems to be • the nature of the offence; perpetrated by the initial offender.

• the relationship between the witness and Six tasks for the criminal justice system are: defendant; • minimising risks oi intimidation associated • the nature of- the evidence the witness is with involvement in the criminal justice required to give; and system (including reporting intimidation);

• the defendants characteristics (particularly • preparing for the possibility of intimidation; dangerousness). • recognising intimidation: The literature review focuses on three main groups: intimidated witnesses, those with disabilities arid • dealing with intimidation where it occurs; illnesses, and victims of special offences. • preventing further intimidation; and

Intimidated witnesses • mounting a case when no witnesses come forward. Several authors have suggested there are different types of intimidation. For example, case-specific Measures can he identified to assist each task: for intimidation involves threats or violence intended example surveillance operations and professional to discourage a particular person from helping a witnesses may be used to mount cases where particular investigation. Community-wide witnesses will not come forward. Such approaches intimidation covers acts intended to create a have limitations though: it seems preferable general atmosphere of fear and non-cooperation therefore that more attention is given to the other with the criminal justice system, within a tasks listed above. particular area or community. These categorisations are useful because they suggest different approaches may be needed to tackle Witnesses with disabilities and different types of intimidation. illnesses

However, information on the scale and nature of More literature was found on the experiences of the problem is very limited, partly because of the witnesses with learning disabilities in the criminal nature of the problem. Despite this, there is some justice system than for those with physical evidence suggesting: disabilities or mental illness. However it appears

103 that at lease five areas of personal hinctioning may violence, racial incidents and hare crimes agaiusr be affected by disabilities and illnesses: sexual minorities.

• memory; it is difficult to estimate the scale of these crimes, for various reasons including under-reporting • communication skills; (particularly in sexual and domestic violence) and the way these offences are recorded by the • emotional resilience (including response to criminal justice system (particularly domestic perceived aggression); violence and racial incidents). Nevertheless, the evidence suggests that these offences are relatively • mobility; and rare. Sexual offences and domestic violence appear to be much less common than property crime for • soci.il skills. example, and racially motivated crimes much less common rhan crimes where there was no Some means of distinguishing those who are apparent racial motive. vulnerable is needed. Three approaches are discussed: Against this, the recorded rates of all three offences have increased in recent years and incidents of • drawing up a list of groups who qualify as domestic violence reported to the British Crime vulnerable; Survey have increased. However, increased reporting could at least partly account for this. • detailing one or more tests to determine whether a particular witness qualifies; and A number of areas were found where the literature .suggests die criminal justice response to witnesses • leaving the decision to the discretion or the of special offences could be improved. Some of various criminal justice agencies: expert these are particular to one offence: for example, assessments could be used to this end. having a choice of a female doctor to conduct medical examinations for female sex offence Seven themes to improve responses to witnesses complainants. Other concerns were common. For with disabilities or illnesses can be identified: example one common concern is that seeing the alleged offender in court may be upsetting: • encouraging reporting; screens are one possible response. Some efforts have been made to improve the criminal justice • idenniying vulnerability; response. However, very little research was found evaluating the effect of these changes. • facilitating communication; Nevertheless it appears that although some improvements may have been made, there is srill • recognising that a crime has occurred; concern about how these groups fare.

• increasing understanding; Conclusions • providing support; and Numerous possible measures to improve the • preventing future offences. situation of vulnerable witnesses are identified within the report, requiring varying levels of A thread which runs through many of these intervention. Some are specific to a particular type concerns is the role of carers. of vulnerable witnesses, but many could be applied to more than one group. A number of practical issues about using special measures for Victims of special offences vulnerable witnesses are identified, such as whether measures should be granted as a right, Victims of special offences and possibly others and who should have responsibility for providing who have suffered repeat victimisation may be them. More fundamentally, the use of special seen as vulnerable witnesses. Four main groups are measures to protect vulnerable witnesses has considered: victims of sexual offences, domestic implications for justice.

104 Section 1: Introduction

Background providing any such .special measures would be reduced. This report summarises the findings or a literature review on vulnerable and intimidated witnesses. Despite the importance of carefully defining The literature review was commissioned by the vulnerable witnesses, there is little literature on Home Office to feed into and inform an inter- this subject. This is at least partly because most of departmental government review of this area the literature examines how individual groups of (hereafter referred to as "the review"). The review's people experience the criminal justice process, terms of reference were as follows: rather than considering vulnerable witnesses as whole. This neglect could also perhaps be related UI laving regard to the interests of justice: the to a concern (noted by the Western Australian importance of preventing and detecting crime, Law Reform Commission I 990: 64) that singling the needs of witnesses and cost effectiveness, out certain groups as vulnerable may be experienced as patronizing or discriminatory. and taking into account the National Standards of Witness ("are in England and Wales, However, failure to recognise and compensate for inequalities between witnesses seems both • to identify measures at all stages of the inhumane (when this results in stress or trauma criminal justice process which will improve for the witness) and unjust. There are other the treatment of vulnerable witnesses, problems: although a group of people may be including chose likely to be subject to potentially vulnerable, in practice not all members intimidation; of the group will actually be vulnerable.

• to encourage such witnesses to give evidence Those authors who have looked at vulnerable of crime and enabling them to give best witnesses as a whole have generally tackled these evidence in court; problems in the same way:

• ro consider which witnesses should be • first, by suggesting certain conditions where classified as vulnerable; a witness may be vulnerable; and

• to identify effective procedures for applying • secondly, by recommending that the courts appropriate measures in individual cases; be given the discretion to decide whether those conditions are met in the individual • and to make costed recommendations."

Some examples of the definitions which have been Definitions suggested ate given overleaf.

The definition of a vulnerable or intimidated witness is a key concern for the review. Deciding who should be classified as a vulnerable witness has important practical implications. In particular, where special assistance is available to vulnerable witnesses generally, it will be an important factor in determining whether those most in need of assistance receive it. Clearly there are dangers in drawing the definition too broadly or too narrowly: In both cases, the cost effectiveness of

105 Example 1: Queensland Evidence Act Example 3: Western Australia Law 1977 S.21A (quoted in Law Reform Reform Commission Report (1991: 116, Commission of Western Australia 121). Discussion Paper,1990: 65-6)._ The Commission recommends a dual approach, Vulnerable witnesses are described as ''special" in which some groups face more strenuous tests witnesses. These are defined as a child under 12 than others. years, or: For witnesses with an ''intellectual handicap or "a person, who in the court's opinion - other mental or psychological disorder or physical handicap" the test suggested is whether (i) would, as a result of intellectual impairment the witness "is likely to be unable to give or cultural differences, be likely ro be evidence in accordance with the traditional rules disadvantaged as a witness; and practice of the court".

(ii) would be likely to suffer severe emotional In other cases, "the court should be able to trauma; or declare any witness a special wirness if, taking into account: (iii) would be likely to be so intimidated as to be disadvantaged as a witness, (1) a person's age, cultural background, or relationship to any other party in the if required to give evidence in accordance with proceedings, the usual rules and practice of the court". (2) in a criminal case, the nature of the offence, or Example 2: Western Australia Law Reform Commission Report (1991:11) (3) any other relevant factor.

The Commission begins by defining a the court is satisfied that the person vulnerable witness as: "any competent witness (either for the prosecution or for the defence) a) would be likely to suffer unusual emotional for whom the giving or evidence is likely ro be trauma, or especially traumatic or even impossible". However, it then goes on to develop the more b) would be likely to be so intimidated or stressed detailed definition given in Example 3- as to be unable to give evidence

There are a number of similarities between these if required to give evidence in accordance with definitions, suggesting some consensus that the traditional rules and practice of the court" certain groups of people may be particularly (emphasis added)". vulnerable. Most commonly, rhe witness's persona] characteristics are identified as a These definitions have a number of limitations: potential source of vulnerability, namely: • Most fail to specify whether they apply to - physical or mental handicaps or illness both victim and non-victim witnesses (the exception here is example 2, which refers to - age (the elderly and children); and both prosecution and defence witnesses). It seems likely that most vulnerable witnesses - cultural background. will be victims of crime, but it is possible that other witnesses could be vulnerable. The nature of the offence, nature of the evidence the witness is to give and the relationship between the witness and the defendant have also

been suggested as possible causes of vulnerability. A virtually identical definition was subsequently adopted in the Western Australian Acts Amendment (Evidence of Children and Others) Act 1992 (section 106R).

106 • Intimidated witnesses are nor usually from special measures, so their advice could specified, but seem to be covered in most be useful for the court. However, it is also cases by references TO the rraunia suffered by possible thai some witnesses might reject vulnerable witnesses1, and those concerning measures which could help them, for ability ro give evidence. This accords with esample to avoid being labelled as the reviews terms of reference (see above, vulnerable. bullet point one), which treats intimidated witnesses as just one class of vulnerable In addirion to these criticisms concerning the witnesses. groups covered, there are some limitations worth noting about how vulnerability is defined. First, • None or" the definitions encountered most of the definitions are restricted to suggested the witness's sex or sexual vulnerability arising from giving evidence at disposition could be a source of vulnerability. court. As the reviews terms of reference suggest, Yet there is some literature suggesting that witnesses may experience trauma ar other stages in both female witnesses and lesbian and gay the criminal justice system. Trauma can arise as witnesses may encounter prejudice in their early as the initial report or at any rime until after contacts with the criminal justice system. the case is heard in court (if it ever is). This The same is also true of race, although indicates that a broader criteria of vulnerability is '"cultural background" (examples 1 &C 3) needed than having difficulty in giving effective does touch on this. (See .section 4 for further evidence at courr. discussion of the evidence regarding sex, sexual disposition and race). Example 5: Crime and Punishment (Scotland) Act 1997 s29 • Xor have any of the definitions considered offender characteristics: in some cases The definition adopted in Scotland is much witnesses may be vulnerable because the more narrow that the Scottish Law Commission offender is dangerous. recommended (see Example 4). Under the Act, special arrangements for child witnesses (such as Example 4: Scottish Law Commission screens, live television links and video- (1990). recordings) are extended to adult vulnerable witnesses. Vulnerable adult witnesses are defined The Scottish Law Commission suggested the as people who: following issues should be considered when gran ring rhc special measures it recommends for • ate sixteen years or older; and vulnerable adult witnesses (video-taping pre-trial • are subject 10 a court order under the depositions, screens and live closed circuit Mental Health Acts on the grounds of television links): suffering from a mental disorder; and • appear to the court to have a "significant • the age of the witness; impairment of intelligence and social • their physical/mental capacity; functioning" (emphasis added). • the nature of the offence; • the relationship between the witness and The Act also details three issues for the court to the defendant; take into account when assessing an application • the possible effect on the witness if required for special treatment: to give evidence in open court; and • the probability that the witness would give • "the possible effect on the vulnerable person better evidence if not required to do so in if required to give evidence, no such open court. application having been granted; • whether it is likely that the vulnerable • Only one (example 5) allows the witness's person woid be better able to give evidence views to be taken into account. None of the if such an application were granced; and literature found considered whether • the views of the vulnerable person' (emphasis witnesses' views should be considered. This added). is a difficult issue. Witnesses may be best placed to judge whether they would benefit

107 Example 6: Report of the Advisory Example 7: Literature review definition Group on Video Evidence 1989 (The "Pigot Report") For the purpose ok the literature review, a vulnerable witness is any witness (whether a The Group suggested a "test of vulnerability" victim or not) who is likely to find: whereby the courts can declare a witness is vulnerable if they are: • witnessing a crime; • any subsequent contact with the criminal "likely to suffer an unusual and unreasonable justice system; degree of mental stress if required to give evidence in open court, having regard to: unusually stressful, upsetting or problematic, because of: • the witness's age; • their physical and mental condition; • their personal characteristics; • the nature and seriousness of the offence; • rhe nature of the offence; and • the nature of any evidence they are called • the nature and .seriousness of the evidence upon to give at any stage to assist the justice they are to give". process; • the offenders characteristics; For all victims of serious sexual offences, the • any relationship between them and the Group recommended that there should a defendant; or rebuttable assumption that they were vulnerable • intimidation. witnesses (1989: 3.5).

The examples given also tend to associate Report Scope and Structure vulnerability with trauma. 1c could be argued that vulnerability should encompass practical problems Scope in dealing with the crime, or contacts with criminal Justice system or both. Another form of The scope of the report has inevitably been limited vulnerability is the possible risk of future by practical constraints. The focus of the report has victimisation. There is evidence chat past been restricted in two main ways. Firstly, child victimisation increases the risk of victimisation in witnesses were excluded because this sizeable area of the future (see for example, hirrell & Pease 1 993- literature would have been impractical to include and Lloyd Farrell and Pease 1994), suggesting that given time constraints. This is, however, an area the victim-witnesses may be more vulnerable than Review has examined separately. non-victim witnesses . Secondly, by necessity the report covers only those In practice of course, a witness may experience- groups whose experiences of witnessing have been more than one form of vulnerability. Different documented. This means that some groups, for forms may be mutually reinforcing: practical whom rhere is very little or no relevant literature problems may add to trauma already ielt by cannot be considered in detail although certain contacts with rhe criminal justice system. It may characteristics specific to them will be covered: therefore, be worthwhile defining "vulnerability" more loosely. • The elderly No erature was found examining the Definition used for the literature review lit experiences of elderly witnesses. It seems To avoid unduly restricting the scope of this reasonable to assume that, where elderly witnesses literature review, the definition used in this report are vulnerable, at least in some cases diis may is a broad one (see example 7). This definition is stem from mental or physical disabilities/illnesses not proposed for the review in considering any- which become more common in old age (covered future legislation: it is however intended to avoid bv section 3). excluding any potential groups of vulnerable witnesses the review team might wish TO consider.

108 • Repeat victims on special offences, which also discusses issues of sex and race. Again it is not clear whether repeat victims tend to find witnessing more upsetting or problematic than other groups. It is possible that they may Finally, it is worth drawing attention ro the have special difficulties, for example in receiving possibility of multiple membership of the various an inappropriate police response because they groups of vulnerable witnesses (for example, a have not been identified as repeat victims. black disabled rape victim). It is likely that However, there is now significant awareness of membership of more than one group of rhls issue, and efforts are being made to give vulnerable witness may compound the negative repeat victims .special assistance (tackling repeat experiences and perceptions involved in vie rim is anon is one of the polices key witnessing. performance indicators). Repeat vienms are discussed in section 2 in relation ro witness Structure intimidation, and in section 4 regarding some The report is divided into five sections. Sections 2 special offences such as domestic violence, where - 4 each examine a different group of vulnerable repeat victimisation is common. witness. Section 2 examines intimidated witnesses; section 3 looks at those who are vulnerable • Lesbian and gay witnesses because of physical and mental disabilities and illnesses: and section 4 cases where the nature of There is a small body of literature examining the the offence can make a witness vulnerable. Each law (primarily in relation to the age of sexual of these sections begins by examining the nature consent) and police attitudes to and treatment of of vulnerability associated with the group in this group. This suggests that witnessing may be question, and then discusses possible measures to upsetting or problematic for these groups. Lesbian ameliorate the problem. Finally, section 5 draws and gay witnesses are considered under section 4 some conclusions.

109 110 Section 2: Witness Intimidation

The Problem • The perpetrator must know or believe that the other person is helping or has helped an Witness intimidation can discourage some investigation, is a witness or potential witnesses from reporting crime or coining forward witness, juror or potential juror'. with other evidence, and could cause cases char do go ahead ro be losr or abandoned. At a more • The perpetrator must have acted or general level, ir is thought to undermine both threatened the person because they believed public confidence in the criminal justice system this, and with the intention of obstructing and its effectiveness. Until recently witness the course of justice. In many cases it may intimidation could only be prosecuted under the be difficult to prove that the perpetrator common law offence of perverting the course of intended to pervert the course of justice. justice, which also covers other acts such as Consequently, the Act says that this making false allegations of crime. The 1994 intention will be presumed unless the Criminal Justice and Public Order Act created contrary is proven. two new offences: The difference between the two offences is that • intimidating a witness; and witness intimidation applies to current investigations, but harming or threatening to • harming or threatening to harm a witness. harm a witness applies only after the trial has ended. So for example, if the intimidation Legal Definition occurred because the perpetrator believed the witness had helped with the investigation of an The legal definitions or witness intimidation and offence at some point in the past, the offence harming or Threatening to harm a witness under would be harming or threatening to harm a the Act both cover: witness'1. • threats to harm someone and acts to harm them;

• physical and financial harm to the person or the property; and

• acts or threats against a third party (for example a relative of the person they want to intimidate). in addition:

• The act or threat muse lie intended to intimidate: so for example, a casual remark that someone should not bother going to Juror intimidation is not specifically examined hy this report, the police, intended merely as comment or but it a worth noting that some of the measures reported in persuasion would not be counted. In relation to witness intimidation may also be applicable to furor intimidation. practice of course, it may be difficult to No lime limits have been set for the prosecution to bring such a distinguish at what point an effort to case. There are however time limits on the presumption of intent, persuade becomes intimidation. meaning that intent cannot be presumed in cases where the alleged harm or threat occurred outside the relevant period (usually within a year of the act or within a year of conclusion or the trial or appeal when intimidation occurred during the trial).

111 Types of intimidation probably consisting of attempts til persuade rather than attempts to intimidate. At least some Several authors have identified different types of community-wide intimidation, where the witness intimidation (see examples 1 to 3 below). behaviour is not be directed towards individual witnesses, may also be excluded. These categories Example 1: Case-specific and are worth considering however, given that they community-wide intimidation (Healey may all produce the same effect in undermining 1995:1) the criminal justice process,

Healey (1995:1) distinguishes two types of Example 3: Three tiers: small core, intimidation: middie ring and outer ring (Maynard 1994:1) • Case-specific intimidation involves threats or violence intended to discourage a Maynard (1 994: 1) takes a slightly different particular person from helping a particular approach, identifying three tiers: investigation; whereas • The small inner core consists of the most • Community- ide intimidation covers acts w serious cases, where intimidation is life- intended to create a general atmosphere of threatening. These witnesses need high fear and non-cooperation with the criminal level protection such as changes of identity justice system, within a particular area or and relocation; community. • The middle ring comprises those witnesses Two points are worth noting. First, Healey who have experienced non life-threatening emphasises that community-wide intimidation is intimidation; and potentially as harmful to the criminal justice system as case-specific intimidation. Secondly, • The outer ring covers people who are the two forms are interrelated: each example of discouraged from reporting by the case-specific intimidation reinforces community- perceived risk of threats or harm, even wide intimidation. where they themselves are victims of crime. (There is some overlap between this outer Example 2: Traditional, cultural and ring and both personal and community- wide intimidation). perceived intimidation (ABA 1981:1) There are some similarities between the categories. Another classification has been suggested hy the For example: American Bar Association (1981:1): • case-specific intimidation (example 1) is • Traditional iniimidmon occurs when much the .same as traditional intimidation threats or acts are made against a witness, (example 2); their property or a member of their family; • community-wide intimidation (example 1) • Cultural intimidation occurs when friends is similar to perceived intimidation (example or family of the witness rry to dissuade 2). (The main difference is that the former is them from assisting an investigation; and deliberately fostered by offenders, whereas the latter is not necessarily Intentional); and • Perceived mtxmi&axion occurs when fear of possible intimidation or retribution is felt by a witness. • perceived intimidation (example 2) is similar to Maynard's outer ring (example 3). It should be noted that neither cultural nor These categorisations are useful because they perceived intimidation are covered by the English suggest that different approaches may be needed legal definition of witness intimidation (see to tackle different types of intimidation. For above). Most cases of cultural intimidation would example, Maynard observes that of the three tiers also be excluded by the English definition, he identifies, the inner core has been given most

112 attention. However, the high-level protection Even when convictions for perverting the course schemes designed for them are nor suitable ior of justice: are included, less than 1% of offenders most witnesses: convicted in 3 996 were convicted for witness intimidation. This probably greatly • the schemes are expensive; underestimates the scale of the problem: intimidation may mean that the initial offence is • they require major lire changes (such as never reported, let alone the offence of witness moving to a new area and severing all intimidation. contacts from the past) by witnesses, which could be viewed as penalising them; and 2. Crown Prosecution Service Survey

• most witnesses would not be willing to Another source of data is the Crown Prosecution make such drastic changes to their lives. Service, which takes decisions on whether to prosecute cases in consultation with the police. This suggests that alternate measures need to be Maynard reports the findings of an unpublished examined For the middle and outer rings survey by the CPS. Crown Prosecutors completed (Maynard 1994: 2-3). Other measures would also a questionnaire whenever a case was discontinued be desirable for the inner core witnesses who are in the Magistrates' Courts during one month in not prepared to take part in high-level protection 1993. In three quarters of cases where the schemes. As Maynard notes, this needs to be prosecution was unable to proceed, the reason based on a .strong understanding of the problem given was that a key witness was missing or of witness intimidation affecting these groups, but refused to give evidence. According to Maynard at present very little is known. (1994: 4-6) this accounts for over 1% of the cases dealt with by the CPS each year, although not all Research on the scale and nature of the problem these will have been the result of intimidation. The literature review found very little published 3. Police Research Croup Study research on the size and nature of the witness intimidation problem: The Home Office Police Research Group (PRG) did some research in 1993 to provide more 1- Criminal Statistics information about the scale of the ptoblem. Insufficient resources were available ro conduct a In 19%, there were almost 3^0 convictions for large scale survey to obtain a representative witness intimidation and harming or threatening sample. Instead, the upper limits of the problem to harm a witness in England and Wales (internal were measured using a house-to-house survey in note, 31/10/97). A further 2,000 offenders were five high crime housing estates. This found that found guilty or cautioned for perverting the 13% of crimes reported by victims and 9% by course of justice (Criminal Statistics England and other witnesses were followed by intimidation. Six Wales 1996). per cent of crimes experienced by victims and 22% of those mentioned by other witnesses were Or course, not all the convictions for perverting not reported to the police because of intimidation the course of justice will have involved witness (Maynard 1994: 12-14). intimidation. Other forms of interference with justice will have been included, such as bribery and supplying false information to a police officer. These findings should be treated with caution. Unfortunately it is impossible to say what While it is reasonable to believe that the rates of proportion of these cases were for witness intimidation found were higher than that in the intimidation on the information available'1. general population, the research highlighted a potential problem in using small sample surveys to measure crime: sometimes witnesses perceive intimidation where none exists. For example, It seems likely that bribery of witnesses is less common than Maynard found that some of those respondents intimidation: bribery inevitably voiti the perpetrator something, reporting intimidation were actually repeat victims. whereas intimidation ran he achieved free. However, it is In some crimes it can be difficult to decide whether impossible to estimate Imw common other forms of interference with justice (fur example supplying the police with false intimidation is real or perceived as the result of info: mutton) are m comparison. repeat victimisation (1994: 17). Occasions where

113 acts arc intended to intimidate but are nor Seventeen percent of vicrims and 4% of perceived as such by rhe victim are probably fever, other witnesse;. reported intimidation suggesting that this approach truly tend ro (Dowd.s and Budd 1997: i & ii). This would overestimate the extent of witness intimidation. he expected, given that victims ate more likely to know the offender rhan non-victim 4. British Crime Survey witnesses. However, this probably overestimates the size of the difference. The Finally, the British Crime Survey asks people figure for victims is probably higher than in about their experience of crime over the previous the general population of victims, because year and whether they made a report. The survey the victim .sample was restricted to people does not routinely ask people about their who had some knowledge of the offender. experience of witness intimidation, bur does Dowds and Budd (3 997: 5) estimate that at examine victims' reasons for failing to report minimum 6% of all crimes reported in the offences to the police. One of rhe .strengths of this 1994 survey were followed by intimidation .survey is its size: the core sample for the 1996 of victims. survey covered 15,000 people. Another advantage is the survey's high response rare (83% in 1996)'. • Most intimidation involved verbal abuse Both factors give confidence that the survey and then threats, with physical assaults and findings are likely to be typical of the general damage to property less common. population. Incidents often involved more than one type The 1996 survey found that: of harassment (Figure 1). For victims, 71% of intimidatory incidents involved verbal • The most common reasons tjiven were that abuse. 41% involved threats, 16% physical the offence was roo trivial (.40%) or that the assault and 9% damage to property. A very police could nor do anything (29%). similar pattern was found tor other witnesses (73% reporting verbal abuse, 34% threats, • Fear of reprisals accounted for only 4% of 1 2% physical assault, and 9% damage to all cases not reported, but 11% for both property). assault and robbery cases (Mirrlces-Black, Mayhew & Percy 1996; 2). • In most cases the original offenders were thought to have been responsible for the Maynard (1994:8- discussing the 1992 survey) intimidation. suggests rhe high rate for non-reported assauit is probably because the victim and defendant are •Questions about the identity of harassers more likely to know each other, so the were only asked of respondents who had opportunities for intimidation arc greater. information about the offender, so - as the Similarly, in robbery the victim is more likely to researchers state - this finding is not be able to identify rhe offender than in other surprising. The relationship between the property offences. victim and victimlzer is more interesting, i-emale victims were more likely than men Special questions were included on witness to be intimidated by ex-partners or partners. intimidation in the 1994 .survey (Dowds & Budd Men were more likely to be harassed by 1997: i). These included rhe extent of other relatives or household members, and intimidation, who was responsible lor it, and the by work contacts. Similar proportions of nature of rhe crime witnessed. The main findings were: men and women were intimidated by friends or neighbours (Dowels & Budd 1997:7). • Victim intimidation was much more common than non-victim witness • intimidation was more common if the Intimidation. witness reported the initial offence to the police, but failure to report did not. Figures taken from rhe 1996 Hrimh Crime Survey Technical guarantee immunity. Report by J.Hales & N. Stratford, ami table from Social and Community Planning Research, London.

114 Harassment was almost twice as likely where • Risk of intimidation varied according to the the initial incident was reported to rhe nature of the initial offence. police: 20% ot those victims who reported were intimidated compared to 14% oftho.se Victims reported higher rates of who did not report. The original offence was intimidation following sexual offences, twice as likely to be reported as vandalism and assaults than tor other intimidation, and the likelihood of reporting crimes. Similarly, non-victim witnesses were intimidation was tied to reporting patterns more likely to be intimidated following tor the initial offence. Those who had not assaults and vandalism. A possible reported the initial offence were much less explanation is that intimidation might he likely to report harassment (3%) than those greater in "expressive' offences (directed who had reported the initial offence (19%). against a particular victim) than Non-victim witnesses were more likely to "instrumental" offences (motivated by report intimidation (39%) than victims personal gain). (Dowds & Budd 1997: (23%) (Dowds & Budd 1997: 8-9 & 13- Tables 2 & 9). 14). These findings are useful, but the Women were more at risk than men. representativeness of the victim sample (as opposed to the witness sample) is questionable. Female victims were more likely to be The sample was restricted to victims who knew intimidated than men (19% as against 14% something about the offenders identity. of men). This difference applied whether or Intimidation is less likely in cases where victims nor the initial offence was reported and even do not know the offender's identity, but the though a higher proportion of cases against prerequisite for intimidation is that the offender women were not reported because of fear of can identify the victim in some way. The offender reprisals (Dowds & Budd 1997: 4 & 9). may fear rhe possibility of a police investigation This could have been at least partly because uncovering his/her identity, not just what the women were more likely to know the victim could tell rhe police. offender than men.

115 The BCS also reinforced Maynard's finding thar Sampling error could also account for [he repeat victimisation and intimidation are bound apparent increase. up with each other. In repeat victimisation, harassing the victim may be part of rhe motive Similarly, the number of convictions for witness and effect of both the initial and subsequent intimidation rose more than threefold from offences. Viewing only second and subsequent almost a hundred convictions in 1995 (internal victimisations as intimidation may be note, 18/9/97). This may well have been related inappropriate. Dovvds and Budd acknowledge to increased awareness of the new measures. The that this depends though on what we want to number of offenders found guilty or cautioned for measure, if the concern is to measure intimidation perverting the course of justice has also increased as defined by the law (in other words where rhe sharply, from 275 cases in 1985 to over 2,000 in intent is to obstruct rhe criminal justice system) it 1996 (Figure 2). This may indicate that witness is necessary to separate victimisations intended to intimidation is growing, but it is not certain that dissuade the victim from cooperating with the cases of intimidation accounted for the increase. justice system (Dowds and Budd, 1997: 10-11). Even if cases of intimidation do account for this, the increase may not reflect an increase in witness From this point of view, including all repeat intimidation in general. Greater convictions victims as victims of intimidation may inflate the might be related to growing awareness of, and 'real' level of intimidation as defined by the law. harsher attitudes towards, witness intimidation Excluding them may underestimate the problem. within the criminal justice system. More This suggests further information is required: specifically, police officers and Crown Prosecutors Dowds and Budd suggest one measure might be may be more willing to pursue a case of perverting whether victim-witnesses felt the aim was to the course of justice now that greater support discourage them from the criminal justice .system (from specialised police units) is available to (1997: 11). i iowever, complicating matters intimidated witnesses. Further the law .says thai intent to obstruct the course or justice can be assumed if the offender Perhaps less plausibly, reporting rates may also believed that the victim was a witness or a have increased. Rape reporting is said to have potential witness. If intent can be assumed by the increased as police treatment of rape victims courts, a question is raised about whether improved (through setting up rape suites for researchers need to examine intent when trying to example): the creation of specialist units m some measure intimidation. Deciding on the best wav forces ro deal with witness intimidation might to measure intimidation is clearly far from simple. have hud a similar effect.

Outstanding questions 2. I low does the scale of the problem in England and Wales compare with other The paucity of information means little or countries? nothing is known on several important questions: he lack of research scale of witness 1. What is the trend in the rate of witness T on the intimidation exists in other countries, it is likely intimidation? that some countries have a greater problem with witness intimidation than others, just as crime Several authors have claimed that the problem of rates in general vary. For example, it has been witness intimidation is rising (see for example suggested that we have a greater problem than the Clarke 1994, 12; Palmer 1994, 10; Reville 1995, rest of Europe (Clarke, 1994: ]2), although no 1775) but hard evidence is lacking. The two most firm evidence was given co support this. Similarly, recent sweeps of the British Crime Survey (1994 witness intimidation may be greater in America, and 1996) have found that increasing numbers of where it has been acknowledged as a common people attribute their failure to report crimes to problem. Suggestions that rhe problem is rising the police to fear of reprisals. However, rhe have also been made in America. For example increase (from 2 to 4%) may not reflect a real rise Healey (1995: 1) interviewed a number of in fear of reprisals but may simply be due to criminal justice professionals from 20 chance. For example some offences are more likely jurisdictions, and bund that most agreed that to be reported than others, so differences in the intimidation was rising. However, she also types of offences reported could explain this. observed that:

116 "A decade ago, commentators no red that only assault than some other crimes: I lealey goes unsuccessful intimidation attempts ever came further and suggests that a violent initial to the attention of the police and prosecutors. crime generally increases the chance that a Today, prosecutors report that extremely victim or witness will be intimidated. More violent intimidation attempts - which are information could be collected. It seems almost always successful - are coming to their plausible that intimidation may also be attention with increasing frequency" (Healey, prevalent where the stakes are highest, .such 1995:2). as in organised crime where large sums of money are involved. In America, drug crime This suggests that witness intimidation may be is thought to be strongly associated with becoming more serious in America, bat it is also witness intimidation, although again no plausible that increased reporting is a factor. firm evidence of this was found by the literature review (Healey 1995, 1: Los 3. Who is at greatest risk or intimidation? Angeles Times 14/1/97: Reuters News Sen-ice 12/1/97). In theory any witness might be a victim of witness intimidarion. The possibility of obtaining a b. Some groups of witnesses are at greater risk criminal record and being punished may lead to of intimidation than others. witness intimidarion however trivial the initial offence. In practice, there is some support for the The BCS report suggested that victims were idea that the risk of intimidation is nor equally moie vulnerable than non-victim witnesses, distributed: and female victims (particularly those aged between 31 and 60) were more vulnerable a. Intimidarion is more prevalent in some than men. Victims who knew the offender types of crime than others. wete also more likely to be intimidated, although this may be due to bias towards The types of crime involved may help decide this group in die survey design. Again what the most appropriate response is. The further information would be useful, it BCS and PRG data suggests that seems plausible that witnesses who are intimidation is more likelv for witnesses of vulnerable for other reasons (such as

117 learning disabilities) may make easier targets and offender live or work near each other and therefore be more prone to intimidation than if they spend their lives miles apart. than ''ordinary" witnesses (Healey 1995:3). This does not however, appear to have been covered by any of the research to date. c. Intimidation will be more common when the offender is known to the witness. Sometimes these factors will be interlinked. For example: The BCS findings give some support to this idea, although (as noted above) the survey • victims of violent and sexual crimes are design was restricted to victims who knew more likely to be acquainted with the the offender. For intimidation to occur, the offender than vicnms of property crime; and harasser must be able to locate the witness: this is obviously more likely when they were • geographical proximity to the offender may previously known to each other. be more common where the witness and offender know each other. It does not follow, however that the victim must know the offenders' identity- Nor does This will clearly increase the risk of intimidation. it follow that the original offender In addition, victim witnesses of some sexual necessarily intimidates the witness crimes such as rape may be vulnerable for other him/herself. Relatives or friends of the reasons. The difficulty of proving rape, and original offender may be the harassers, prospects of an internal medical examination, having learnt the witness's identity from the public examination of intimate sexual behaviour original offender. For example, a Victim in the courtroom and media reporting may all Support study (1996: 12) found that a third make rape victims particularly vulnerable (see of Victim Support schemes repotted chapter 4). intimidation of rape victims by family or friends of the defendant, against about a 4. When are witnesses at greatest risk of quarter who reported intimidation by the intimidation? defendant (see below for more discussion of this study). Research on repeat victimisation has found that risk or subsequent victimisations is greatest A related consideration (not examined by immediately after an offence. This has been very the BCS report because the sample size was useful in deciding when to intervene: preventive too small to dtaw firm conclusions) is efforts will be most effective: when implemented whether people who are intimidated by as quickly as possible following an offence. partners, ex-partners or other relations are Information on when the risk of intimidation is less likely to report than those intimidated greatest might also have important implications in by others. Fhcre may be particular pressure determining what the most effective response when relatives are involved, where there may from the criminal justice system might be. Only be concerns about bring shame on the the PKG study examined this, but little ramily and about how other relations would information was given, just that intimidation react. Against this, it could be argued that occurred at two main points: either "soon after threats or intimidation from people who are the initial crime'' or "at the time of the court known to the victim might be perceived as appearance". less real or serious than when strangers are involved. 5. How do intimidated witnesses view the criminal justice process and (where relevant) d. Geographical proximity to the offender what are their experiences of it? increases risk of intimidation. Only two pieces of research were found that Healey (1995: 3) also suggests that risk of examined intimidated witnesses perceptions and intimidation will be greatest when the experiences of the criminal justice process, and witness lives, works or studies near to the how it could be improved. One was the PRG offender. It is likely that opportunities for study which found some dissatisfaction with the intimidation will be higher if the witness police response, including the information and

118 advice qivcn ar the time and .is rhe c-ise proceeded generally, which is only now beginning to be (Maynard 1994: 23). Some of these concerns, rectified. Nevertheless, it is possible to identify a such as not being told whether a suspect hud been number of areas where measures could be given bail, may be common to other witnesses implemented to enhance the prevention, although they rake on .special importance in detection and prosecution of witness relation ro intimidation. (Maynard's findings on intimidation. this subject are discussed further below). a. Reporting The second was research on rape victims' As noied earlier, it seems likely that most cases of experiences of the criminal justice system, based witness intimidation and many of the initial on questionnaires sent to Victim Support crimes witnessed are not reported. This suggests schemes. Of chose schemes reporting contacts that measures are required at an early stage both with victims who had been re-assaulted or to prevent witness intimidation, and where it harassed since the original attack, almost half felt occurs to encourage and support witnesses in police protection was insufficient. About half, of reporting. In practice, the earliest point for which witness services reported rape victims generally measures have been proposed is when a witness were frightened of facing the defcndant(s) and reports an offence to the police. These can be supporters in court. Some of this may have been divided into: about bringing back bad memories rather than concern about their own .safety though. The survey has two main weaknesses: a low response • minimising the risk of intimidation rate (25%) common to this kind of research, and associated with reporting: the face thar women's experiences were reported secondhand through Victim Support staff rather • recognising intimidation; and than from the women themselves (Victim Support 1996: 13, 16). Further research on how • preventing further intimidation. intimidated witnesses view and experience the justice system is clearly needed. Minimising the risks of intimidation

Reporting arrangements can provide 6. What s (he extent and nature of i opportunities for witness intimidation. For community-wide and perceived intimidation? example, Maynard (1994: 18-19) details one case where a witness's identity was inadvertently revealed to the offenders after the crime was It should be noted that all of the existing research focuses on case-specific intimidation: that is reported. In this case, offenders were listening ro intimidation against a particular person. Nothing police radio frequencies to know how quickly they is known about the scale and characteristics at had ro escape, when the witness was identified community wide intimidation - intended to over the radio to officers being dispatched to the create a general atmosphere of intimidation - nor scene. The recommendation here was to ensure or perceived intimidation, where the possibility of that as little information as possible be given over inrimidarion suffices. This is perhaps partly the radio to enable officers to respond. because these forms of intimidation will be more difficult co measure. Another case is reported where a witness, who reported a burglary and was visited by police to take a statement, received a brick through her 2. The Response window shortly afterwards. To avoid similar occurrences, Maynard (1994: 19-20) recommends several alternatives be considered: The literature review found that very litrle material has been published on how witness intimidation is being tackled, either here or • delaying visits to take statements and use of abroad. This may partly be explained, by the need plain clothes officers where possible; for security, and fear that details or the approaches used ro counter the problem could be used by the " conducting house-to-house calls on perpetrators, it may also be related to the neglect neighbouring properties to avoid picking of the issue of witness intimidation more out the witness; and

119 • inviting the witness by telephone to visit a further victimisation and suggests they should police station :o make the statement. have an opportunity to explain how the crime has affected them. Pilot projects looking at how this Maynard recommends that [he choice is left co rhe might work are currently being evaluated by the witness wherever possible, bur clcariy views the third Home Office. option as rhe most viable. He notes thai rhe first option runs contrary to the Audit Commission's Preventing further intimidation recommendarion that only one visit should be required, and could be ineffective if offenders can Once intimidation has come to police attention, identify plain clothes police officers as such. The various measures can be implemented to prevent second option would increase demands on patrol further harassment. The traditional approach has officers, but the third option would reduce them. been to offer some form of police protection However, asking witnesses to visit a police station to (Maynard 1994: 1) such as: make a statement would require more effort for the witness and for the police: some planning would be • increasing police patrols in the witnesses required by the latter to ensure that suspects are not neighbourhood; being interviewed or held at the station at the same time. To avoid witnesses and suspects meeting, the • police transport to take the witness to and custody officer and investigating officer will need to from work, school, shops and so on; liaise. • 24 hour police presence; Recognising intimidation • emergency relocation; As well as measures designed to avoid intimidation resulting inadvertently from • long-term relocation, which may be reporting procedures, other measures might be accompanied by a change of identity; or considered to help identify intimidation at this stage. No suggestions were found in the literature • protective custody. on this, but several possibilities can be proposed: All are expensive, and some might be viewed as • profile raising: increasing police awareness of penalising the victim. Unsurprisingly, Maynard's the problem through police training, force (1994: 1) research suggests that such measures are newspapers, posters and the local policing only used in the most serious cases. plan: Only one suggestion was found for assisting the • guidance: issuing guidance on how to spot larger number of witnesses subject to non-life signs of intimidation; threatening, but potential deeplv upsetting intimidation. Farrell, Jones and Pease (1993: 131- • statement taking and interviewing: requiring 132) suggest char witnesses could be loaned officers to ask witnesses if they have received personal alarms, which will notify police if they any threats or harm to dissuade them from need help and generate a rapid response. This assisting the police. would have several advantages:

Of the three suggestions the first seems the most • It would be less expensive than the practical. The second option may be hindered by traditional approaches, removing the need the limited stare of knowledge of witness for a constant police presence while still intimidation at this stage. The third would require offering protection round the clock. some system of checks to ensure that the requirement was adhered to. It also seems • Some police forces are already accustomed plausible that officers will be unwilling to ask to using these alarms to protect other witnesses if they are being intimidated without vulnerable witnesses, particularly repeat being certain that back-up support was available victims. for those who respond positively. However, the Victim's Charter (1996: 3) does state that victims • As well as reassuring victims and ensuring can expect the police to ask about their fears of that further intimidation will receive a fast

120 and informed response, there is some (albeit A few days after the positive identification the limited) evidence that they may deter witness's bicycle wits damaged and the witness offenders (Lloyd, barrel! and Pease 1994: verbally abused. In another ca.se, the suspects were 10-20). caught trying ro gee away with stolen goods. 1 he two witnesses were asked to look in the van where • Following on from this, if those witnesses the suspects were held in order co identify them: in who Lire most risk of intimidation can be this case they refused to do so for fear or reprisals. identified, alarms might also be used to prevent it occurring in the first place. It is difficult to say whether these were isolated incidents. Recent enquiries by the Home Office At present our limited knowledge of witness found that all police forces in England and Wales intimidation could make it difficult to identify now either have their own screens or access to who should benefit from such measures. Despite screens for identification parades, some of which this it may be possible to identify some cases are one-way mirrors (internal note, July 1997). An where pre-emptive action would be warranted, tor unknown number of forces aJso have purpose example where: built identification .suites or use video- identification. However, the mere existence of • the suspect has a record of witness these facilities is not enough: they obviously need intimidation; to be used routinely. IF. is not known how much use is made of these approaches at present, nor • the witness has a previous relationship with whether there are any obstacles to their use. the offender; Clearly interviews with suspects may also present • the witness is vulnerable in some other way; another occasion when officers can place witnesses at risk by revealing witness's identities. Even if the • no other witnesses have come forward, and suspect is held In custody there is no guarantee of the witness's testimony is essential in an the witness's safety: the Victim Support study important case. referred to earlier suggests the suspect's friends or family may decide to act on their behalf (Victim b. The Investigation Support 1996: 12). Nevertheless, it seems likely that some officers reel justified in telling suspects Three areas are discussed below concerning the witness's identity on the grounds that it may investigation: reveal that a witness has some motive for fabricating claims. In some eases, the suspect may • reducing opportunities tor intimidation; ask officers this question outright. In others, they may declare strong suspicions about the witness's • preparing for the possibility of intimidation; identity, which il left unanswered could be taken and as a tacit admission that their suspicions are well to uncled. • mounting a case when no witnesses will come forward. Some possible measures to discourage the practice are: Reducing opportunities for intimidation • raising police awareness of witness The manner in which an investigation is intimidation (see section a above on conducted can place witnesses at risk, in particular reporting); by revealing the witness's identity to a suspect. One point where witnesses can be placed in a • guidance on when it is appropriate ro reveal vulnerable position is at an identification parade. a witnesses' identities; Maynard (1994:20) reports such two cases. In the first, the police asked a witness to identify the ' requirements not to reveal witnesses' suspects at the scene, following a rapid response to identities in certain circumstances. the witness's telephone call. 1 he witness was asked to walk past a bus queue where the suspects were Against this, it should be noted that pressure to standing along with other members of die public. identify witnesses before interviews may come

121 • concern rhat defence counsel will not have from solicitors. The solicitors' argument is that an opportunity to cross-examine without knowing all the evidence against their clients, they are unable to give adequate advice. The second point Is addressed in relation to Failure by police to disclose the evidence they admission of evidence in section c - "the trial" have may mean rhar solicitors advise their clients (below). Of most relevance at the investigation 1 not to answer questions, frustrating the purpose stage is the second point , which refers to claims of the interview'. This is based on the idea that thai witness statements are not a \cry exact suspects may need to know the identity of their record. Statements are prepared by the police, accusers to defend themselves (for example, so condensing information obtained from interviews char they can say if the accuser might have some into legal jargon which the witnesses may not reason for falsely accusing them). The question of recognise as their own words. Unlike interviews whether police officers should disclose all the with suspects, interviews with victims do not have evidence against a suspect prior to an interview, to be tape recorded or written down verbatim. including the witnesses identity, is therefore a There is therefore, some potential for distortion. complicated issue. (The issue of anonymity will be considered again later on). Taping or video-recording interviews might Increase the credibility or this evidence Preparing for the possibility of intimidation (Wolchover & Hearon Armstrong 1997: 855- 85~). Of the two, tape recording .seems more Measures can also be taken at the investigation viable, The police are used to taping other stage in case a witness is subsequently successfully interviews, and already have much of the intimidated from testifying, or into changing their equipment required. Video-recording would evidence (known as "flipping"). Admitting other require additional equipment which the police are evidence, such as signed statements and taped or not used to operating, and also raise greater video-recorded interviews, is one option. In rhe problems in storing tapes. past this has been problematic: there is a general ban on admitting "hearsay" evidence because the Mounting a case when no witnesses come forward defence has no opportunity to cross-examine to test the evidence. The 1 988 Criminal Justice Aci Of course in some cases, no witnesses are prepared moved some way to rectify this: under .section 23 to come forward. In these situations, there are at of the Act written .statements are now admissible least two options: if: 1. Surveillance operations • the statement was made to a police officer; and The traditional approach has been for the police to mount surveillance operations. This however, • the person who gave it is not giving oral depends on the police having some other source evidence ''through fear or because he is kept of information about likely suspects. It also out of the way". requires cooperation from certain members of the public, such as those who allow the police to use Figures on the use oi this provision are kicking, their premises. Clearly those people who do assist but anecdotal evidence suggests that little use has the police in this way may be at risk of been made of it. Two main explanations have intimidation if their identify or the fact that a been suggested: surveillance operation is being conducted is revealed. The former may be a particular problem • concerns about the validity of statements if a home or shop is used for just a few minutes prepared by the police; and without prior planning.

2. Professional witnesses

A more recent idea has been to employ Preliminary findings from research currrmly being conducted professional witnesses. At present, their use for the Home Office suggest that this practice of pressuring police appears to have been limited to housing officers to reveal evidence before interview has incm/tsed since the authorities wishing to evict problem tenants. The right to silence provisions were introduced (internal note, 28/10/97)

122 professional witness is employed co move into a These criticisms suggest that rhere is very limited nearby property under the guise of being a normal potential for extending the use of professional tenant. They then record anti-social behaviour by witnesses. the subject until sufficient evidence has been amassed. c. The trial... By far the greatest number of measures proposed The literature review did not reveal any published concern the court. Most are intended to reduce material on this practice, but was able to draw on the potential for intimidation at court, or information submitted to the review. According to following a court appearance. Others concern the Department of the Environment (submission how intimidation is dealt with by the courts when June 1997)- the service is usually provided by the authority's own housing; officers, environmental it occurs. health officers or a private company. The Local Government Association (submission August Preventing intimidation in court 1997) also suggests police officers may be used. Together they identify' a total of seven authorities Opportunities for intimidation occur as soon as thought to have used professional witnesses. Both the witness gets to court. Measures for highlight a number of drawbacks: preventing intimidation within the court buildings include: • Certain conditions will need to be met for 1. Structural changes to court design the use of professional witnesses to be feasible. For example, the behaviour causing Some opportunities for intimidation at courr concern must take place in a restricted geographical area. The area has to be arise from the physical layout of the courr and sufficiently open and well-lit for offenders to court buildings. In most courts, there are be identified, and an empty property has to common entrances, common waiting and be available nearby. There are unlikely to common refreshment areas for defence and many cases meeting all these conditions. prosecution witnesses and ordinary members of the public. The Lord Chancellor's Department recommends in its 1993 design guide that • The courts have, on occasion, been reluctant separate waiting facilities be provided in all new ui make a repossession order without and refurbished Crown Courts. The process of testimony from the victim. As a refurbishment will however be lengthy, and does consequence, the 1996 Housing Act included a provision intended to validate not apply to magistrates' courts where the vast evidence from non-victim witnesses. majority of criminal cases are heard. In many cases it may be difficult and expensive to adapt However chis only came into effect in existing buildings. February 1997, so it is too soon to determine whether or not this has had the intended effect. 2. Keeping witnesses on "standby"

An alternative and less expensive measure than • Expense: many of the cases involve a number of small incidents spread over a long changing the physical layout of all courts would period. To convince a court of the gravity of be to keep witnesses on •'standby". Instead of the situation evidence will need to be calling all witnesses to the court for the duration collected over a long period. of the case, witnesses can be asked to stay close to the court and given pagers to notify them when • According to the Local Government they arc needed. This was experimented with in Association, a "destructive cycle" may one area in the PRG study: as Maynard observes, develop '"in which complainants' although it would require some initial expectations are raised by the use of expenditure, it would be relatively inexpensive. professional witnesses, only to be dashed by the response of the court" (LGA submission, Aug 1997). All cases currently go through magistrate's courts: a small proportion (the more serious ones with a greater chance of hai-ina witnesses) art committed to the Crou 71 Court for trial.

123 While it might be impractical to call all witnesses • Reporting restrictions (such as hannmg m rhis manner, a would be appropriate where a publication of rhe witnesses name, address, witness is particularly vulnerable or has suffered photograph or any other identifying detail) inr.imidar.ion. Reporting restrictions are already available in 3. Screens, cctv and voice distorters certain circumstances, such as when the crime is a particularly serious one (eg. murder) or involves a Once the witness is in the court, another raft of young person. This provision could perhaps be measures serve ro pt event intimidation by extended to enable reporting restrictions obscuring the witness by screens, closed circuit concerning any witness who may be vulnerable to television (cctv) and voice distorters. The former intimidation (The Scotsman 3/1/97). measures allow judge, jury and counsel vision or the witness, but obscure that of other people • Identifying witnesses in court present. These measures serve a dual purpose. On one hand, they enable direct confrontation Under the Statement of National Standards of between witnesses and rhe defendant and the Witness Care in die Criminal justice System defendants' supporters in the courtroom TO be (1996: para 17.1), witnesses should not be avoided. On the other, they can enable the witness required w state their address in open court unless to maintain some anonymity to prevent this is necessary as evidence. Applications can also intimidation outside the courtroom. be made in exceptional circumstances (such as terrorist cases) to avoid the witness's name being Figures are lacking on how commonly these used in open court, providing certain guidelines measures are used, but anecdotal evidence suggests are adhered to. However no research was found they are rarely employed". This may in part be due examining how frequently such exceptions are to the preparation needed for such measures: in granted. I he idea of removing this requirement in the case of cctv and voice distorters the equipment some causes (in conjunction with using other needed is not available in every court. In addition, measures such as screens) runs against the idea the use of voice distorters for this purpose (as that justice should be conducted without secrecy opposed to terrorism cases) is very new: to enable public scrutiny. I lowever, perhaps more trepidation on the part ol the courts is fundamentally, it contravenes the principle that understandable in these circumstances. The Law the defendant should know the identity of his/her Commission (1997: 3.35) .suggests that there is ! accusers because this may effect his/her ability to also a perception that 'it is more difficult to tell a defend him/herself. This is a concern for many of lie about a person "to his face" than "behind his the measures discussed, and is considered in back'. The Law Commission agrees that it is greater detail at the end of the chapter. desirable that the witness can sec the defendant, if only because of tins perception. However, they 5. Reforming the defendant's right to cross- also suggest that other factors (such as the likely effect on the witness's ability ro give the best examine evidence they can) can outweigh chis. Ir has been suggested that the defendant's right to cross-examine witnesses be reconsidered in 4. Protecting the witnesses anonymity relation to rape. This followed media reporting of a case where a defendant was allowed to cross- Protecting the witnesses anonymity may take two examine a rape victim over several days and in a forms: manner rhat. allegedly would not have been deemed acceptable in prosecuting counsel (sec chapter 4). Cross-examination of a witness in such a man tier might constitute intimidation, and could be just as upsetting for a witness who has been subjected to serious intimidation. There may Phinikoff'and Woolfson (1995: 51) have examined the use of be grounds then for considering amending this these measures for child witnesses. They found that w links were right regarding other vulnerable witnesses such as used in about two-thirds of the child witnesses in their study intimidated witnesses as well as victims of serious and about one in ten had screens. However, a large proportion of applications were unsuccessful; for example applications were sexual offences. made for cctv in 63 cases but used in only 15.

124 6. "Friend in court" schemes health would be seriously endangered by a court appearance (Spencer 1995: 114)". In addition, Finally, as well as measures to discourage and deal until recently it was possible for a magistrate to with intimidation when it occurs, some authors receive a deposition from a dangerously ill adult". have suggested "friend in court' schemes. 7 hese 1 he Pigot Committee (Advisory Group on Video involve volunteers accompanying witnesses Evidence. 1989: 5.9) recommended that pre-trial throughout their attendance in court, from sitting hearings should be extended to both child and in waiting areas with them to sitting in the court vulnerable adult witnesses, from which the itself- I he witness service run by Victim Support defence should be excluded. As yet this has not and other organisations already provides this but been implemented. exists mainly in the Crown Court rather than in magistrates1 courts, where the majority of However, the Law Commission (1997: 194-203) prosecutions are dealt with. has recently recommended that the law on hearsay should be reformed. Automatic admission of Dealing with intimidation evidence from frightened witnesses was considered, but rejected on the grounds that this Intimidation can be dealt with by the courts in a might lead to some witnesses feigning fear to number of ways: avoid cross-examination (1997: para 8.58). Instead, the (Commission proposed that although 1. Admission of evidence the general ban on hearsay evidence should stay, exceptions to this rule (such as the admission of It was suggested earlier that one approach at the evidence from frightened witnesses) should be investigation stage might be to prepare for the changed. Recommendations included that (1997: possibility of intimidation, for example in tape- 1.40-1.41): recording interviews and witness statements. This strategy will only be effective however if such • statements should not have to be made to a evidence is admitted in court:. Written statements police officer or equivalent to be admissible; should now be admissible in some circumstances, but it does not appear that much use is being • the definition of fear should be clarified, to made ol1 this provision. One reason could he include fear of injury to others and fear of concern about the validity of written witness financial loss: in the past fear has been statements: approaches could include improving interpreted more narrowly; witness statements, routine tape recording, and extending the provision to cover other forms of • the courts should have discretion to admit evidence such as video-recordings. evidence from any frightened witness: ai present, evidence can only be admitted from Despite this, perhaps a more important those witnesses who fail to come to court, explanation is that admitting written evidence in and not from those who attend court but the absence of the witness denies the defendant change their minds about testifying, or [he opportunity for cross-examination. This change their testimony on the stand; and appears to contravene the European Convention finally on Muman Flights (F.CHR), which states that *'fe]veryone charged with a criminal offence has • the courts should be given discretion to the minimum right to examine or have examined admit evidence which is otherwise the witnesses against them" (quoted by Wolchover inadmissible but where it would be in the & Heaton-Armstrong 1997: see also Spencer interests of justice to allow it. 1995: 114). The Law Commission argues that if In some jurisdictions such as Holland (Clarke, implemented, these proposals would enable the 1994) this problem is avoided by allowing oral testimony from intimidated witnesses at a pre-trial hearing, where the defence has opportunity to Under s42 and s43 of the Children and Youtio Persom An cross-examine. There are some circumstances in 1933. which this can already be done in England and Under sl05 of the Magistrates Court-, Act 1980. According to Spencer 11995: 114) this was repealed by the Criminal Justice Wales, for example, for a child whose lire or and Public Order Act 1994 for no apparent reason".

125 evidence of frightened witnesses to be admitted However, the English courts do not have more frequently. The Commission maintains that equivalent powers to exclude the public when this does not contravene the European aduh witnesses are called. In Scotland rhe courts Convention for the Protection of Human Rights do have this power in relation ro alleged rape (ECHR), which allows evidence to be admitted victims (Spencer 1995: 1 14). This could perhaps where the witness is unavailable for cross- be extended to other vulnerable witnesses in examination. According to rhis argument, cross- England and Wales. examination is nor always necessary although it should still be possible to challenge hearsay 3. Penalties for witness intimidation evidence. Consequently frightened witnesses would still have to be identified, in addition, Under the 1994 Act, the maximum penalty is five judges would have a duty to warn juries of the years imprisonment and/or a fine on indictment, dangers of convicting on hearsay evidence alone. and six months imprisonment and/or a fine However, it is debateable whether judges would summarily. In practice, in 1996 or those convicted use their increased discretion to allow more such for either of rhe two offences, 48% received evidence to be admitted. immediate custody, 32% community sentences arid 12% discharges. It is too early to say whether 2. intervention by judge/magistrates eg. to this pattern is set to continue. In 1995. 61% of clear public gallery, warnings those convicted received immediate custody and 20% community sentences, but the number Once a trial is underway. Intervention by the convicted the following year was more than three judge or magistrate may be needed in response to times higher (internal note, 4/3/98). However, it intimidation such as the defendant's supporters may be worth monitoring sentencing patterns in staring at, gesturing or calling out to the witness, future. Further research on the characteristics of judges and magistrates have powers: these offenders and how senrencers view witness intimidation might also be useful. • to Issue warnings about the penalties for witness Intimidation; Penalties for intimidated witnesses

• to exclude poorly behaved members of the At present judges can and do imprison some public from me courtroom. witnesses who come to court and then refuse to give cadence for fear of reprisals. Failure to attend Again figures are not available on how frequently couu is punishable by up to three months these measures are used. One limitation is that imprisonment. Attending court and then refusing judges and magistrates may not always be aware to answer questions can attract up to two years. that intimidation is occurring, f'or example, Evidence of intimidation has to be taken into Henley (1995:4) suggests that intimidation could account, but does not preclude such penalties. involve a large number of gang members I his Is justified on die grounds that they are attending court wearing black (to symbolise necessary ro maintain die authority of rhe courts. death) and using coded hand gestures. Some jurisdictions in America use video-cameras raping This reasoning is questionable however when people coming into the court to discourage this applied to intimidated witnesses. It can be argued (Heaiey 1995: S). However ir is not clear how thar cases where scared witnesses are imprisoned widespread nor how effective this is. and offenders walk free produce the reverse effect: undermining rhe authority of the courts and In addition, the courts have powers: public faith in the whole criminal justice system. No figures were found on how frequently these • to clear the courr entirely in cases where a powers are used, but such events appear ro be rare. child has to give evidence of behaviour However, the publicity surrounding these cases "'contrary to decency or morality"11''. may discourage some witnesses both from reporting crimes and coming forward to assist police in their enquiries.

"Children and Ynuno Person; Art 1933, 37, quoted by Spencer Improving prevention, recognition and support to (1995: 114). witnesses throughout the criminal justice process

126 may help reduce the number of cases when Conclusion wimesses are too scared to attend or answer questions. I)e.spire this, rhe continued existence At present, very linle is known abour witness and use of .such penalties sends out negative intimidation. However, awareness of the problem is messages about the criminal justice system's growing, fuelled partly no doubt about concerns approach to witness intimidation. Possible that the problem maybe increasing. Further responses to this dilemma could include: information is important if we are to Identify cost- effective measures to prevent intimidation and • reducing the maximum penalty available for support those who suffer from it. failure to attend court and refusing to answer questions; and Despite the paucity of literature on the subject, a number of possible measures can be identified. • making both offences non-imprisonable These measures are summarised in Table 1. It where there is evidence of intimidacion. seems unlikely that any one measure would .suffice: different measures are needed ar. different d. ... and beyond stages of the justice system. To achieve maximum benefit, a package of complementary measures Often the criminal justice system's role is seen to tackling witness intimidation at all stages of the have ended after a trial. I lowever, the potential for criminal justice process may be needed. In intimidation does not end there. Even if the offender is convicted and imprisoned, there may addition, responsibility for dealing with witness be a risk of retaliation by friends and family or by intimidation could be clarified. Some police forces the offender upon release. The risk to the witness already have specialist units or officers to perform may be even greater than before, wirh an angry- this function. This approach could perhaps be offender .seeking retribution. This stage has mken further, by giving one agency responsibility however, been neglected by the literature, Only for coordinating measures throughout rhe three approaches TO deal with post-trial criminal justice process. intimidation have been suggested: In deciding the response, consideration will have to be given to the common concern running • New identities & relocation: sonic of the throughout the above examples: the preservation '"high level' police protection measures discussed earlier (such as new identities and of wimesses anonymity. In some cases the relocation) may be initiated at this stage, witness's identity is known to the of tender before particularly if the witness has not been the offence, and fewer protections will be allowed anonymity within the criminal available. I lowever, in many cases, the witness's justice process. However, such measures may identity only becomes known co the offender not be appropriate for the majority of through contacts with the criminal justice system. intimidated witnesses. The issue of anonymity raises a dilemma: denving witnesses anonymity can place witnesses in • Imprisoned offenders' use of telephones: danger, but granting anonymity may also make it concern has recently been expressed in the more difficult for defendants to prove their media that offenders can telephone victims innocence. There are at least two possible and harass them from prison (Guardian, responses co this dilemma: 4/9/97J. This suggests that monitoring of outgoing telephone calls from prisons might be tightened. • Option 1: accept the dilemma and identify ways to cope with it • Information about offender's release: intimidated witnesses or those at risk of The first approach is to ask at what point intimidation may find information about should the witness's Identity be revealed. offender's release arrangements valuable, The aim here is to delay the revelation until especially if this is backed up with other as Jate as possible to minimise the risks to support. the witness but still allow rhe defendant to use the information to assist their defence.

• Option 2: seek to resolve the dilemma

127 The second approach is to agree that the anonymity. This clearly depends on how many defendant lias the right to know the cases ol intimidation involve witnesses who arc- evidence against them, but 10 deny that it is not known so the offender, if this group is in die always in the interests of justice that the minority then focusing on protecting anonymity witness's identity be revealed. In some cases may nor be the most effective way of [adding the the witness's identity may be important to problem. Although existing research supports this, the defence, if. for example, it can be shown the evidence is limited: further information is chat they have a grudge against the needed to ensure that any resources devoted to defendant and might therefore have reason witness intimidation are used as effectively as to perjure themselves. The argument is that possible. if a witness is intimidated against testifying, or changes their testimony because of It seems likely that whatever measures are intimidation, then justice has also failed to implemented to encourage reporting and be done. There may then, be grounds for identification of witness intimidation, some cases creating exceptions to allow witness's will always fall through the net. Nevertheless, anonymity to be preserved throughout. efforts can be made to ensure that the holes in the net are as small as possible. This is a difficult area. It is worth therefore evaluating the importance of the issue of

128

130 Section 3: Disabilities and Illnesses

The Problem Box 1: The number of mentally and physically vulnerable people in the UK It is widely agreed that a person's physical and mental health and abilities may influence their Learning disabilities: according to Mencap, there experience as a witness. The literature are over a million people with learning encountered focuses almost exclusively on disabilities in the UK, of whom about a fifth intellectual (Inabilities, and the discussion below have severe learning disability (submission to reflects this. However, mental illness and physical Review, 27/8/97). disabilities may also make some people particularly vulnerable witnesses, and are Physical disabilities: according to the 1988 considered wherever possible. Many of the issues OPCS (Office of Population Censuses and discussed are relevant to all three groups, whereas Surveys) Survey of Disability, there are 6.2 others arc distinct. Tn addition, it is worth noting million people in England and Wales with that they are not mutually exclusive. Physical and physical disabilities. intellectual disabilities can be associated, although they do not always accompany each other. Less Menial illnesses: according to the Psychiatric commonly recognised, intellectual disabilities and physical disabilities may at some point be Morbidity Survey (Report 1 1995. Table 6.1), accompanied by mental illness. When a witness about one in five of the population living in has more than one of these conditions, they may private households reported suffering some be especially vulnerable. psychiatric disorder. These included phobias and depressive episodes, and drug and alcohol dependence2. Definitions I he definition used (or the literature review was a The definition that the Review and any broad one: any physical or mental condition that subsequent legislation adopts will be important in could render a witness particularly vulnerable. influencing which groups receive special Both disabilities (normally permanent} and assistance. There are several possible approaches: illnesses (usually temporary and treatable) tail within this definition. 1. Listing the conditions eligible for special t reartmen t.

This would have the advantage of clarity: criminal justice practitioners would be left in no doubt The term intellectual disability are used loosely. There are about whether n particular witness qualified under numerous names for such conditions, including mental handicap andretardation, learningduahility, developmnnaidisability, these terms as vulnerable. It might also be possible and intellectual impairment (which may be used to cover both to say which measures would be helpful for each mental disability and illness). A variety of names are used group. However, diagnoses may differ between throughout the report. However, the terms can carry practitioners and change over time. Listing all political connotations: mental handicap and retardation are eligible conditions could also be cumbersome. A both rety old-fashionedattd now seen as stigmatizing: these are therefore avoided. Mental disability is similarly a little old- very long list of eligible conditions could be fashioned, but is used here to enable discuaion of "mental envisaged, and it would need occasional updating disabilities, and illnesses"as opposed to the more long-winded to reflect developments in medical science (such "intellectual disability and mental illness ". It should also be noted that although some of the terms may imply more serious conditions than other. For example learning disability might be thought to cover dyslexia, whereas mental handicap may conjure About 16% reported "neurotic disorders' 'such depressive up image of much more serious handicap. They are used here as episodes within the past week. A further 2% reported broadly and are not intended to refer simply to one end of the scale. "functional psychoses" mch as drug or alcohol dependence within the past year. (OPCS Psychiatric Morbidity Survey Table 6.1).

131 as the identification or new conditions, or 3- Assessment of individual needs. improvements in understanding that might alter what is considered an appropriate response). lr The third approach also involves assessing addition, some criteria would be needed to decide individual cases, but leaves the criteria of which conditions to include and which to vulnerability open. For example, Sanders et al exclude. suggest that the individuals needs should be assessed, preferably by experts. This approach 2. Derailing one or more tests to determine again recognises that the impact of disabilities and whether witnesses qualify. illnesses varies greatly between individuals, but it would be more useful for identifying appropriate An alternate approach is to specify one or more measures for each case. It would, also enable tests to decide a particular witness's eligibility witnesses with disabilities/illnesses to be assisted for special treatment, such as tests of mental without singling them out from other groups age. memory, and reliability. Assessing needing special help. It is not without problems individuals could be fairer than deciding though. If the assessment is not based on cleat, eligibility on the basis of group membership: common criteria or guidance, people with similar individuals with the same disability or illness needs could be treated very differently. (This may be very different in terms of memory skills approach is discussed further in part 2 - "the and the like. This approach has been response"). recommended by the New South Wales Law Reform Commission (1996) in relation to A final consideration is what definitions have been intellectual disability - .see Box 2. used by past legislation. There is an argument for consistency: however, it may also be argued that Box 2: New South Wales Law Reform different areas of social policy and different Commission (1996, paras 3.2 & 3.20) contexts can require different approaches.

" 'Intellectual disability' means a significantly The impact of disabilities and illness below average intellectual functioning, existing Disabilities and illnesses can create two kinds of concurrently with two or more deficits in vulnerability: adaptive behaviour". • first, vulnerability to crime as victims: and Three possible "adaptive deficits" are listed: • secondly, vulnerability as witnesses assisting • level ot communication; the criminal justice process. • social skills; and 1. Vulnerability to crime. • ability to live independently. I he precise extent ot these groups* vulnerability to crime is not known. Official statistics on criminal This does have the advantage of barm concise. proceedings and convictions do not contain The Commission also claims ir is specific details of whether witnesses (or for that matter enough to prevent someone feigning a disability suspects and defendants) have disabilities or to gain an advantage (although it does not illnesses. Even the British Crime Survey, which provide any evidence that this actually occurs covers unreported crime, does nor. cover people in with other definitions). There are however, some residential homes, psychiatric wards or care. problems with such an approach. For example, the cut-off point where functioning becomes However there is some Australian evidence (Wilson "'significantly below average" is vague and can be 1990, cited by SSW Law Reform Commission drawn differently by different people. The 1996 para 2.23) that people with an intellectual inclusion of two or more deficits in adaptive disability are twice as likely to be victims of a behaviour also seems problematic: the number personal crime (eg. assault), and one and a half seems arbitrary, and the definition of an times more likely to experience a property offence. adaptive deficit also seems to suffer from the It seems reasonable to assume that this greater cut-off point problem. vulnerability to crime exists here as well as in

132 Australia. Similarly Sayce (1995: 141) suggests that Even if an offence is reported to carers mentally ill people in this country are particularly and/or the police, it may be met with vulnerable. Temkin (1994: 402-403) lists a number scepticism or may not be perceived as of reasons why people with disabilities may be criminal. Temkin argues there are several particularly vulnerable. Although she focuses on myths which contribute to this. For disabled (both physically and mentally) children example: some of these factors will also apply to disabled adults and some people wirh mental illnesses: - that "no adult would be so callous as ro abuse a disabled child"; • Unable to run away, easy to overpower. - that abuse is committed by strangers and not carers; People with physical disabilities may be - that they may have misunderstood or physically less able to resist and fight off an invited the offence; attacker, making them easier targets than the - that they couldnt be the object of sexual able-bodied. attention; and - that people with learning disabilities • The numbers of those involved in care. commonlv tell lies or fabricate stories.

Temkin argues that the sheer numbers of Sayce (1995: 143) says that people with people involved in caring for the disabled mental illness may have their allegations increases the risk of abuse, whether they live explained away as delusions. with their family or in a residential home. • Societal attitudes. • Dependence. Finally, Temkin argues that disabled people People with disabilities may depend on are treated as second-class citizens by society, other people more than usual, creating and mav be singled out as soft targets as a 'learned helplessness"' which makes it more result. Sayce (1995: I4l| argues that the same difficult to resist psychologically. problem afflicts psychiatric patients. On the former, it lias been suggested that such • Recognition of abuse as such by the victim. prejudices tend not be challenged because the people who hold ihem rarely meet those with When people are physically dependent on disabilities (see Cohen 1994: 21). others, ideas about bodily integrity (for example that other people do not have the It has also been suggested that right to louch them) may be alien to them. deinsdtutionalisation under the "care in the Similarly, Temkin suggests that disabled community'' policy has led to over-representation people are sometimes poorly informed about of mentally ill people among the homeless, thus sexual matters and are not aware of ideas of increasing rheir vulnerability to crime. Walker sexual choice. (1992: 177-200, 124-135) supplies some evidence that deinstittitionalisation has increased the • Difficulties communicating offence to carers numbers of mentally ill people among the and/or the police. homeless in America, and some limited evidence that police contact with mentally ill people in Physical problems, language problems and England may have increased in recent years (see- lack of knowledge of where to complain or also Palmer 1996: 635). However more research is how 10 report may hinder communicating needed to ascertain whether mentally ill people victimisation to others. In addition, some really have become more vulnerable to crime as a (eg deaf children) may lack self confidence result of deinsrirutionalisarion. and self esteem, and may therefore be more open to efforts to dissuade them from 2. Vulnerability as witnesses assisting the reporting. criminal justice process.

• Recognition of abuse as such by carers and Health and abilities also affect people's experience the police. of contacts with criminal justice process and their

133 performance as witnesses. Sanders ei al (1996a: 2; areas of personal functioning may also be an issue 19%b: 7-9) identify three main areas of personal for people with physical disadvantages. For hmaioning which can be affected by learning example, people who lack or have partial hearing, disabilities: speech, or sight may suffer communication problems. Even people whose physical disabilities i. Memory- some people with learning are nor an obvious impediment to the four areas disabilities can take longer to absorb, of personal functioning identified may be affected comprehend and recall information. Recall or by medication or pain, in addition, some people details may be particularly effected. Greater with physical disabilities may be affected in a time and patience than usual will help. fourth area: mobility.

ii. Communication skills- they may have a A fifth and final area which can be identified is limited vocabulary and remember things in social skills. People with learning disabilities may pictures rather than words, leading to have more limited life experiences: for example difficulties in understanding and answering some may lack experience of paid employment, or questions. '1 hey may also find it difficult ro of financial responsibility. They may not therefore explain things in a way other people rind have the same level of social skills as other people. easy to follow. On the other hand, social skills may be more devebpedthan other areas of personal iii. Response to perceived aggression - Sanders et al functioning which may lead strangers to over- suggest some people with learning estimate their other capabilities. disabilities are especially sensitive to negative emotion, and may be suggestible. They may Of the five areas, communication skills are respond to tough questioning by trying to probably the mo.st important in a legal culture please the questioner. For responses to be which 'takes oral communication for granted and reliable questions should be kept simple and relies heavily upon it'' (Temkin, 1994: 402). non-threatening. Tern kin distinguishes two categories of (both physically and mentally) disabled people according Sanders et al (1 996a, 2; 1996b. 9) observed that to their communication skills (see box 3). few learning disabled people have all these problems, arid some may be advantaged in some respects. For example although many people with autism have communication problems, they may also have better than average memories.

The third of the.se areas could be expanded to cover emotional resilience more generally. Sanders et al f ] 996a: 9) suggest that witnesses with learning disabilities can find being a victim of a crime and contact with the criminal justice system particularly stressful, and that one reason for this could be learned helplessness (see above), ft seems plausible that this may be exacerbated by lack of confidence, and low self-esteem coupled with problems with the three areas listed by Sanders et al which may cause frustration. This suggests that it is particularly important that the criminal justice system responds sensitively and appropriately to their needs.

No similar research was found by the literature review for people with mental illnesses or physical disabilities. Despite this, it seems plausible that some similar problems may affect people with mental illnesses such as clinical depression. These

134 Box 3: Categorisation of disabled resilience, mobility and social skills) can affect the children (Temkin 1994: 405) criminal justice process. These problems are discussed for each stage of rhe criminal justice Category 1: process in the next section.

• "rho.se able to communicate orally, albeit with some difficulty", The Response

• "those who can do so with assistance from a. Reporting other complementary systems", and Several studies suggest that a large proportion oi sexual crimes against people with intellectual • "those who may be described as alternative disabilities are unreported ro the police (see communicators who can communicate Sanders et al 1996: 1 5). It seems plausible that effectively using some system other than this also applies to property offences, and to some oral communication". people with physical disabilities or mental illness. Four issues are discussed below: Examples given include: these with spina bifida and with learning disability, deaf people fluent • recognising that an offence has occurred: in sign language, and the partially deaf who are able to speak. • encouraging reporting;

Category 2: • facilitating reporting: and

• "those who are unable to communicate • identifying vulnerability. effectively, cither orally or by using some alternative system". Recognising that an offence has occurred

Examples given include: most deaf blind, some Clearly before a en me is reported It has lo be with language disorders, and some with cerebral recognised as such by the victim or witness. It was palsy who cannot speak, use computers or noted earlier that this may be problematic for communication boards. people in positions of physical or emotional dependence, or if they have not been educated Although her focus is on children, the categories about such matters. This suggests several possible seem equally applicable to disabled and mentally measures: ill adults. As Fern kin comments, whatever changes are introduced, the criminal justice system is • improved "crime education" ro cover unlikely to be able to offer much to those in the reporting crime as well as teaching right second category. She suggests chat the best way to from wrong'; help these people is to improve alternative communication systems, for example by teaching • informing service professionals about basic British Sign Language more widely. However, law: for example that staff confining people there is also a small group for whom "effective to their rooms against their will may be communication will never be a realistic guilty of false imprisonment, and that possibility" (Temkin 1994: 406). Equally it seems unwanted touching may be assault'; and likely that there are some sufferers of mental illness who may never be able to give effective testimony, either because of communication • writing formal policies for professional problems or because of deficits in some other area carers and care institutions to tespect the of personal functioning.

In the discussion that follows the focus is on the Williams {1995a: 2} quot a victim of racial harassment who first of Temkin's two categories. Although Temkin es had learning disabilities: "At special school I was taught not to focuses on communication difficulties, pinch other children's sweets and money. I was not taught, if I disadvantages in any of the five areas of personal am in trouble, to tell the police". functioning (memory, communication, emotional See Williams (1995b. 2) for further examples.

135 bodily integrity or rhe people "they have care reported by A third party" (cited by Sanders et al of, for example by asking permission before 1996b: 1"). Sanders et al (1996b: 17) found figures moving or touching someone. of 78% for personal crimes and 66% for property- crimes. The decision of the carer (or other third These are not measures that rhe criminal justice party) on whether to report ro rhe police is therefore system can institute alone: To gain the cooperation extremely important. of other agencies, for example in encouraging disabled/ill people co report, ir may be necessary However it has been suggested that a non- to convince them that they will be treated reporting ethos exists in Britain (see for example sympathetically and their special needs catered for. Williams 1995b: 55). There are numerous reasons Nor are all these measures straightforward iti why offences may not be reported to the police. themselves: sex education for example may be Some of these concern the perceptions and beliefs controversial for some people with severe held by the person who initially learns of the intellectual disabilities. offence:

The perceptions of cirers (who may be the first • failure to define an incident as contact) and the police, particularly scepticism criminal/scepticism that it occurred; that the incident occurred or was criminal, also need to be tackled. Williams (1995a, 1995b) • the ethos that all discussions with clients conducted a two-year study or. victims with should be treated confidentially; learning disabilities, involving interviews with victims, carers and police officers. He suggests • belief that the victim does not want the that the language used by social service offence to be reported to the police; professionals can disguise the criminal nature of incidents: • belief that the offence is too trivial or that the police could do little; "Women with learning disabilities are more likely to be described as being "sexually abused' • concern that the police and the rest of the than raped; men with learning difficulties arc criminal justice system will be 'physically abused' not assaulted; stealing unsympathetic, not meet the something from someone with learning victims/witnesses needs for special care and 1 difficulties is 'financial abuse , not theft. not produce a conviction (ie that it wouldn't Offenders against (he general community are be worth the trouble); and criminals, those who victimise people with learning disabilities are "abusers ; victims with • concern about the offender: m some cases learning difficulties are 'survivors and the offender may be a colleague or another 'sufferers'; 'sufferers' do not report crimes ro ihe client, creating (misplaced) conflicts of police, they 'disclose abuse' to professionals." loyalties. (Williams,* 1995:2). Other reasons for not reporting concern the Again, this might be dealt with through education organisation: to raise awareness of the increased vulnerability of these groups and ro tackle the myths discussed • lack of formal reporting procedures, creating earlier: this could be approached in the through uncertainty about whose responsibility this training, guidance, newsletters, leaflets and is: posters. • complex reporting procedures: Williams Encouraging reporting (1995b: 66) found that there are often long linear reporting chains, whereby one person It was nored above that when people in institutional refers the matter to his/her superior, who care do report a crime, in the first instance many refers it to his/her superior and so on, each may make the report to a carer rather than to the person perhaps consulting others along the police. Wilson and Brewer (1992) found that 56% way. This may deter or delay reporting, and of personal crimes and 63% of property crimes if one link fails the case may fall by the against people with learning disabilities were wavside; and

136 • internal investigative: procedures: in some There are advantages m drafting formal organisations, an internal review may be policies for the organisation as well as for the carried our to determine whether a criminal victim and the criminal justice system: or (if a starr member is involved) disciplinary offence has occurred before ''policy guidelines are useful for reassuring reporting ro police. This may hinder clients and their carers that claims of ill- collection of viral evidence, as well as treatment will be taken seriously, .\m} for increasing the number or occasions on establishing a base-line of good practice which victims and other witnesses will be against which complaints or allegations or required to tell their stories. negligence by staff can be measured" (Sanders et al: 1996,21). Some or these reasons may be legitimate. For example if the victim was aware that the incident Although such an approach cannot was criminal, that recourse could be available guarantee good practice it might encourage through the criminal justice system and that the it. person they reported to (or someone else) could report this to the police on their behalf, but did • Creating a legal requirement on service not want to pursue the matter. Sanders et al professionals to report all alleged cases of (1996b: 23) quote the policy of Cumbria social sexual abuse. services: ''In most situations the rights of the individual would be respected but there will be A legal requirement on professionals ro situations in which agency staff will be bound by report all alleged cases of sexual abuse has their professional ethics or agency contract to been suggested by Cervi (I 992: 1 5). This disregard the individual's, wishes''. Deciding could perhaps be extended to cover other whether to report may not always be a simple offences such as crimes of violence and matter. property crime, framed, to take into account the victim's wishes. 1 he key question is Some measures were discussed above concerning whether such a requirement would work. the definition of incidents as criminal. Possible Some system of monitoring reporting approaches to combat the other problems listed practices and sanctions to impose for failing include: w observe she requirement would be needed to ensure adherence. 1 lowever, it is difficult • Creating/reviewing formal policies for to envisage an effective monitoring system: investigating and reporting incidents. service professionals would hardly have an incentive to keep records of disclosed crimes Ideally, an internal investigation should not they did not repon. It might be necessary to delay reporting an offence to the police. A ask the people in their care whether they policy of joint investigation by the police had been victims of crime over a set period and the organisation concerned might and whether they disclosed this to staff, and overcome such problems and also limit the then to check that it was reported. number of times the victim and other witnesses have to be questioned. Formal • National guidelines for professional carers polices could also contain a presumption in and care agencies on reporting. favour of reporting and specify what might constitute a legitimate reason not to report. A less radical alternative would be to create This could perhaps be backed up by non-statutory national guidelines on disciplinary sanctions. Responsibility for institutional reporting. This might reporting should be clearly defined. In encourage consistency in reporting policies, addition, MENCAP (1997: 3) recommends and would not require the monitoring and that "web-like" structures be used instead of enforcement arrangements of a non- linear structures ro maximise opportunities statutory system. for offences to be reported. Finally, there is some evidence that even minor crime can be traumatic for some vulnerable victims (Sanders et al 1996b: 9). Sanders et al

137 found chat this can create a disjunction between equivalent skills to enable communication. I here the expectations and perceptions of the victim are two ways of tacking this: (who sees the crime as serious and expects the police response to reflect rhis) and those of the • using interpreters, cither volunteers or paid police (who see the offence as minor and treat it for by the police: and as such). People with learning disabilities arc taught to trust people in authority to act in their • training police officers in alternative best interests: if they find rhis crust misplaced. communication systems. they may be deeply upset. Sanders et al argue that this disjunction in attitudes undermines The first opiion would appear to be the easiest. confidence in che police, and discourages future However, it may be difficult to find interpreters reporting. This suggests that sympathetic for even the most common alternative treatment (in which time is taken to acknowledge communication methods. Temkin's research the victim's feelings and to explain the processes) (1994: 40) suggested that there were only: may be particularly important not only by the police but throughout the criminal justice system. • 12 lip-speakers in England trained to apptopriate level for this type of Facilitating reporting interviewing, 1 in Wales. 1 in Scotland, and none in Northern Ireland; and Even if people with disabilities/illness want to report an offence, there may be obstacles which • 39 fully-qualified sign freelance language prevent or deter them from doing so, or mean interpreters in England, four in Wales, 1 in that they have to rely on third-parties to report for Scotland, and 1 in Northern Ireland. them. Crimes arc usually reported to the police by telephone or in person at a police station. In some As a consequence of the scarcity of interpreters, it cases the obstacles are physical. For example could take some time to arrange one. They may people with impaired mobility and wheelchair then need specialist advice about the person's users may find physical access to some police particular disability/illness, and might also need stations difficult. 'I hese kind of obstacles mav be specialist equipment. No figures were found on remedied by changes in the physical design of the use of interpreters and lip-speakers by the police stations (such as the installation of ramps police or any other pan ol the criminal justice and lifts) and provision of suitable transport to system. It seems likely that these additional and from stations where necessary. requirements may discourage the recording and investigation of cases involving disabled victims, Other problems are less straightforward. Some and also discourage the use of disabled witnesses. disabled people (such as many people with cerebral palsy) may have no or limited speech The second option would help overcome these f iemkin 1994: 404), which could make reporting availability" problems, even if only a handful of by telephone impossible and reporting in person officers were trained in alternative communication at a police station difficult. Various alternative systems. Again, this would need to be backed up communication methods have been developed to by information about how to interview people deal with some of these problems of with disabilities/illnesses. communication. The most well-known are probably sign language' and lip reading and are Identifying vulnerability used by some deaf people. Other methods include picture cards, communication boards (which Of course, to communicate effectively with the contain pictures and symbols he person can point t witness, the police may need to identify that the to) d computers. an witness has a disability ov illness that might be relevant. This can be a difficult task. Physical However, even when they do attend a police disability can be immediately obvious, but this is station to report, the police may not have not always the case: for example arthritis can be severely disabling, but is not usually apparent from the sufferer's appearance. Similarly the New There is more than one type -for example British Sign Language and Sign Supported Language are quite different South Wales Law Commission (1996: 55) states, (Temkin 1994; 404). intellectual disability is "not., necessarily obvious

138 from a person's appearance", and the same is true contacts between the police and groups involving of mental illness. However with mental illness the learning disabled should be increased. there are .some additional complications. The illness could develop during the criminal justice Another approach is suggested by Sanders et al process. In addition, treatment could mean thai (1996b: 25-27). Their study included a survey of the witness is fine when the crime is first reported, police forces and found char none of those who but char their condition deteriorates later on. responded had guidelines to help officers tell whether victims had learning disabilities. Some Studies of police contacts wirh learning disabled referred to guidelines on the treatment of suspects suggest that police are not very good at mentally disordered offenders (under the Police identifying them as such. As Sanders er al argue, it and Criminal F.vidence Act 1981), but as Sanders seems likely that the same is true of victims and et al point out, these onlv outline what should be non-victim witnesses. I fowever, Walker studied done once rhe disability has been identified and police reported contacts with mentally disordered nor how to identify this in the first place. Some people, and suggests "police officers in the United efforts have been made to resolve this problem in Kingdom receive virtually no training in the other countries such as America and Australia. recognition and management of mental disorder" However, Sanders et al argue that (as Gudjonsson (1992:226). ct al 1993 observed regarding suspects), whatever tests are used there will always be some people Under the Police and Criminal Evidence Act who are not identified. They argue that some 1984, the police have to call a police surgeon if people with learning disabilities can cope with the they think a suspect may have a mental illness criminal justice system as it is. Instead of trying to (although officers may call one out when the identity people wirh learning disabilities, they suspect has a learning disability because of suggest: difficulties in distinguishing symptoms). One measure misj.lit be to extend this requirement to "the aim should be to identify those witnesses, but there are also problems with this characteristics which tend co be associated with approach: learning disability (poor memory, communication, increased irauma etc.], bur • police surgeons are nor always easily which may be present in other people too. This available, which may cause delay (Deloirre would enable criminal justice agencies to focus and'louche 1996: 3-3: 4.1); on the identification of these characteristics, rather than co allocate people to categories...It • according to Palmer (1996: 637) not all will is the communication problem that should be have experience in mental health and they identified and addressed. Identification should may only have had basic training in the area be seen as an end in itself"" (Sanders et al. during their medical qualification; and 1997:16). finally Hence instead of focusing on recognition of • most police surgeons see their primary role particular disabilities, they advocate that guidance as deciding whether the suspect is fir to be and training should focus on identifying detained or charged. characteristics associated with vulnerability and. how to handle them. This approach is supported Walker (1992: 201) found some evidence that the by the Law Society's Sub-Committee on Mental police, the medical profession and people with Health and Disability (submission to the Review. mental disorders believe improved training is 11/9/97). needed. He argues that improved training should be given to all new and serving police officers, but b. Investigation recognises that a more practical approach would Once a complaint has been reported officers have be to start by targeting supervisors (usually 1 co decide whether co investigate. Sanders et al sergeants). Similarly, MF.XCAI (1997: 4-5) have (1997, 17) suggest that communication suggested there is some desire among the police difficulties can be used as a reason not to pursue for greater training. They recommend that the case further. In their study many of the cases training should be introduced for all new recruits, reported to the police were either "no crimed" (ie with compulsory refresher courses, and that

139 not recorded a.-, crimes) or led ro no further action suitable locations and a gteater likelihood of (Sanders et al 1997: 6). This research was based interruptions. on case studies, so U is important ro note that if might not be representative. However it does The literature review did not uncover any accord with Williams (1995a: 3) research in this information on police practices or policy in this area. Some ways of overcoming communication area. problems were considered earlier in relation to reporting, btit apply equally here. 2. Who else should be present

When complaints are investigated important An associated issue is whether mentally vulnerable considerations include: witnesses should be accompanied during an interview. Under the Police and Criminal • the location; Evidence Act 1984, police must arrange an '"appropriate adult' to accompany any suspect • who else should be present; and who they believe may have a mental illness or handicap. The role of the appropriate adult under • how the interview should be conducted this scheme is threefold: (including responsibility for interviewing and whether the interview should be • to advise the suspect; recorded). • to assist communication with the suspect; I. The location and

The location where interviews are conducted is • to ensure that interview is earned out fairly. important for both mentally and physically vulnerable witnesses. There are two main issues: This can be carried out by n parent, guardian, accessibility and comiort. other carer or someone with experience in mem al health (eg. an approved social worker). • accessibility: the location for the interview Under the revised Codes of Practice issued in needs to be accessible for people with 1995, this excludes a solicitor representing the mobility problems. I his needs to be built suspect. into the design of police buildings- including special interview suites built in It has been suggested that this scheme should be some police forces for children and victims extended to mentally vulnerable witnesses (for of sexual offences such as rape. Regarding example by the Law Society Sub-Committee on the latter, Temkin (1994: 416) claims that Mental Health and Disability op eit, para 10). "all too often they are situated in locations However, there are problems with the current where access for the disabled is hard or scheme. First, it relies on officers to identify impossible'- vulnerability. As observed earlier, this may be a difficult task. Adults with learning disabilities or • comfort: one of the aims of special interview mental illness may be reluctant to make this suites is ro help pur certain groups of known to the police, and symptoms of mental vulnerable witness at ease in a stressful disability/illness may be lacking or attributed to situation. Of course, providing similar consumption of alcohol or drugs. To encourage facilities or adapting existing facilities for the use of appropriate adults in cases where mentally vulnerable witnesses mav be there is some uncertainty about the suspects' expensive and rime-consuming. One means vulnerability, and as a safeguard, any confession of bridging the gap may be to interview made without the presence of an appropriate learning disabled witnesses in their own adult may be excluded from court. Nevertheless, homes or (if the offence happened there) in there is some evidence that appropriate adults another similarly familiar and comfortable environment. However, any cost-savings accrued by such an approach would have to Juveniles are also covered by the provision. In this case, the be balanced against problems in identifying appropriate adult can be a parent, guardian or social worker. If none of these are available, a responsible adult aged 18 or over who is not employed by the police may perform the role.

140 are not always called for mentally vulnerable 3- Responsibility for interviewing adults, even in ca,se.s where custody officers are aware of the disability/illness (Palmer 1996: There is some evidence that special interviewing 640-641). skills may be needed for some mentally vulnerable witnesses. The importance of treating witnesses In addition it has been argued that appropriate sympathetically was mentioned earlier, adults do not gain enough experience to develop particularly of recognising the potential real expertise (Sandell 1992: 18). There is also importance of apparently minor incidents ro concern about the confidentiality of any vulnerable witnesses. Repetitive questioning may comments made by the suspect to the appropriate give mentally vulnerable witnesses the Impression adult. It is unclear whecher an appropriate adult that they are not believed by the investigating has a duty to inform police of any potentially officers, or that they are not listened to properly useful or important information the suspect tells and given the attention they deserve. them ot whether they should keep thi.s confidential (Littlechild 1993: 15-16). Finally, it The use of props (such as toys) is well established may be difficult to find an appropriate adult, in interviewing child witnesses, particularly causing delay. Parents and guardians may not be suspected victims of sexual abuse. Using leading willing to perform the role, and other carers and questions and prompting carefully have also been social workers may have other commitments for raised as a possible interview technique for child their time (Palmer 1996: 641). witnesses. Such approaches may be appropriate in interviews with some mentally vulnerable If the appropriate adult scheme is extended to witnesses, but could be a source of problems in mentally vulnerable witnesses, these problems court. The need [or special interview approaches need to be taken into account. Palmer 0996: may also be difficult to reconcile with the need to 635) suggests several measures to improve the gather evidence. For example, although a single scheme for suspects, which might equally be of interview might limit the stress the witness benefit if extended to witnesses: undergoes, it might also fail to reveal evidence that will only be fully disclosed over a period of • specifically asking the witness if they need time (Temkin 1994: 416). help; Despite the apparent need for specialist skills, the • asking the witness which school they evidence suggests interviewing officers usually lack attended (to help determine whether they them. Sanders et al (1997: 20) found that have special needs); interviews for sex offences were usually conducted by officers from force family/child protection • familiarity with common medication and units: these officers will not necessarily have the conditions they arc used for; appropriate skills for interviewing mentally vulnerable adults. Xon-sex offences were usually • greater use of civilian custody officers or dealt with by officers without specialist training in regular custody officers to improve working with any vulnerable witnesses. Sec against experience; this, social workers who have expertise in working with these groups tend not to have investigative • basic training for all police officers in skills. identifying special needs. The Law Society (submission op cit para 16) These measures all apply to the first problem - recommends that interviews should always be identification. The expertise and role of conducted by officers with appropriate training appropriate adults are also important who should have access ro communication aids considerations. For example, appropriate adults and specialist advice and support. To meet this for witnesses could perhaps have an additional recommendation, specialist training will be role in giving the witness emotional support. needed for some officers, and consideration will have to be given to how specialist support can best be organised.

141 Another approach would be to conduct joini to (Temkin 1994: 408). 'lenikin suggests that investigations, using a team or both police officers special guidance on video-techniques for interviews and others with expertise in mental health. To be with disabled people is needed. In addition, ir effective, such an approach would need ro be might be useful to improve training and guidance supported by information about the advice and on writing statements, perhaps incorporating support available from other organisations: tor advice from experts on learning disabilities. example, lists of qualified sign language interpreters (Tenikin 1994: 407). c. The decision to prosecute It has been suggested that an excessive number of Special skills are also needed in taking witness cases involving witnesses with disabilities are lost statements. Statements are usually drawn up by at this stage. This is given some support by rhe police once the interview has been conducred research in one health region by Hillary Brown and signed bv the witness. Statements are useful (cited by Cetvi, 1992:14: Cohen 1994: 167). Of 167 cases of alleged sexual abuse against people with learning disabilities reported between 1989- • to help witnesses refresh their memories at a 90 only 10-1 5% resulted in a court appearance later date (they are usually given to witnesses despite three-quarters being accompanied by outside rhe courtroom), and corroborative evidence. One factor may have been that half were allegedly carried out bv people with • to check consistency with oral testimony at learning disabilities. Williams (1995a:2) suggests trial, as evidence of the witness's reliability. that the CPS is discouraged from prosecuting offenders with learning disabilities, whose victims However, it was noted in chapter 2 that the are often people with learning disabilities. statement is not a verbatim record of an interview. The information included is selected by the Other reasons ior deciding not to prosecute investigating officer who may inadvertently omit include: information that subsequently proves important or introduce errors or distortions (Wolchover & • lack of evidence following delayed reporting Armstrong Jones 1997: 855-856). If it is difficult or internal investigations by care fot ordinary adults to recognise the account organisations (a number of measures ro created as their own, this may be an even greater tackle these problems have been considered problem for witnesses with disabilities or mental above); illness. Furthermore, some people with disabilities or illnesses may have communication problems: credibility: similarly, people with disabilities some people with learning disability may have • may not be seen as credible witnesses limited or no literacy skills (Mencap submission (several myths which may contribute to this to the Review 27/8/97). This may make it were discussed earlier: training and guidance difficult or impossible for them ro read and check may help tackle them.); and the witness statement or use it in court, but this may not always be drawn to the officers attention. • competence: it has been suggested that people with learning disabilities are Several measures can be envisaged to help overcome automatically regarded as incompetent. such problems, in particular increased tape- recording or video-recording of interviews with Sanders et al (1997: 35-36) found that although mentalJy or physically vulnerable witnesses (rhe the police often liaised with the CPS before this latter is recommended by I-aw Society, op cit). stage, the CPS and the police rarely consulted These measures were discussed in chapter 2 on experts about learning disabilities. This also seems witness intimidation, but the main arguments to apply to illness. Most cases do not reach the apply equally here. Nevertheless, additional issues prosecution stage, particularly when it may not be are raised by video-recording for people with in the public interest to prosecute cases involving disabilities. Special Techniques may be needed vulnerable witnesses (for example, because the where alternative communication methods are proceedings would be too traumatic). used: for example split screen filming to show simultaneously the interviewees face and the communication board the interviewer is pointing

142 Another issue is what evidence the decision to A number of measures originally designed or pro.secute is based on. Williams (199%: 72) proposed for other vulnerable witnesses (such as reports that the CPS can take this decision children and victims of sexual offences) may be without meeting the witness. However the validity appropriate for witnesses with disabilities/illnesses. of'witness statements us the witness's own account For example: is questionable. The problems of using witness statements as an indication of the ability to stand • pre-trial preparation (a pre-court witness up in court, were and the suggestion char video- pack has been produced by the campaign recordings, audio-tap ing; or verbatim transcripts group Voice for the Home Office); be used instead. This is supported by MF.NCAP (1997: 7) who recommend compulsory video- • friend in court schemes; taping of all interviews with learning disabled people. • • the removal of wigs and gowns;

Other possible measures include: • design changes in the architecture of court buildings (for example to make them more • a requirement for prosecutors to meet victim accessible wheel-chair users and people with before they make a decision (perhaps mobility problems); accompanied by changes to police statement writing); • screens and cctv;

• guidelines or information for prosecutors to • pre-trial hearings and admission of written assist their decision-making: the Law depositions; and Commission reports that the CPS has an internal memorandum providing such • clearing the public gallery. guidance that the Law Commission has argued should be published (para 19); Some of these measures (such as pre-trial information, removing wigs/gowns, screens And • CPS training (MF.NCAP 1997: 7). A cctv) have been recommended by MF.XCAP variation on this theme would be to have (1997: 9). They also suggest some people with designated and trained Crown Prosecutors learning disabilities may require more frequent in each CPS office to deal with these cases breaks during court hearings. ([.aw Commission op cit para 20); Several points are worth noting: • use of expert advice to identify ways oi' overcoming problems, to assess reliability • These measures will not always be and to assess the effect of the decision appropriate. Individuals needs vary. For whether or not to prosecute on the victim example Sanders et al (1997: 64) report that (Law Commission op cit para 20); some witnesses with learning difficulties were disappointed not to see wigs and • pursuing civil cases: parents whose disabled gowns having seen them on TV. Different children's cases have not been prosecuted witnesses can require different types and have been encouraged to launch a civil case amounts of support. where the burden of proof is lower (Cervi 1992: 15). • The content of these measures and/or the way they are used is also important. Sanders d. The trial et al argue that not all pre-trial preparation is effective. They argue that although Measures at trial appear to have two main aims. effective preparation can contribute towards The first is to reduce the fear and trauma of success for the victim, success is not simply a attending court. The second is to ensure that the conviction or removing all stress - both are quality of evidence suffers as little as possible setting the aims for pre-trial preparation too because of the witness's particular vulnerabilities. high. They define success as any case where In practice, the two aims can be related. the witness gave evidence considered by the jury. However they later suggest that the

143 mosr.successh.il preparation identified and. In addition, Sanders et al (1997: 53-54) suggest acred on the witnesses particular that the problem is not lack of support for such concerns/weaknesses which helped the measures - they found widespread support for this witness give better evidence or race giving pre-trial preparation for example. Instead they evidence in court more confidently. suggest the problem is confusion about which agency is responsible for arranging it, If this is • Ineffective preparation is not necessarily die correct, a more appropriate response may be to fault of rhe agencies involved. Sanders et al define responsibility for organising needs (1997, 51) identified one ca.se where a assessment and appropriate measures more clearly, person with a mild learning disability perhaps through legislation. There are several refused offers of assistance but found the agencies/organisations who might take on this trial difficult. Support cannot be forced on responsibility: people, even when it is felt to be in their best interests. * the police: at present, the National Standards for Witness Care suggest that the • Even the best preparation does not police should organise some of these guarantee that the witness will not find the arrangements (eg court visits). trial traumatic. Unforeseen events such as adjournments or a change to another court * the CPS: this has been recommended by the can arise: it is impossible to prepare for Law Society (submission op cit), although every possible eventuality (Sanders et al they acknowledge this may not be ideal. 1997:52). * Victim Support/court witness services: these just as many of the measures discussed in relation schemes also provide some assistance: for to orher groups of vulnerable witnesses may be example, all court witness services arrange applicable to people with disabilities/illnesses, so court familiarisation visits. However, as many of the issues raised are common, albeit with noted in chapter 2 they are largely restricted some distinct connotations. For example xo the Crown Court, admission of evidence was discussed in chapter 2 concerning witness intimidation. Pre-trial Sanders et al (1997:54) question whether lawyers hearings and, perhaps to a lesser extent, written have the skills to perform this task effectively, and depositions may also be helpful for people with this may apply to the police. I lowever, Victim disabilities/illnesses. Pre-trial hearings could for Support schemes or court witness services arc not example, be used for learning disabled witnesses yet available in all courts. They suggest that who are likely to find cross-examination m the assessment of needs of all vulnerable witnesses formality of the courtroom particularly stressful, should be assessed by experts. to help them give more effective evidence. Written depositions could be useful for people with As well as those measures which have also been physical disabilities or mental illness who may suggested for other groups of vulnerable witnesses, find it difficult m attend court. there are some distinct to witnesses with disabilities and illnesses. One example is the Many of these measures are currently matters at provision of interpreters: this was discussed earlier the judges' discretion. However as noted in in relation to reporting, fhree main areas are chapter 2, ir is not clear that these measures are considered below: routinely considered. Cervi (1992:15) supplies some anecdotal evidence that such measures rarely * cross-examination; used for people with learning disabilities. Reducing judicial discretion by creating a legal * expert evidence; and assumption in favour of such measLires might be one means of overcoming this apparent obstacle. * summing up. This would not however sit comfortably with the idea that measures should be based on an 1. Cross-exam ination assessment of individuals needs. Competency and credibility are particularly important issues for learning disabled and

144 mentally ill witnesses who may display an element Tn addition. Sanders et al (1997: 78) argue for of suggestibility. greater use of expert evidence (discussed further below) and advocate the creation or rules on when Judges can intervene to assist mentally or judges can intervene to prevent questioning which physically vulnerable witnesses jn two ways, first, is unfair or, because of the witness's vulnerability, by calling tor breaks so that witnesses am rest. likely to produce unreliable evidence. More and secondly to prevent inappropriate radically, they suggest a neutral examiner might be questioning. No research was found on the use or considered. The idea for this comes from Israel, breaks ro help mentally or physically vulnerable where neutral examiners are used for child witness adults, bur some was found on witnesses, and provisions based on this idea have intervention to prevent inappropriate questioning. been introduced in New Zealand and Ireland. The Sanders er al (1997: 78) found that many of the role of the "child interpreter'' in these two learning-disabled witnesses they interviewed felt jurisdictions is to sit next to the child and translate bullied or pressured, and some felt that their questions put to them into language that the child testimony had suffered as a result. However judges can understand. A similar idea was also suggested had rarely Intervened, and some found it difficult by the Pigot Committee (Ref to be added). TO adapt their own language. In some cases the judges themselves were perceived as bullying by 2. Expert evidence the witnesses. Sanders et al observe that this may be the unintentional result of failure to It has been suggested that expert evidence could understand the level of learning disability and its be used more commonly to help inform the court relevance. (including the jury) about the witness's particular learning disability (Sanders et al 1997: 77). This This .suggests that training for the judiciary and might equally apply to mentally ill witnesses, and magistracy on learning disabilities, as well as for to people with physical disabilities: for example the legal profession more generally, would be sight impairments may prove problematic if the useful. MEXCAP (1997: 9) recommends training legal profession give less credence to such for all those called to the bar as well as the testimony than is actually warranted. This may magistracy and judiciary. It also seems likely that however, be easier to tackle (for example through equivalent training about other disabilities and expert advice and training) than prejudice or mental illnesses could also be of benefit. I his misconceptions held by juries. could cover the nature and implications of disabiiities and illnesses. It could also look at how Sanders et al (1997:77) argue thai expert evidence to question people with intellectual disabilities may be useful to help deflect attacks on the (for example to slow the pace or questioning and witnesses credibility, ii it addresses specific issues use more appropriate language) and encourage raised by the case. However, they also see that the intervention by judges and magistrates to curb udmissibiliry of expert evidence under current inappropriate questioning. provisions ma)" need to be reviewed to enable more widespread use. Some efforts have already been made in this direction: for example the Bar Council has set up 3- Summing up a network of barristers who have some experience/knowledge of cases involving learning Finally, in cases which reach the Crown Court, disabled (Cervi 1992: 15). It in not known how before the jury retire to consider their verdict, the much use is being made of this, nor how effective judge usually sums up the case. A careful and it is. Tf it is successful it might be worth sensitive summing up is important. However, at considering inviting some members of the this stage the judge may issue a corroboration judiciary and magistracy in each Petty Sessional warning, There is no general requirement that a Division/Crown court area to specialise to some witness's testimony has to be corroborated by extent in cases involving mentally disabled/ill other evidence. Nevertheless, when a witness's witnesses. Such an approach would of course rely evidence is thought unreliable corroboration is on cases involving people with learning usually desirable, and judges may use their disabilities/illnesses being flagged in some way, discretion to warn the juries of this. and appropriate court scheduling.

145 No figures were found on the use of corroboration after the trial. The slow nature of the criminal warnings in cases involving witnesses with justice process means that victims are sometimes physical or mental disabilities. However Sanders denied counselling until long after the offence. et al f 1997: 60) report that only one or the cases Therapists are not usually trained to deal with the they studied involving learning disabled witnesses learning disabled, complicating matters further. came to court on the witness's evidence alone. They suggest that the police and CPS may It is difficult to identify measures x.o overcome currently filter such cases our. It seems plausible these problems. Some possibilities include: though that if responses to witnesses with learning disabilities improve, the numbers of cases • farther investigation of whether and to what involving uncorroborated evidence from witnesses extern therapy can contaminate witnesses, with learning disabilities will improve. Sanders ct and how this weighs against any al (1997: 60) argue that judges here, as in other improvements in performance at court; jurisdictions, should be prohibited from issuing corroboration warnings when this is based solely • reducing the length of time the criminal on the fact that the witness has a learning justice process takes: the recent Narey review disability. on delay in the criminal justice system made a number of recommendations with this e. ...and beyond aim. Some of these are to be introduced in the forthcoming Crime and Disorder Bill; After the trial there are two further areas for consideration. The first is that of therapy, and the • reviewing police and CPS perspectives and second of future crime prevention. policies on therapy: although the decision may affect the strength of the prosecution case, advice needs to be balanced: the decision about therapy should ultimately Although therapy may help witnesses perform in re.se with the witness And his/her family; and court, at present it may be delayed until after the trial because of fears that their testimony will be • training for therapists in working with undermined by the charge that the witness has people with learning disabilities: they could been "contaminated" or ''coached''. Counselling may have very limited value if the incident is not then be listed in a national register available discussed, but discussion might distort the to all interested agencies, with some system witness's memory of the event, lividence from a of referral to learning disabled people and witness who has received therapy is admissible but their carers. only if they have not been conr3min.ar.ed (Sanders et al 1997: 52). The problem then becomes Prevention proving whether or not the witness has been contaminated. To assist this, counsellors may be In the longer term, improvements in the criminal required to disclose notes and other records which justice system's treatment of mentally ill or would normally be treated confidentially and disabled witnesses may reduce their vulnerability would not have been written with this purpose in to crime by making them less of an easy target. mind: comments taken out of context might then However, some aspects of their increased be used to undermine the witness's credibility7. vulnerability (such as dependence on carers) cannot be addressed by the criminal justice system except through crime prevention efforts. Proving chat contamination did not occur can be difficult. According to Sanders et al (1997: 52) the • Statutory police checks on ail professional CPS and police usually discourage therapy until carers working with disabled or mentally ill people. The rules governing third-party disclosure have been changed under the Criminal Procedures and Investigations Act 19%, John Newing, then chief consrable of Derbyshire and are due to i?c implemented in January. One of'the intentions has called for statutory police checks on carers of of the changes is to wake it more difficult for defence barristers ui use third-part)' disclosure for "fishing expeditions " in which vulnerable adults (Cervi 1992: 1 5), but there possible lines of defence are sought, rather than support for seems no reason why such a measure should not existing lines of defence. also apply to those working with vulnerable

146 children. The main problem wirb this proposal is Williams' (1995a: 4) conclusion is worth quoting that it might give a hiise sense of security: given almost in full (see box 4). that a targe proportion or crime against people with disabilities/ illnesses is probably unreporred, Box 4: Williams' conclusion (1995a:4) the police may only know of a small number of offenders against this group. Against this it can be "redressing rhe stereotyped view of people argued that any protection possible would be with learning difficulties, in relation to crime, desirable. In addition, if this was implemented as is the key element in changing the present parr of a general package of measures improving situation. Justice is frustrated noi only confidence in the criminal jusrice system, rhe because of the response of the separate number of offenders known to the police (and agencies, but of the effect they have on each hence the effectiveness of this particular measure) other. The police do not record crimes • miaht increase. because they believe the CPS will not prosecute, staff do not report to the police because they 'do nothing' and victims do not Conclusion tell staff because 'they say the police wont help'. Consequently the courts are The literature review found a paucity of literature unpractised at dealing with vulnerable on physical disabilities and mental illnesses- as witnesses, and perpetrators see people with well as on the impact of multiple health problems. learning disabilities as safe targets. Positive This may partly be a reflection of .societal action...could break this spiral." attitudes generally, of the Invisibility of some of these constituencies (for example residential care, Numerous measures are possible: many of these separate education facilities and discrimination in could also be applied to other groups of Industry may all serve to hide these groups) and vulnerable witnesses. Table 2 summarises the their powerlessness (associated with dependence measures discussed. As with witness intimidation, on others). Whatever the reason, this needs to be some cases will always fall through the net. The addressed: by not adequately meeting the needs of task currently facing the criminal justice system, is these groups in the criminal justice system, we to ensure that there is a net which will catch as mav be increasing their vulnerability to crime. many of those witnesses who are physically or mentally vulnerable as possible.

150 Section 4: Special Offences

The Problem not examined further here, although the Review may consider it when finalising their definition of Definition of "special offences" vulnerable witnesses.

Some of the definitions discussed in section 1 Repeat victimisation suggest there are certain special offences where witnesses may be vulnerable. The only offences Another question is whether repeat victimisation specified by any of these ;ire sexual offences!. This It may aiso be argued that some particular Ls supported by a large literature that suggests that offences such as stalking should be covered. 1 his sexual offences are particularly upsetting for the raises the question of whether repeal victimisation victim. Nevertheless, other possible "special offences' can be identified, including domestic generally should be covered. Repeat victimisation violence, racially motivated crime and hale crimes is known to be common in domestic violence and against sexual minorities: all or these are considered racial attacks, but can also occur for victims who below. However, the literature review found more would not otherwise be considered vulnerable, information on some of these subjects than such as burglary victims (for a discussion of the others. In particular, very little Information was distinction between repeat victimisation and found on hate crime against sexual minorities: che intimidation see section 2). The finding that discussion below reflects this. crime is concentrated on a small proportion of the population (see for example Farrell and Pease 1993: 5-15) suggests that repeat victims' Other offences experiences of the criminal justice system will be more frequent and different in nature to other The selection used has some limitations. It could victims of crime. 1 lowever research in this area is be argued ihat witnesses of other violent offences relatively new, and has focused on measuring should be covered, but it seems inappropriate 10 repeat victimisation and identifying ways to label all witnesses or violent crime as vulnerable. prevent revictimisation. rather than on how this For example, much violent crime involves pub affects the victim's emotional vulnerability. Repeat brawls or orher assaults committed in public be ween young males, not1 all of whom will victims are not considered separately below, necessarily view themselves as. victims of crime. although many of rhe measures discussed may be 1 lowever, it could be argued thai" witnesses of relevant for them. some specific offences (such as the family of a murder victim) could be seen as vulnerable too. Source of vulnerability

The suggestion that repeat victims should be Against this is can be argued that their considered raises another isstie: by drawing vulnerability stems from bereavement, which may attention to particular offences the real source of affect witnesses of other crimes, suggesting that the witness's vulnerability may be obscured. bereavement should be another criterion of Vulnerability may not stem simply from the vulnerability. Due to time constraints this issue is qualities of the offence itself. Just as the repeated nature of victimisation may be important, so too may the personal characteristics (such as sex, race 1. It can be argued tome victims of sexual offences, such a< thoie who are raped abroad, are especially vulnerable (fill Sewnrd or sexual orientation) of the witness and (submission to the Review, 29/8/97). Tin- is not considered here perpetrator(s) and the relationship between them. became no wievaut literature was found, a/though the Revieiv "Hie significance or these factors may in turn may wish to consider that some witnesses are vulnerable became derive at least partly from public attitudes, both the offence was committed in a foreign county. 2. See the Victim Support report by Brown, Chmtie, C" Morm outside and reflected within the criminal justice (1990) "hamilies of Murder Victims Project "for a discussion of system. Mote detailed examples are given below. the needs of this group

151 Example 1: Sex Example 3: Sexual orientation

It can be argued that power relations between Vulnerability may be associated with sexual the sexes causes vulnerability for victims of orientation: both because public attitudes to sexual offences and domestic violence. According sexual minorities may increase their vulnerability to this argument; their vulnerability cases stems to crime (for example, "queer bashing") and not from the sexual or violent nature of the because of possible discrimination in the offence. Rather, it derives from the fact thai criminal justice system (both in the law and on most of these cases involve female victims and the part of some criminal justice professionals). male defendants, in a society which has The subject has been largely overlooked. This traditionally privileged the male over the female, may be because studying discrimination against and reflected this in its institutions. sexual minorities can be difficult (information about sexual orientation may nor be offered as Of course some such offences may involve male freely as information about racial origin) but victims who may also have unusually negative may also reflect lack of awareness. experiences and perceptions of the criminal justice .system. One explanation for the problems chat these witnesses experience is that the challenge the victims pose to sexist stereotypes Each of the four special offences identified above (such as those characterising males as strong, (sexual offences, domestic violence, racially powerful and assertive) and homophobic motivated crime and hate crimes against sexual attitudes. minorities) are discussed below in turn. Three final points should be noted about this selection: Thus, instead of labelling witnesses of special offences as vulnerable, other characteristics such as • The concern here is primarily about victims: sex, race and sexual orientation could be used as lew non-victim witnesses will be vulnerable criteria of vulnerability. One difficulty with this because of the nature of the offence (as approach is that it would label a very large opposed to other factors such as intellectual proportion ot witnesses as vulnerable. In addition, disabilities). there is itttle research examining whether these factors make witnesses vulnerable across the full • The categories of special offences discussed range of crimes. below are not mutually exclusive. For example, sexual offences may be racially Example 2: Race motivated, or may be inter-linked with domestic violence. It can be argued mat vulnerability in racially motivated crimes is just one of a number of • The Review may decide thar a different vulnerabilities associated with race, in a society selection would be more appropriate (or which has traditionally privileged whites. indeed to employ different criteria of Consequently, it may be more useful to examine vulnerability than the nature of the offence). how witnesses from ethnic minorities fare in general, rather than just those who fail victim to For each "special offence', definitions are racially motivated crime. Cultural and language examined first, followed by evidence (where barriers can be seen as examples of problems that applicable) from '"Criminal Statistics", the "'British may apply equally in offences which are not (Crime Survey", and other studies on the scale and racially motivated. However, such problems may nature of the problem. The one exception to this not apply to all people from ethnic minorities. pattern is hate crime against sexual minorities, where research is much more limited.

1. Sexual offences Definitions

The range of sexual offences is wide, including both keeping a brothel and rape. The most

152 common factor defining sexual behaviour as • the widening of the law to include male rape criminal is lack of consent. Where the victim is (under the Criminal Justice and Public voung, this translates into whether the victim was Order Act 1994); and above or below the age of consent. However, legal definitions include some behaviour between • legal recognition thai boys under fourteen consenting male (bur not sexual behaviour can commit rape (under the 1993 Sexual between consenting female) partners. I his may be Offences Act). addressed following a ruling European Commission on Human Rights, which rejected Another explanation for this increase is that the idea that the homosexual age of consent more victims are reporting rape. It is widely should differ from that for heterosexuals. A free rh.ou.ght chat public attitudes to rape and the way- vote is now planned on the homosexual age of rape victims have been treated by the criminal consent in the House of Commons. However, this justice system in the past, discouraged many will still leave anomalies to the general rule that from reporting. It is also generally accepted that lack of consent defines sexual behaviour as changes in public attitudes and police treatment criminal (eg. soliciting). of rape complainants may have contributed to the increase in reported rapes. However, it i.s also Table 3 lists the main offences and the maximum possible that the actual number of rapes penalty for each: it should be noted that under the committed might have increased (Lees 1996: Crime (Sentences) Act 1997 a life sentence is now 24). required after a second serious offence such as tape and attempted rape (unless there are Te conviction rate for recorded rapes fell from exceptional circumstances). 24% in 1985 to 9% in 19% (see Figure 3). Given that many rapes are likely to go unrecorded Scale and nature of the problem (either because they are not reported, "no-crimed"' or recorded as a lessor offence), and some i- Criminal Statistics convictions will be quashed on appeal, the real conviction rate must be even lower. There is some Sexual offences account tor a very small evidence that conviction rates for rape are also low proportion (less than 1%) of all alleged offences elsewhere in Europe and in America (Lees 1996: recorded by the police in England and Wales. xii). Research for the Home Office into the Most sexual offences appear to be committed by reasons for the increasing attrition rate for rape is men. In 1996 over eleven thousand men were due to be completed in June I 998. However, prosecuted for such crimes and about two-fifths initial findings suggest that this might be related found guilty: about a hundred women were to a larger proportion of rapes involving intimates prosecuted for such offences, and less than a third being reported, where the likelihood of a convicted (Criminal Statistics England and Wales conviction may be lower (Harris, 1997: 3). 1996 Supplementary Vbls 1 & 2, Tables S1.1 & S2.1). for those convicred of sexual offences, on the face of it sentences seem to have become more severe The number of sexual offences recorded has risen over the past decade. 1 he proportion of sex at a similar rare to recorded crime as a whole (about offenders given immediate custody rose from 35% 3% a year since ] 986). Nevertheless, in recent in 1986 ro 55% In 1996. This apparent increase years, the total number of rapes recorded by the in severity was encouraged by new sentencing police has increased almost threefold (Criminal guidelines issued in 1986. Changes were also Statistics England and Wales 1996 Table 2.16). introduced so thur only senior members of the There are several possible reasons for this increase, judiciaiy could hear rape cases. Despite this, the including various changes in the law over this proportion of convicted rapists sentenced ro period (Harris, 1997: 1). Examples include: immediate custody fell from 95% to 89% (see figures 4 and 5). Some sentencers do not always • legal recognition of marital rape (in 1991 in the case of R v R, the House of Lords upheld a Court of Appeal ruling on this); The number of convictions and cautions for sexual offences as a whole has tended to remain steady, at less than 2% of all convictims/cautions.

153 154 155 adhere to the guidelines (Robertshaw 1994: 343- It has yet to be seen how frequently judges will 345). The new measures requiring a life sentence use cheir discretion in "exceptional drcumssances' after A second serious offence such as rape and to impose sentences other than life. attempted rape may increase sentencing severity.

156 157 i. British Crime Survey explanation is that whereas initially they only mentioned those incidents they viewed as crimes, The BCS provides an indication of rhe amount of further questions revealed some sexual incidents crime not reported ro the police. However, in which respondents did not perceive as crimes. For practice survey method:* have nor proven a strong example. 6% of women reported having been measure or sexual offences. There are .several forced into sex at least once from age 16, but only problems: 2% described the most serious sexual incident they had experienced as rape. This complicates • people may find answering questions about matters. If only rhose incidents viewed as crimes sexual offences embarrassing; by the victims are taken, the second measure over • some respondents may be reluctant: to counts, but the first measure may still under- disclose sexual victimisation because of lack count. Despite these limitations, this method does of privacy (concern that someone might seem more reliable than previous approaches- overhear, or in some cases fear of retribution Other findings were that (Percy & May hew 1997: if the perpetrator should learn of die 10-15, cables^, 8, 9 & 10): disclosure); • Thircy-nine per cent or those reporting • public definitions, may not be the same as some sexual incident viewed them as crimes. legal definitions: for example, legally rape does nor include penetration by bottles, • The proportion of Incidents regarded as fingers, etc.; and crimes was highest for rape (74%).

• a respondent may come to label an incident • Incidents were more likely to be considered differently over time. crimes when the offender was a stranger (49% compared to 23% when the victim It is difficult to judge what effect these problems knew the offender by name). have overall, but the "embarrassment factor" and concerns about privacy suggest a bias towards • Of all those incidents viewed as crimes, 39% under counting. Although the BOS collects data were reported to the police. Only 5% of on sexual offences, this has nor been published those incidents viewed differently (for recently because of doubts about its reliability. To example "wrong but not a crime"', "just try to improve the 1994 survey (Percy & Mayhevv something that happens", or "not sure") 1997: 6-7) respondents were asked to read and were reported. answer questions on a computer, to try to reduce the embarrassment and privacy problems. • There was strong evidence that rjpe victims were more reluctant to report to the police More offences were reported with the self-keying than victims of other sexual offences. Only method. From the conventional sexual 26% of rapes viewed as crime by the victim victimisation question asked towards the were reported to the police, against 47% of beginning or the interview, fewer than 1 % of attempted rapes and 45% of indecent women said they had been rhe victim of one or assaults. more sexual incidents in the last year. However, with the self-keying method about 8% admitted iii. Oiher studies. sexual victimisation. Similarly, fewer than 1% initially said they had experienced one or more There have been a number of other surveys of sexual offences since age 16. After follow-up victims concerning sexual offences. One of the questions in the self-keyed component, up to criticisms of the BCS is that national figures mask 22% reported sexual incidents of some sort from local differences in risk. One well-known example age 16. of a local survey is the Islington Crime Survey (ICS), in which over one and a half thousand These findings are difficult to interpret. It is questionnaires were completed (a response rate of plausible that after further prompting the women over 75%). The survey found 4% of respondents remembered more crimes, suggesting that the reported some level of sexual assault in the second measure is more reliable. Another possible previous year, although three-quarters described it

158 as low-level pestering rather than assaulr. Women question was problematic (van Dijk and Mayhew were more likely than men to describe this 1992, Travis er al 1995 and Koss 1996 cited'by experience as assault. About 8% of women Percy and Mayhew 1997: 23). reported sexual harassment or assault during the previous year. Eight percent of women and 12% Outstanding questions of men reported sexual abuse before aged 16 (Crawford et al 1 990: 8, 19-20). The number of sexual offences reported is affected by the research methodology used, and caution However, comparing crime survey findings is has to be exercised in interpreting the results. problematic: methodological differences can lead However, rhe literature found on sexual offences to great variation in results. For example, the BCS focuses on offences involving female vicrims and suggested offenders were known to 60% of male offenders. victims 60%, against 23% in the ICS. One explanation for the difference might be that 2. Domestic violence Islington is a high crime area whereas the BCS is a Definitions national survey, but there were also major methodological differences between the two. For The Home Affairs Select Committee in 1993 example, in the ICS male respondents were also defined domestic violence us "any form of asked about women in the household: the BCS physical, sexual or emotional abuse that takes excludes male respondents from questions on place in the context or. a close relationship". This sexual victimisation, following tests .suggesting definition was adopted by the Home Office in an men were unwilling to answer questions on sexual inter-agency circular in 1 995. This said that victimisation or take them seriously (Percy and domestic violence can take a number of forms, Mayhew (1997: 21, 5). not only physical violence but also sexual abuse, rape, and mental and verbal abuse such as threats Another more recent example is the 1996 and systematic criticism (cited in Home Office International Crime Victimisation Survey, which 1996). The Home Affairs Committee definition is involved mainly telephone interviews of nationally representative .samples of between one also used by the CPS policy statement on and two thousand people in each of eleven domestic violence which observes that in most countries. The 1 996 survey found some similarity cases, the relationship will be between partners in risks of sexual assault between the eleven (married, cohabiting, or otherwise) or ex- countries. On average, 2.5% of respondents partners' . It further elaborates that although in reported one or more sexual incidents over the last most cases the offender is male and the victim year. I he highest rates was recorded for female, domestic violence can also involve male Switzerland (4.6%) and Austria (3-8%), with the victims and female offenders, and partners/ex- lowest in France (0.9%). England and Wales fell par mers or the same sex (CPS 1995: 2.1-2.2). in the bottom half of the range at 2%: Scotland and Northern Ireland were lower at 1.3% and This definition is very wide: most studies 1.2% respectively (Mayhew & van Dijk, 1997: encountered in the literature review focused more Table 1). narrowly on physical violence (usually by male offenders against female victims). This appears to be the approach used by the police. However, the Again, these apparent differences should be CPS definition is useful in highlighting that other treated cautiously. The study found evidence or incidents can occur in close relationships, and that consistency between the countries in how these relationships affect the nature of rhe seriously they viewed different behaviours, from incidents. which the authors inferred a large degree of consistency between countries in defining certain Another definition is provided by the British actions as crimes. However, willingness to report Crime Survey, which focuses on violent incidents sexual offences to researchers may have differed: "involving partners, ex-partners, household although this is partly related to perceptions of members and other relatives, irrespective of seriousness, other factors (such as the location". The inclusion of non-partners means "embarrassment" factor) may have affected the definition is quite broad, but has the response differently in different countries. It has advantage of matching police measures of also been suggested that the wording of the initial

159 domestic violence (Mirrlees Black, Mayhew & On the nature of domestic violence, it was found: Percy 1996: 2). However, almost all cases reported to the 1992 survey took place in or jusr • Approaching half (44%) of ail violence outside the victim's home: some occurred at the reported against women was domestic: men home of a friend or rhe offender when the victim more frequently reported being victims of and offender were not currently cohabiting stranger and acquaintance violence. (Mayhew, Aye Maung & Mirrlees Black 1993: 82). • Younger people (those aged between Hi and 29) were at greatest risk. Scale and nature of the problem • Weapons were less likely to be used than in i. Criminal Statistics other violent crimes, but victims of domestic violence were most likely to be injured. This In law, there is no such offence as domestic suggests they only reported the more serious violence. Instead domestic violence is prosecuted incidents. under a number of different offences, such as assaults and breach of the peace (a public order • A third of victims were victimised more than offence). Nor do Official statistics identify the once, relationship between the victim and offender. Consequently, it is impossible to examine • Women found domestic violence more domestic violence using Criminal Statistics, upsetting than men: common reactions because it is not separable from, for example, included anger, fear, crying and difficulty stranger assaults. Some forces keep domestic with sleeping. violence registers, in which such offences are supposed to be recorded, but there are concerns These findings should be treated with caution: the about recording practices (particularly the rate of BCS probably under counts violent crime, "no-criming") and statistics from these registers especially domestic violence. It seems plausible are not published. This makes criminal justice that violence between people who know each responses to domestic violence difficult 10 other is less likely to be perceived as crime. There monitor. Both the National Inter-Agency may also be other reasons for not mentioning Working Party on Domestic Violence set up by domestic violence: as with sexual offences, Victim Support in 1990 (Victim Support 1992: 2.8) and Grace (1995: 53-54) recommended embarrassment and concerns about privacy and improved recording of domestic violence to assist retribution may be important. monitoring. iii. Other studies

ii. British Crime Survey One of the earliest studies on domestic violence (Dobash & Dobash 1979: 164) found that a The BCS provides much more information about quarter of all violence is domestic, and only 2% domestic violence. The 1992 survey found that was reported to the police. Other surveys have 10% of women experienced domestic violence at also suggested that between 10% and 25% of some point in their lives (Mirrlees Black 1994 women have been the victim of violence by a male cited by Grace 1995). The latest survey (Mirrlees partner at some time in their lives. This seems to Black, Mayhew & Percy 1996: 5. 28-35) found tic with the findings of some forthcoming research domestic violence increased 242% between 1981 for the Home Office (Phillip, & Brown 1998: 39) and 1995. Other evidence from the survey on entry into the criminal justice system: suggests that more domestic violence is being domestic incidents accounted for just over a reported to the police, so one reason tor this rise quarter of all violence against the person arrests'. may be that respondents are also more willing ro According to Dobash et al (1996: 2) similar disclose domestic violence ro interviewers. figures have been found in other countries. Domestic incidents accounted for about a quarter of all violenr offences reported: the authors suggest that '"if domestic violence could be Six per cent of all suspects were arrested for domestic violence measured better, the proportion might be higher". related offences (including public order offences and property offences, such as criminal damage, not just violence against the person).

160 The literature ieviewed has tended to tocus on currently two years imprisonment on male on female violence, but it seems plausible indictment (that is at rbe Crown Court), a rhar some domestic violence may be female on fine up to the statutory maximum or both. male or occur in a same-sex relationship. Despite Tor cases tried summarily (in the magistrates this, there have been a few studies of female on courts) the maximum penalty is six months male violence recently. These suggest this is far imprisonment, a i)ne or both. less common than male on female violence, For example, Phillips and Brown (1998: 91) found • possessing racially inflammatory materials: that 10% of those charged with domestic violence this was also created under the Public Order offences were female. Other research in this area Act 1986, and carries the same maximum suggests that violence committed by women penalties as stirring up racial hatred. against men is usually against a background of a history of violence by the male: that it tends to be The Association of"Chief Police Officers (ACPO) less systematic, and less serious (Dobash et al definition of a racial incident is any incident 1992; 71-91: Dobash er A 1996). However, no where: studies were found examining domestic violence within same sex relationships. a. the reporting or investigating officer perceives some racial motivation, or Outstanding questions b. any other person alleges racial intimidation At least two outstanding questions can be (Home Office 1997b: 3D identified: Scale and nature of the problem • to what extent sexual abuse accompanies physical violence; and The literature review found little published research on hate crimes against ethnic minorities. • the extent of domestic violence in same-sex relationships. i- Criminal statistics

3. Racially motivated crime Criminal statistics on racial incidents are limited. Where racial motivation is not reflected by the Definitions offence type, it is not possible to separate it from non-racially motivated offences. Like domestic A wide range of legislation may be used to violence, police forces do record racially motivated prosecute racially motivated crime. For example, incidents, but this has only occurred since 1988. racial harassment and racial verbal abuse can lie prosecuted under the Criminal justice and Public Order Act 1994'and Public Order Act 1986. The number of racial incidents recorded by the However, at present only a few offences with an police has risen each year, from about 4,400 explicit racial element exist. For example: recorded cases in 1988 to 13.150 for the year ending March 1997 (see Figure 6). This increase is greater than that for recorded crime generally, but • racial discrimination: the Race Relations Act it is not clear whether or to what extent this 1968 First made it unlawful ro discriminate reflects an increase in the real rate of racially on the grounds of ethnic origins in the motivated crime. It is possible that the increase in provision of goods, facilities or services. recorded racial incidents is partly due to changes in recording: until the early 1980's officers were • stirring up racial hatred: five offences of this not required to consider other people's views on type were created under the Public Order whether an incident was racially motivated. Act 1986, s27(3), including distributing Against this, Sibbit (1997: 25) found that in written material. The maximum penalty is practice most racial incidents recorded by the police were cases where the victim had claimed racial motivation. For example, this created a new offense of intentionally causing harassment, alarm or distress to deal with racial harassment, although it can be used for other forms of harassment. Another factor could have been increases in reporting rates. The BCS provides some evidence

161 that reporting rates increased between 1987 and included in the ("rime, and Disorder Bill should 1991, but the eihnic minority samples were roo assist assessment of the scale of the problem. The smail tor the increase to be statistically .significant: proposals include new offences of racial violence it may have been due to sampling error. A more (including racial common assault, racial assault important factor may have been greater occasioning actual bodily harm, and malicious willingness by the police to record offences as wounding). This would enable racially motivated racially motivated when racial motivation is violence to be treated more serious!)' than other alleged even in che absences of other evidence of violence. Both the Commission for Racial racial motivation (Aye Maung & Mirrlces Black Equality and the Home Affairs Select Committee 1994: 20). have supported such a measure.

Statistics on convictions and sentences are not A new offence of racial harassment has also been readily available. However, there is sonic further proposed. As mentioned above, although there are information from the CPS racial incident existing provisions dealing with harassment, racial monitoring scheme which began in 1995 (CPS harassment has nor been dealt with separately. 1997: 3-7). For the year ending March 1997 over There are also concerns about the adequacy of the 1,300 completed cases were recorded. Although provisions for dealing with low-level harassment. the ACPO definition of racial incidents was used, At the same time, the new offence would enable more were identified as such by the CPS than the the courts to treat racial harassment more police (63% identified by CPS against 37% by seriously than other harassment. The effectiveness the police)''. This adds weight to the argument or these measures may depend on how easily racial that despite the growth in the police figures, they motivation can be proved. In some cases, racial still under count the number of crimes which are motivation may be obvious: for example, when racially motivated. Concerns about the police racist slogans are painted on the property of figures have been expressed by HMIC, who have ethnic minorities. In other cases, proving racial suggested some police officers may not be clear motivation may be more difficult. about the definition of a racial incident (1997: 2.66). ii- British Crime Survey

The CPS (1997; 7-9) records also show the The BCS is broader in approach than the police charges put by police and some details of the figures, in that it looks at unreported as well as outcomes of those cases which are prosecuted: reported crimes. Against this, the BCS only looks at those crimes where the victim perceived racial • The most common were public order motivation: it does not cover any where the police offences (48%) followed by assaults (27%) detect racial motivation bur none is reported by and criminal damage (14%). the victim.

• In two-thirds of cases the charges were A key findings from the 1996 and previous unaltered by the CPS. Less than 2% were surveys (Percy 1998: 5-6) was that higher rates of increased: the remainder were either reduced victimisation were found among ethnic (11%) or dropped (21%). minorities. This is partly explained by demographic factors. [;or example, ethnic • Of those prosecuted, most (79%) pleaded minorities tend to be younger, of lower socio- guilty, although some (21%) were initially economic status and to live in rented or public contested. housing, all factors associated with greater risk of crime. Smith (1994:1106) suggests high Details of conviction rates and sentences were not victimisation rates among Afro-Caribbeans may supplied. However, Government plans (Home also be related to their higher recorded offending Office 1997c: 7-9) for new measures to be rates: a large proportion of offences against Afro- Caribbean are committed by Afro-Caribbeans. This fails to explain though why Asians (who have relatively low recorded offending rates given their numbers in the general population) have the Although there were variaiiom between areas, only two forces identified more, racial incidents than CPS and even then the highest risk of victimisation. margin was negligible (Anglia 53% and Midlands 52%).

162 However, most crimes against minorities were nor However, the perception of racial motivation perceived by the victim as racially motivated. alone may increase the impact that crimes have on About 4% of Airo-Caribbeans, 5% of Indians and the victim: victims who perceive racial motivation 8% of Pakistanis reported racially motivated even when it did not exist may still be vulnerable. offences, representing about 15% of all crimes committed against them. The number of racially Three-fold categorisation of racially motivated crimes reported was similar ro that motivated crime (Fitzgerald and Hale, reported in previous yean.. Racial motivation was 1996:57) perceived in more personal crimes than property crimes. A larger proportion of those* where racial 1. Offences which are solely motivated by motivation was perceived involved white racism. offenders, more of these crimes involved groups, and the perpetrator was less likely to be known to 2. Offences which have a racial element, either victim. Racial incidents were more likely to be in motivating the offence or arising during part of a series, suggesting repeat victimisation is a the incident, but which are not entirely particular feature of racially motivated crime racially motivated. (Percy 1998: 15-20). 3. Offences which are racially motivated Measuring racially motivated crime is not a simple (wholly or partly) bur where this motivation task though. The box below suggests three is nor perceived. categories of racially motivated crime. Victim surveys depend on rhe victims awareness of racial iii. Other studies motivation, but racial motivation may not always be perceived when it is present. On the other There are a few other studies on racially motivated hand it is possible that racial motivation can be crime. For example, Maynard and Read (1997: 2- perceived when it is not actually the main 4) conducted a postal survey of all police forces motivation for the offence. Victim surveys may asking for figures on recorded racially motivated therefore under-count racially motivated crime, incidents for 1996/7. They received returns from but it is not known how big the problem is, thirty four of the forty-two forces: the remaining

163 eight were able to .supply data other data (for Other research conducted by Love and Kirby example, five gave derails for she calender year (1994: 1-12) looked at racial incidents in council 1996). In total over thirteen thousand racially housing. Self-completion questionnaires were sent motivated incidents were recorded. The highest to all local housing authorities, and a high numbers were found in the Metropolitan Police proportion (82%) responded. However about a (rise ''Met'), at over five and a half thousand third of those responding said that they had very offences'. few or no ethnic minority tenants, and three-fifths said that racial violence and harassment was not a One-explanation tor the large proportion of racial problem. The other respondents were asked to incidents recorded for the Met might be the provide more detailed information about racial concentration of ethnic minorities in the area, incidents. The responses show an increase in the rather than a high rate of (recorded) racial average number of racial incidents recorded by victimisation. Maynard and Read (1997: 5! housing authorities from 35 in 1987/8 to 56 looked at the number of recorded racial incidents offences recorded in 1989/90. There was great per thousand of the ethnic minority population. variation between authorities, with some According to their figures, die race in the Met is reporting much higher numbers each year: nine under four recorded offences per thousand, authorities reported more than a hundred against an average figure of just under seven. The incidents in 1989/90. highest rate was in Northumbria, at almost twenty-three offences per thousand. However, this However, most respondents thought their figures could at least partly reflect differences in recording underestimated the problem. This is supported by practices between police forces. Sampson and Phillips (1992: 4-5) study of racial incidents on an East London estate. During a six- Maynard and Read (1997: 7) also looked at the month period over twice as many incidents were nature of the problem. They found verbal reported to the housing authority as to the police, harassment was most common (38%), followed but more than twice as many again were reported by assault (21%) and damage to property (20%). to the homeless families campaign/law centre. Figures varied between police forces: this could be Although this was a small study, it does suggest explained both by differences in the nature of that housing authority figures should also be crime in different areas, and by difference!) in treated with caution. In particular Sibbit (1997: reporting rates. However, the overall pattern is 63) observes the types of offences reported to the supported by Sibhit (1997; 27-28), who looked at police and housing departments differ: only those 140 racial incident reports in one area and occurring in or near council housing are likely to identified three main types: be reported to housing authorities.

• contact assaults, where direct contact is made Outstanding questions by the perpetrator intended to cause physical injury or pain, were die most rare; Clearly, our evidence of the scale and nature oi racially motivated crime is only partial. There are • indirect assaults, where contact is indirect, significant limitations to both the official figures eg. where objects are dirown at the victim, a and victim surveys. More information on gun is used or spitting, and reporting rates would be particularly useful in explaining why recorded racial incidents have • intimidatory behaviour, which included increased so dramatically. It is however difficult to racist verbal abuse, damage to property and survey enough people from ethnic minorities to threats, was the most common. draw firm conclusions on this. Other potentially useful information would be evidence of the scale There was some overlap between these categories: of the problem in other countries. Finally, and for example both contact and indirect assaults perhaps most importantly for the Review, little could be accompanied by intimidatory behaviour. research was found on how victims of racially motivated crime experience the criminal justice system. The literature that was found on this primarily concerns the police. These issues are Home Office figures for 1996/7 published since the research show about 5,600 offences in the Met and 13,150 in England explored in rhe following section on "the and Wales as a whole (1997: 31). response".

164 4. Hate crimes against sexual minorities such offences are still prosecuted indicates that there is still some room for improvement. In Very little literature was found on hate crimes addition, Mason and Palmer (1996: 3) argue rhar against sexual minorities. increased tolerance has led to increased visibility of sexual minorities, which may increase Scale and nature of the problem opportunities for hate crime against them. Hate crimes against sexual minorities are not Other sources examined separately from other offences in either Criminal Statistics or the BCS. Contrastingly in Few alternate sources of information were found: America there has been a legal requirement on the Department of Justice to collect and publish 1. Commission on Discrimination survey annual statistics "on crimes that manifest This three year survey (1977-1980) is now very prejudice based on race, religion, sexual old. By examining newspapers (including gay orientation, and ethnic origin" under the Hate papers), evidence of 250 attacks against people Crime Statistics Act since 3 990 (Conyers in I Icrek thought by their attackers to be gay were found in & Berrill [Eds] 1992: xiv). Great Britain and Northern Ireland. Of these 15% reportedly led to the death or disablement of Recording systems within the criminal justice the victim. However, '"it is clear that the cases system in England, and Wales have not making the news are only a small and biased traditionally collected such data separately. This is sample of what goes on" (fvleldrum. 1980: 1). beginning to change in some police forces. For example, the Met now separate such offences on 2. Lewisham survey their computerised crime recording system, which logged almost 250 such offences in the first 9 More recent research was carried out (in 1 992) as months. HMIC has called for all forces to follow part of the I.ewisham Safer Cities project to help this lead (1997: 3.19). address lack of research in this area. There were two main elements (Safe Neighbourhood Unit At the same time though HMIC have 1992: 1): acknowledged that perceptions of the pervasiveness of homophobia in the police are • a self completion survey covering gay men likely m discourage sexual minorities from only who lived or worked in the borough or reporting crime generally. Concerns about were regular visitors; and confidentiality are also important: examples have been reported of officers informing family, • interviews with gay men and key agencies employers or neighbours about the individual's (including the local police). sexuality. Similarly, there is some anecdotal evidence chat reporting may be met with Key findings from the study (1992: 35, 40-41) accusations of wasting police time or even were that: violence (Galloway 1983: 106-107). In Manchester anonymous self-report forms are • The vast majority (8 1 %) reported being used to help gauge the extent of experience of verbal abuse. homophobic crime (HMIC 1997; 3.24; 3.27), but it is not known how effective this approach is. • Approaching half (45%) repotted being physically attacked because of their sexual Despite these problems, official statistics on orientation, two-fifths of whom required prosecutions for sexual offences such as indecency medical attention as a result. between males can nevertheless provide an indication of attitudes towards sexual minorities • Multiple victimisation w;is common. within the criminal justice system. "Nationally, prosecutions for gay sex offences have declined • Property offences were less common: 13% year on year throughout the 1990s, and are reported being robbed, 17% damage to currently running ac Jess than half the level of 10 property. years ago'1 (The Guardian, 26/11/97). Nevertheless, it could be argued that the fact that

165 • Vulnerability appeared to be lower for those and bisexuals. They distributed fifty thousand who were in regular relationships, perhaps questionnaires through gay publications and gay because less time would be spent in public mailing lists and received over four thousand places frequented by gay men. completed forms. Although the sample was large and che authors suggest the response rate was high Harassers were most commonly reported as for this type of survey, there is a danger that the unknown groups or individuals, but in a third of sample may not have been representative. In cases verbal abuse came from people in close particular those who had been victimised might contact with the respondents such as neighbours, be over-represented. To try to minimise this risk It colleagues or relatives (1992: 42-43). was stated at the top of the questionnaire ''whether you have experienced violence or not, The report suggests there is a perception that we need YOU to fill in this questionnaire". sexual minorities are targeted because they are- thought less likely to report crime. Reporting rates Almost all respondents reported using some kind to the police (and to advice agencies - both gay of avoidance tactic, such as not kissing or holding and straight) were low. The highest reporting rate hands in public or telling people they were gay. to the police was 16% for violence, followed by Mason and Palmer (1996: 68-72) argtie that such 13% for verbal abuse. Attackers were interviewed strategies reduce the visibility or sexual minorities, in only seven cases, and five charged. iMost helping reinrorce the idea that they are a tiny victims (66%) were very dissatisfied with the minority or even that they don't exist in some police response: common complaints including areas. Yet it was found that there was little insufficient or no action and not treating correlation between these strategies and risks of complaints seriously. These seems unsurprising victimisation. Other key findings (Mason and given that 21% reported verbal abuse from police Palmer 1996: 1-2, 45) were that: officers and 3% being violently assaulted by them (1992:43-45). • Just over a third of men and about a quarter of women reported experiencing violence in Contact with the police was more likely to conic the last five years because of their sexuality; from police surveillance of "cottages" and "cruising areas" than reporting crime. About a • About a third of all respondents reported third of respondents reported being asked ro move being harassed (including threats, blackmail, on, and a quarter being charged. Most vandalism and hate mail); respondents thought a more sympathetic police response to attacks and prosecution of • Three quarters reporied verbal abuse on at perpetrators would have the greatest impact on least one occasion: more than a quarter the safety of gay men. Action by the council (for (29%) reported six or more occasions. example by using tenancy agreements) and improved liaison between the police and gay clubs The researchers also found that some groups had were also strongly supported (1992: 46 & 49). greater risk of victimisation than others: in Interviews with the police suggested they had particular black, asian and disabled respondents little knowledge of attacks on gay men and did were more likely to report violence than average. not see it as a priority. However, they also Problems were also greater for young people: reported difficulty for officers in identifying the almost half of those under I 8 reported violence, motive for an attack as anti-gay. There were approaching two-thirds reported harassment, and particular problems in asking a victim directly whether he was gay (1992: 52). nine out of ten reported verbal abuse within the past five years. Some respondents reported that it was easier to move house than to involve the 3. Stonewall survey police: it was commonly reported that most agencies were either indifferent to their problems Stonewall is a national pressure group which or sympathetic but took no action (Mason and campaigns for the civil rights of lesbians, gay men Palmer 1996:8,54,27).

The Stonewall report also cite a small but representative survey by Snade, Thomson and Chetwynd (1995), and another small study by Truman et al (1994) in Mariihestei: regrettably there was not time to include these in the literature review.

166 4. American research rhar chore may have been a bias towards those with experience of violence: those with negative There appears ro be more lircramrc on hare crimes experiences may have been more inclined ro against sexual minorities In America. For example, respond. -A survey or over 2,000 gay men and lesbians across eight American cities found that almost all had suffered some form of victimisation. More The Response specifically a filth reported suffering physical violence at least once, and almost half had been 1. Sexual offences threatened with physical violence. Many reported repeat victimisation: 92% of those who reporting a. Reporting verbal abuse, and 47% of those who said they had been physically assaulted (National Clay and Failure to report Lesbian Task Force cited by Berrill 1992: 19-20). Berrill summarises the findings of a further It was noted earlier that there is some evidence twenty-three studies. The median figures (Berrill char reporting of sexual offences has increased, 1992:'20) were: which has been attributed to improvements in police responses to sexual offences. Nevertheless, • Tour fifths reported verbally harassment; the reporting rate is still low compared to that for othet offences. There are various possible reasons • A third reported being chased or followed; tor failure to report sexual offences including (see for example, Williams 1984): • A quarter reported having objects thrown at them; • shock, in particular Post Traumatic Stress Disorder (PTSD) which covers a variety of • fust under a fifth reported vandalism; symptoms including repressing thoughts, feelings and memories about the offence (see • Seventeen percent reported physical assaults; Parkinson 1993; Peterson, Prout and d an Schwarz 1991 for further discussion of PSTD; Resick 1993 considers PSTD in rape • Thirteen percent reported being spar upon. victims); fhere is also some evidence suggesting hate crime • fear that the media would publicise their against sexual minorities is growing. case and blame them; Outstanding question s • fear that family, friends or colleagues would not be sympathetic, rhac they might be Further research would be useful on all aspects of disbelieved or even ostracized: hate crime against sexual minorities in this country, This needs to cover both male and female • a desire ro protect family and friends from victims. The Lewisham study did not look at the knowledge and possible media attention; women because it was commissioned by the Lewisham gay Alliance, which had little contact • intimidation or fear of reprisals; with lesbians in the area. It is also questionable how representative it was of homosexual men in • concern about the response from the die area. The self-report approach generally has a criminal justice system: for example, that poor response rate (28% in this case as a complaints will not be believed, treated proportion of questionnaires sent out). The sympathetically or as seriously as the victim manner in which the questionnaire was circulated feels is justified, and lack of confidence n the (through gay bars, clubs and networks) may also ability of the system to convict the have meant that the survey did not reach all perpetrator. sections of the local gay population. This is a very difficult area to research though, and the Williams (1984: 461-465) looked at 246 rape shortcomings of the Lewisham study were cases in Seattle and found reporting is more likely recognised in the report. In particular, it suggested

167 if the rape corresponded ro the "'classic rape bncou raging reporting situation", that is If: Changes in the treatment of vulnerable witnesses • the victim was raped in public, abducted throughout the criminal justice system may from a public place or raped by an assailant encourage reporting, but this starts with who entered her home by force or without improving experiences at the first point of her consent; contact. It is especially important that when an offence is reported the response is sympathetic • the assailant was a stranger to the victim: and supportive. Some measures have already been taken to improve the police response to sexual • the victim was threatened with, or subjected offences, and to encourage reporting. For example to a high degree of force; Lees observes that training for officers dealing with rape complainants is now more common • the victim was seriously injured. "although chis is often pretty minimal". Some forces also have a chaperone system where one The relationship between the victim and offender female officer is assigned to the complainant was the most important factor. throughout the investigation (1996; 23).

Anecdotal evidence also suggests that the cases As discussed in relation to witnesses with most likely to result in a conviction and be treated disabilities and illnesses in section 3, developing most sympathetically by the criminal justice system inter-agency cooperation may be another way are those which most closely fit this "classic rape forward. Other agencies such as Rape Crisis may situation". Concerns about the criminal justice be the victim's first point of contact and influence- response to rape victims were highlighted in the decisions about reporting to the police. In I 970s and 1980s, when it was suggested that addition, they can provide support which may- criminal justice responses tend to be based on a assist complainants through the criminal justice number of rape myths which inform the concept of process. I Iowever, llape Crisis schemes coverage is the "classic rape .situation" (see box). patchy, limited ro a few hours a day in some areas: it has been argued that there is a need for more Rape myths (London Rape Crisis 1984: funds to provide nationwide 24 hour support 1-7) (Lees 1996: 5). Other approaches include a pilot scheme recently announced in Merseyside The London Rape Crisis Centre identifies a enabling rape victims visiting hospital to report to number of rape myths: widely held but the police at the same time. Repotting can be misconceived ideas about rape, including: anonymous if the complainant ptefers: a unique reference number can be used on records instead • women enjoy rape; oi their name until they feel ready to proceed with the case. They need not even sec a police officer: • rape is committed by mad strangers (most forms are supplied and faxed to the police station rapists are known to the victim and few are (Jenkins, 1997: 10). found to be mentally ill); Investigation • women provoke rape (for example through b. the way they dress); Attrition (the number of offences which are lost, either because they are not reported or because • men cannot control their sexual utges they are dropped at some stage between reporting (most rapes appear to be wholly or partly and conviction) is a particular concern in sexual planned in advance); offences. Attrition before reporting Is discussed above. However, less attention has been given to • false and malicious allegations are common attrition after reporting. Grace, Lloyd and Smith (the evidence suggests the rate of false (1992: 7, 25-27) found that of about 300 alleged allegations is the same as for other crimes); rapes police recorded in 1985, a quarter were no- crimed and only half the original sample were • only certain types of women get raped prosecuted or cautioned. Thirty-five percent (women from all age groups, classes and resulted in a conviction of some kind but onlv a races are raped).

168 quarter were for rape or anemprcd rape. The three • the police decide whether to tcfer a case to most important attrition points were where: the GPS;

• the police decided whether to "no-crime" an • the GPS decide whether to proceed or incident; reduce the charge: and

• the police decided whether to prosecute (the • the jury (or magistrates) decide whether to study was conducted before the introduction convict the defendant of rape. offheCPS);and The impact of the CPS on the attrition process, • the jury decided whether or not to convict both in making decisions and influencing police the defendant or rape. decisions, will be one of the areas covered in a new Home Office study on attrition in rape cases Those cases where there was some acquaintance due to report in June 1 998. However, initial between the complainant and defendant were findings from Harris's study suggest that "cases more likely to be dropped at each stage than those were more often no-crimed for other reasons'' involving .strangers. than evidence that the complaint was false (1997: 3). The finding that some offences were downgraded is supported by Lees and Gregory's examination of In some cases of course, the complainant may police records tor rape, attempted rape, buggery decide not to proceed with the case during the and indecent assault at two London police investigation. The manner and content of stations between 1988 and 1990. This found that interviews may be one contributory factor. rape and attempted rapes are sometimes Although police questioning is now believed less downgraded to indecent assault, but more brutal than in the past. Lees suggests interviews surprlsindv in .some cases the sexual nature ol the may still be more upsetting than necessary, and offence was removed (for example an indecent actually undermine the prosecution case. She assault classed as robbery). As Lees scares, ic may argues that ''police unwittingly assist'" attacks on be in the interests of the complainant to reduce complainant's reputation in court by "anticipating charges where there is not enough evidence for ihc the defences line of questioning in interviews '. higher charge. In some cases it may also be L'or example, questions on the complainant's difficult to decide where the dividing line between medical history (eg abortions) may be especially offences lie, for example in deciding whether damaging in court. The defence are given records intention to rape existed. Nevertheless the of police interviews, and then use this as frequency and nature of some downgrading may ammunmon in court. hus although the police be cause for concern. Ironically, one reason for T may complain that low conviction rates frustrate recording a rape as an indecent assault may be to make it easier to no-crime (Lees 1996: 99-101). their efforts to treat sexual offences more seriously, Lees suggests thai they sow the seeds for this poor success rate. In contrast such records ate Home Office circular 1986/69 advised complaints confidential in the US (Lees 1996: 102, 239). should only be no-crimed if proven false. If the complaint is withdrawn or eorroboration lacking, Identification parades may also be upsetting for it should srill be recorded as a crime. However, the victim of a sexual olfence corning face to face Lees and Gregory found over a third of cases were with the offender. Various measures can be used no-crimed: reasons included the complainant to minimise this, including the use of screens, deciding not to continue and police perceprion of mirrored glass (submission from North Staffs & lack of eorroboration, contrary to the official South Cheshire Rape Crisis 11/9/97) or video- guidance. In addition, Lees and Gregory (1996: identification parades (submission fromjill 95-96) found that there were now four key Saward 29/8/97). The first two options were attrition points, where: discussed in section 2 regarding witness intimidation. The third carries the advantage that • the police decide whether to no-crime an the victim can progress at his/her own speed thus incident; minimising the possibility of trauma, but may be more expensive.

169 At the investigation stage much criticism has also arrangements for screening victims for S 1 0% been levied at medical examinations. Medicals are recruitment of more female doctors to conduct routine when the victim has reported within the examinations, and rhe creation of examination sufficient time for some physical evidence to be suites. Another I lome Office Circular followed in collected. Past criticisms have included that 1986 (1986/69), highlighting the need to recruit examinations have been carried out badly, losing more female police surgeons, the value or special vital evidence: this may be partly due to lack of victim examination suites and need to provide collaboration between the investigating officers complainants with more information. and doctor. It may also be related to another complaint, that police surgeons have formed their Since these developments there are some own views about the complainant's veracity. One indications that matters have improved. One is consequence is that doctors have been reported as the number of rapes reported to the police (see making insensitive comments to the complainant. above). Another example is the widespread Lack of sympathy and even hostility have been creation of rape examination suites, although it is reported. Finally concerns were expressed that not known how many of these suites exist. A victims were not being given important advice on survey by Victim Support of victim support pregnancy and sexually transmitted diseases schemes (1996: 12-13) found that 75% thought (S'I'Ds), or other follow-up support. victims were always or usually seen in a rape suite, 11% "'sometimes" and only 1% "never" (the rest These criticisms have been hacked up by research were "don't knows"). Many schemes also suggested with rape victims. In each case, the comments there was little contact from the police after received were mainly negative. For example, in reporting and little information about progress. one Scottish study (reported by Temkin), forty- However, it is nor known how representative these five sexual assault victims who had medicals were findings are. interviewed. This yielded 47 negative comments, 1 5 neutral and 12 positive. The negative One of the few studies evaluating progress was comments mostly concerned procedures being undertaken by Temkin (1996: 1-20), who painful and unpleasant: others concerned the interviewed fourteen women about their police surgeons' manner and conduct. Temkin experience of medical examinations from 1991 to also reports another study (Clare Corbett 1987) 1993. This small sample demonstrates the with 22 rape victims which confirmed some of difficulties of finding sufficiently large samples of these findings. Other concerns raised included rape victims to draw reliable generalisations. that: Originally lemkin aimed to interview between twenty and thirty of the 149 women who had • many victims were not asked it they would reported rape to the Sussex police in 3 992 and prefer a female doctor, but being touched by 1993. f lowever, most were ruled out, because a man so soon after the rape oficn added to cases were still pending, or the victims could not their trauma; be contacted for example. To boost the sample additional cases from 1991 were included. Temkin found no clear distinction between those who • some were examined in a police cell/ or agreed to be interviewed and those who did not. office; Nevertheless, there is no way of knowing whether the experiences of the women interviewed were • in some cases officers walked into room representative of rape victims across the country. during examination, or even that the door was left ajar; and finally Temkin found that most victims were more • no washing facilities were available positive about their treatment by the police than afterwards. by doctors. "The medical examination appears to be experienced by some as a further sexual assault These complaints led to a number of responses, and an ordeal in its own right" (Temkin 1996: including a Home Office Circular being issued in 14). More specific findings included that: 1983 (1983/23) which emphasised the importance of allowing complainants to was and • There are still problems in the provision of change as soon as possible. The Metropolitan female doctors. Provision is patchy but this police set up a Working Party, which led to special is at least partly because there is a dearth of

170 female doctors willing to do such work. Metropolitan police abandoned raking Police may face a difficult balancing act, routine samples of head hair in 1990. in between difficulties finding a female doctor addition, recent Home Office research and [he time added to the complainant's indicates the police routinely take non- wait. intimate samples such as mouth swabs from suspects, and chat intimate samples such as • Most complainants were still negative about blood and pubic hair are rarely needed doctors' attitudes, particularly char they were (Bucke and Brown, 1997: 41-47). perceived to have formed their own beliefs about the complainant's honesty. Being A submission to the review from London Rape believed is very important to victims. Crisis (LRC, 11/9/97) goes even further. They However, there was some evidence that the argue thar not just doctors but all officers who are women felt less negative than in some likely to come into contact with a survivor, from previous studies. the reporting stage onwards, should be female. The only exception would be if the victim states • Overall, the study suggested that there had that they would prefer a male officer. In addition, been some improvements. For example, LRC argue that all these officers should undergo none of the women were examined in a rape awareness training. These measures could police cell or office, most were examined by however, present some organisational problems: female doctors, and in most cases a leaflet considerations include whether there would be on pregnancy and S I Da was provided. sufficient female officers willing to do this work, However, the vast majority still made how it would be renumerated and how it would negative comments. fit into current career structures. Similar issues are raised by the recommendation that victims should Based on these findings, Temkin made a number be provided with trained chaperones throughout of recommendations: the criminal justice process (ie. with the same person assigned to a particular victim • More female doctors should be recruited. throughout). In addition, it is not clear who Female doctors mas' not always handle a would take on this responsibility. medical sensitively, but victims should have a choice about who conducts the medical. Other suggested measures have included medical This may require reconsidering the way such follow-ups tor victims and providing victim packs work is funded. At present in most forces, as soon as possible after the report (submission fees are only paid for each examination from Jill Saward. 29/8/97). The pack could conducted: paying an additional fee for include toiletries, a booklet on legal processes and doctors to be exclusively available over a set other sources of support- and other information. period might improve availability of doctors. Separate packs could be provided for male and female victims. • Doctors' training should cover rape trauma syndrome and counselling techniques. Finally, Lees argues that HIV raises a number of Questioning by doctors should also be issues yet to be addressed, such as whether a minimised by greater collaboration with complainant should be able to demand a .suspect investigating officer. This will help the is tested for the virus. There are also questions victim avoid being upset by having to retell about whether there should be a duty on the their story, and also avoid possible problems police and other agencies to inform the at trial presented by having too many complainant when they know the suspect is HIV- accounts of the same events. po.sitive. Lees describes a case where the complainant was nor informed by the police that • Some procedures could be eliminated. For the suspect was HIV-positive, even though they example routine plucking of pubic hair for knew this for some months (1996: 17-18). DMA tests is unnecessary and upsetting for Government plans have recently been announced victims (DNA can be obtained from blood to introduce new laws targeting people who and if a hair sample is needed it could be deliberately spread life-threatening infectious taken at a later dare). There is a precedent diseases such as the AIDS virus (The Independent for this kind of change: for example, the 8/2/98). However, Lees argues that the various

171 jusrice agencies need formal guidelines and • consent; procedures to help deal with the special issues char HIV raises (19%: 256). • cross-examination: and

c. The decision to prosecute • corroboratton warnings. The literature found research on the decision ro Pre-trial preparation prosecute is lacking, i iowever, the main problem in deciding whether ro prosecute sexual oHences A number of measures have been discussed in appears to be proving consent. This is previous chapters which might also he considered complicated further by the idea of undeserving for victims of sexual offences, including: (acquaintance or date rape) and deserving ("real'; stranger and virgin) victims. Changes in [he legal • court familiarisation visits; definition lie outside the Reviews terms of reference, and are not considered further here. • access to statements in good time before I lowever, it is worth noting that it has been attending court, suggested that the standard ol" proof in rape cases should he altered for different types of rapes. For • separate waiting facilities; and example, that date rape should be distinguished in law from stranger arracks (Jill S award submission • friend in court .schemes. ro the Review, 29/8/97). Other measures which have been suggested Other measures which have been suggested specifically res-aiding sexual offences include include: meeting with defence/prosecution before trial (see for example Lees 1996: 253). The scheduling of • establishing a separate unit in the CPS to the court case may also be important. Concerns deal with all cases of sexual violence include that this should take into account how (submissions from London Rape Crisis and well a witness feels able to face court (London North Staffs and South Cheshire Rape Crisis Rape Crisis submission 11/9/97), and also the 11/9/97); need for faster court dates to avoid prolonging the victims suffering and aid their recovery {Cleveland • notification of the victim if the case is Rape and Sexual Abuse Counselling Service dropped, explaining why (Cleveland Rape submission 10/9/971. The latter is particularly a and Sexual Abuse Counselling Service concern when counselling is postponed until after submission 10/9/97); and the [rial 10 avoid possible witness contamination (see chapter 3): this is discussed further in the • a right of appeal against CPS decisions ne section. (CRSACS submission op cit). xt Anonymity d. The trial The trial has widely been described as equally bad Protection of anonymity is also an issue In these an experience as the original offence for victims of cases. The extent to which anonymity can be sexual offences. Lees (1996: 36) found that eight provided is a complex issue. As noted In earlier out of ten rape complainants felr rhat they were chapters, a balance must be struck between on trial rather than the defendant. In some protecting the complainant, but also being even respects the trial was actually seen as worse than handed. Simple measures discussed in more detail the rape itself: "more deliberate and systematic, elsewhere include: more subtle and dishonest, masquerading under the name of justice". Five areas are discussed below:

• pre-trial preparation; It was noted in chapter 3 that the fact that statements are not written by the complainant may cause problems for people with • anonymity; disabilities or illnesses. Lees (1996: 107} found that some rape complainants also complained about this, particularly the assumption in court that the statement was their own words.

172 • reporting restrictions and clearing the court; One response could be to require defendants to prove that they obtained proper consent, rather • not having address or other Identifying than char they did nor (CRSACS submission information read out in court; 11/9/97). Again it could be argued that this would conflict with the principle that a defendant • use of screens, cctv and the like to avoid the is innocent until proven guilty. Another approach victim having to see the accused. would be TO change the burden of proof by requiring that a reasonable person should have However additional issues have been raised known the victim was not consenting (Lees 1996: concerning sexual offences. For example, it has 256). been argued that even when defendants are acquit led, suspicion may remain among work Other possible measures relate to the doctrine of colleagues, family, friends, and other local people recent complaint and the admission of sexual (the "no smoke without fire1' factor). To avoid history evidence, which both address consent: penalising people who may in t]\ci be innocent, it has been suggested that defendants should also be 1. The doctrine of recent complaint given anonymity in the media unr.il and unless they are convicted. Denying them anonymity The common law doctrine of recent complaint conflicts with the principle that a defendant is originated in the middle ages. Under this rule, the innocent until proven guilty, which stems from fact rhe alleged victim complained shortly after the idea rhar it is preferable for ten guilty people the offence is admissible as evidence for the to go free than for an innocent person to be prosecution co enhance the complainant's convicted. credibility. The details of what was said are also admissible. 'I his doctrine is based on two assumptions: Against this it can be argued that the conviction tate for rape is so low that an acquittal should not • that a normal woman would naturally necessarily be regarded as proving the defendants complain quickly after being raped, lor innocence (see for example Lees 1996: 132). According to this line of reasoning, allowing sexually assaulted in some way); and defendants' anonymity could do more harm than good. The balance has swung too far in favour of • that women are prone to making up false the guilty: the number of guilcy people walking allegations of rape. free must now be .so great char the assumption of innocence needs to be reconsidered regarding Both of these ideas have been proven false, bur rape. The argument is that although a few judges may continue to warn juries of rhe danger innocent men might .suffer by having their of convicting when a complaint has been identities made known, the far greater number of delayed. A recent judgement on this doctrine in guilty men who are acquitted would pose a greater New Zealand called this a "perverted survival'" danger if granted anonymity. (The Queen vll, 1996). Lees (1996: 252) suggests rhat juries should instead be warned chat absence of recent complaint should not be seen as Consent evidence that rhe complainant is lying, and that there may be good reasons not to complain. This It was noted earlier that most sexual offences approach is followed in New South Wales, hinge on consent. Given thai the opportunity for Australia. consent (or failure to consent) rends to occur in private, this is a very difficult matter to prove Another measure would be to admit evidence on beyond a reasonable doubt. Lees (1996: xvii) the effect of the offence on the victim. This could argues that the onus is usually on the complainant be in the form oi a victim impact statement. Lees to prove she (or he) did not consent rather than (1996: 31) found victims commonly complained on the defendant to prove she (or he) did consent. they were not allowed to explain fully what had This stands in contrast co other offences such as burglary where it is usually assumed that the complainant is telling the truth (Temkin 1997: 116). "Preliminary findings from research currently being conducted for the Home Office suggests that similar issues also arise in child abuse cases (Davis 1997: 7).

173 happened to them. Alternatively (or in addition) the Court of Appeal case Viola (1982), it was expert evidence could be admitted, in particular ruled that if evidence of past sexual history was on Rape Trauma Syndrome (RT5). Expert relevant to the issue of consent, it was admissible. evidence on RTS has been used in some American Deciding relevance has proven problematic jurisdictions in two main ways: to prove lack or though: "there are certain areas of enquiry where consent and to explain behaviour the jury might experience, common sense and logic are informed otherwise view as evidence that rape did nor by stereotype and myth" (Supreme Court of happen. Canada ruling cited by Temkin 1993:5). In practice, it is widely acknowledged that such In practice, there is some disagreement over evidence is frequently allowed, even where this whether evidence of RTS should be admissible to appears to be in contradiction of the spirit of the prove lack or consent. Where it is admitted, irs legislation (for example where it is used to blacken role is limited. For example, in West Virgina in the complainant's character rather than relating ro the case of the State v McCoy (1988), it was stated consent). This is supported by some research: for "[t]he expert may testify that the alleged victim example Adler (1989: 73) found that applications exhibits behaviour consistent with rape trauma for admission of sexual history evidence were syndrome, but the expert may not give an made in 40% of fhe rape trials she studied, and opinion, expressly or implicitly, as to whether or 75% of these were allowed. More recently, Lees not the alleged victim was raped'' (quoted by (1996: 31) found that over half of all female Myers & Paxson 1992: 3). However, Myers'& acquaintance rape complainants in her study were Paxson suggest that most courts in America allow questioned about their sexual history with men evidence of RTS to explain the complainants' other than the defendant. She found that in some behaviour where the jury might misunderstand it cases questions on sexual history are asked (eg, delayed reporting). No research was fount! on without even requesting the judge's permission admission of RTS evidence by the courts in this (1996: 160). country, but it appears that admitting this evidence may be problematic. Once the Similar problems with sexual history evidence prosecution admits such evidence, the detence have been experienced in other jurisdictions may find their own expert witnesses to contradict them. (Temkin 1993: McDonald 1994). Some such as New South Wales, Canada and Scotland have attempted to tighten up the rules, and the Labour ii. Sexual history evidence Party made a commitment while in opposition to do the same here (cited by Temkin 1993: 20). The Sexual Offences (Amendment) Act 1976 was When in opposition the Labour Party did try to intended to restrict admission of sexual history introduce further restrictions on sexual history evidence in rape cases where this (according to the evidence through an amendment to the Criminal Heilbron Report upon which the Act was based) Procedure and Investigations Bill in 1 996 based "does not advance the cause or justice but on the New South Wales model (Hansard 1 2 June effectively puts the woman on trial" (cited by 1996: 356-368). This approach has its critics who Temkin 1993: 3). The concern was that evidence argue that a more narrow definition of consent is of past sexual relationships with other partners needed to avoid the possibility of sexual history was used to determine consent to the defendant. evidence creeping in through the back door. This This seemed particularly unfair given that raises issues about the definition of rape, which evidence of defendant's past convictions is not are beyond the Review's remit (discussion paper admitted for fear of prejudicing the jury. It was 1? prepared for the Review group). proposed that evidence of past sexual history should only be admissible when it involved the The problem is how to restrict the use of sexual defendant. The one exception would be if there history evidence effectively. Temkin (1993: 3-20) was a strong similarity between the complainant's sexual behaviour on a previous occasion and that on the occasion when the offence was alleged. Adler randomly selected 50 rape trials, representing 85% of all rape trials heard in the Old Bailey in one year. Of these, 5 did However, the wording of the Act was vague, not go ahead "usually because the victim was unable to give suggesting that evidence should only be admitted evidence"(Adler 1989:39, 73). it if was unfair to the defendant not to admit. In See Temkin (1993, 7-20) for a description of the Canadian reforms.

174 argues that judges' use of discretion has delaying reporting the offence, i lovvever, evidence undermined previous c fro res TO restrict the use of has tended to be anecdotal, and trial lawyers have sexual history evidence in this country, suggesting argued that the rules on cross-examination are that any new reform would have to curtail this basically the same for sexual offences as other discretion. Similarly, Lees (1996: 251) offences. recommends that judges discretion should be reduced. She argues that if the defence raise the To redress this Brereton (1997: 242-261) complainant's sexual history or criminal record, conducted a study of trial transcripts in forty rape the defendants sexual hisrory or criminal record cases and forty-four serious assault cases. There are .should be introduced. important differences between the two offences, including that there are more likely to be other Unfortunately the literature review found research witnesses of assaults. Despite this, rape has more on the effectiveness of reforms in other in common with assault than other offences such jurisdictions is lacking. Only one such research as robbery or burglary: for example, in both cases study was found. This examined changes the offender is often known to the victim. introduced in Scotland in 1986, which listed Brereton found some significant differences in certain forms of evidence which should not be questioning rape and assault complainants, such admitted and specified those which could. 1 he as: researchers used information from court records, interviews, observation and forms completed by • sexual history evidence: sexual history (with Court Clerks to collecr details of the use of sexual the defendant or people other than the history evidence. They found that the reforms defendant) was raised in about a third of the seems have had .some success, but there were still rape cases, but only two of the assault three main problems which would need to be cases. addressed if this model is followed in England and Wales (Brown, Burin an and Jamieson 1992: 60- • the amount of time spent on the witness 78): stand: "on average it took about twice as long to cross-examine complainants in the • in some cases the rules on sexual history rape trials as it did in the assault trials" evidence are not being followed; (1997:257).

• the rules are sometimes followed without However, Brereton argues there were also some achieving the aims behind them; and strong similarities in the strategies used in cross- examination: • the rules failed to address subtle character attacks, which continue to be employed. • Assault victims were just as likely to have their character and credibility attacked (for Cross-examination example through questions about drinking and mental stability). There have been two main concerns about cross- examination in sexual offences: first, that victims • Attempts were made to exploit of sexual offences are cross-examined more inconsistencies in the complainants' severely than for other crimes, and secondly, that statements in both cases. defendants have the right to cross-examine complainants personally. • If assault complainants did not behave as expected (eg. reporting soon after the 1. Severity of cross-examination otrence), this was raised in cross- examination, as happened in the rape cases. According to this argument, the focus on the issue of consent and the fact that most sexual offences Brereton acknowledges that, because of the more occur in private, results in greater emphasis and intimate subject matter in rape trials, the length of closer examination of the complainants words, time spenr under cross-examination and the character and motives. Defence cross-examination seeks to show that the complainant did not behave as a real victim would, for example in This wmasbefore changes to rules on the admission of sexual hisrory evidence in Australia in 1991.

175 nature of the offence itself, rape trials are generally for the judiciary as recommended by the more traumatic for the complainant than assault Royal Commission and revoking the rule trials. Brereton concludes that coo much attention that a judge cannot be sued (Lees 1996: has been given to improving rape trials in 247-50, 253)- particular, and that more attention should be given to the weaknesses of the trial process ii. The defendant's right to cross-examine generally. It should be noted that the study was personally conducted in Australia, and it is possible chat Concerns about the defendants' right to cross- cross-examination of rape victims there is not examine the complainant personally have been typical of thar in England and Wales. highlighted by a couple of recent rape cases which were taken up in media. In one case the defendant Other qualitative differences have also been cross-examined the complainant for six days observed by Lees (1996: xxi), who argues that wearing the same clothes as when he committed although both the complainant and defendant's the offences. In another example where there were replications are attacked in rape trials, it is the a number of co-defendants, a Japanese student complainant's sexual reputation but the was cross-examined for twelve days (see for defendant's occupation that are examined. This is example The Guardian, The Daily Telegraph & of particular concern given thac other research The Mirror 23/8/96). Such cases are rare. At suggests sexual reputation may not be based actual present almost all defendants receive legal aid, and sexual activity but on dress, linking independence only 4% are either privately represented or and having a number of male friendships for unrepresented (Home Office discussion paper for example (Lees 1993 cited by Lees 1996: 86). the Review). Nevertheless, allowing the defendant Possessing previous convictions is seen as relevant ro cross-examine personally may be extremely co the complainant's reputation, but rarely upsetting for the victim. In addition, it has been allowed as evidence of the defendant's credibility. suggested that defendants are allowed to pursue Likewise, if the complainant does not have a lines of questioning thar would not usually be criminal record this will not count in her (or his} accepted, raising additional questions about favour, although ic may be seen as relevant to the whether justice is served in these cases. defendants credibility. The obvious implication is that the right to cross- Of course, it is possible for che judge to intervene examine personally should be removed. However, co halt inappropriate questioning. I iowever, this this raises some complex legal issues. The right of has to be set against comments made by some the accused to defend him/herself either members of judiciary in sex cases which suggest at best lack of sympathy (see Adicr 1 987 for some personally or through a legal representative is examples). Although such comments may not be protected by che European Convention for the representative of attitudes among the judiciary, Protection of Human Rights. The only exception media reports of insensitive remarks may to this rule under English law is an automatic discourage people from reporting. Possible prohibition from personally cross-examining child measures might include: witnesses when the defendant has been charged with an offence involving sex, violence or cruelty. So there is a precedent under English law for • training for the judiciary (and perhaps also limiting the defendant's right to cross-examine other court staff and the legal profession personally, and this has not been challenged under more broadly} about rape; the European Convention. • setting up a complaints procedure for If i is accepted that the defendants rights should complainants who suffer inappropriate t be further curtailed in this area, another issue is cross-examination, which would include what should happen when a defendant refuses penalties for the barristers concerned: legal representation. There is no legislation or guidance governing what should happen if a • appointing more women to (currently) male defendant refuses legal representation in cases dominated judiciary; involving child witnesses. There is also little information on what actually does happen. In at • increasing accountability, for example by least one case che judge has conducted cross- instituting a performance appraisal system

176 examination for rhe defendant, bur this could call questioning and often attending idenrificarion che judge's impartiality inro doubc. In another parades" (1996: 1 10-1 11). case no cross-examination was conducted (I lome Office discussion paper for the Review).This Lees argues that judges's discretion ro give a again is problematic: it seems questionable corroboration warning should be removed, which whether it should be possible to convict in the would follow the precedent set in Australia (1996: absence of cross-examination. At the same time 251-252). Adier (1987: 161) similarly argues that though, abandoning such cases could encourage there are plenty of other safeguards both against more defendants to reject legal representation. false allegations generally (such as police questioning, cross-examination) but also for rape One option might be to impose legal in particular {such as medical examinations). As representation on defendants in these cases (David well as being upsetting for the victim, rhe Pannick, The Times 10 Sepu 1996). This is comments made may reinforce rape myths in the followed in some other jurisdictions such as Italy. jurors' minds. In Croissant v Germany, this approach was ruled as compatible with the European Convention on e. ...and beyond Human Rights. This raises the issue of whether defendants should have to contribute towards the Beyond the trial, a number of other issues are coses ok such representation in the usual manner raised, including: under the legal aid scheme. To impose this on them seems unfair, but at rhe same time, it can be • therapy; argued that not requiring defendants to contribute in these circumstances might • compensation; and encourage others to refuse legal representation. Although prohibiting personal cross-examination • information. by defendants in rape cases seems desirable to protect vulnerable victims, conceiving a scheme to accomplish this is a challenging task. The issue of when therapy should it be allowed was discussed in chapter 3 on disabilities and Defendants' access to victim statements raises illnesses. Therapy is a particularly pertinent issue similar concerns. The CRSACS submission in sexual offences. It should not always be (Cleveland Rape and Sexual Abuse Counselling assumed that therapy will be beneficial: for Service 10/9/97) suggests there is evidence that example. Lees (1996: 18 & 20) details two cases these have been given to accused and used as in which responses were unsympathetic and pornography in custody. This again suggests a argues that counsellors need special training, conflict between providing the accused with all which should cover H1V7A1DS. However, there is the information they might need to defend concern that victims who could have benefitted themselves, and protecting the witness. from therapy are discouraged from obtaining it for fear of damaging the prosecution case. It has Corroboratton warnings been argued that therapy should not be discouraged prior to court (South Essex Rape & Until 1995, the law required judges to warn juries incest Crisis Centre submission, 29/8/97). about rhe danger of convicting on uncorroborated However, to ensure that rhis happens some evidence in sexual offences (Adler 1987: 161-2: reassurance will be needed that this would not be Lees 1996, 109). This was again based on used against the complainant in court. misconceptions about the prevalence of false accusations of rape. Although this requirement Compensation has been abolished, it is still possible for .such warnings ro be made. It is not known how After the trial, victims can apply to the Criminal frequently this happens. However Lees' research Injuries Compensation Authority for suggests that where the warning is given judges compensation for the harm done by the offence. often add comments claiming that allegations are L It has been suggested that the current system easy to make but difficult to disprove, 'when in should be reviewed. Specific proposals include: practice it involves a long, arduous process lasting several days, medical examinations, days of police

177 • extension of compensation to include likely to report to the police than those who did emotional trauma; and not. "The 'real' reporting rate, then, may be much lower" (Mayhew, Aye Maung & Mirrlees Black • increasing tariffs for victims or sex offences; 1993: 96). This suggestion is supported by other research evidence (Dobash & Dobash 1979: 164), • clearer definitions of sexual assault (South suggesting that women may experience many Essex Rape and Incest Crisis Centre attacks by their partner before calling the police. submission, 29/8/97). It is difficult to measure the exact number of incidents which have gone unreporred: when a In Formation large number of incidents have occurred, the victim may lose count. This finding does suggest Finally, providing witnesses with information though that when a victim of domestic violence about the progress of their cases throughout contacts the police the response they receive is all system is important. This includes notification of the more important. trial dates in good time: lack of notice has been criticised for adding to victims' anxiety about the Against this, Victim Support (1992: 1.5) argue trial (Adler 1987: 165). After the trial! and that: "there is a brutal but common following a custodial sentence, witnesses need misconception that if women do not leave, the information about the release of the offender. violence they are enduring cannot be all that Under the Victim's Charter (Home Office 1996d: intolerable'' and that this may colour reactions to 12) the Probation Service should notify victims of them. There are numerous reasons why victims of the offender's release from custody, but this only domestic violence may stay. One is the economic applies to those given life imprisonment or effect: if they are not thrown out or the shared convicted of serious sexual or violent crime. One home victims may feel they have to move out to option would be to extend this to other victims. avoid retribution from the offender. The victim However there are concerns that the existing may have to face moving to a new area, leaving scheme does not always work properly, suggesting many of their belongings, uprooting children and that the existing arrangements (including the having to find a new home, schools and possibly requirement on victims to opt m) should be work- Other possible reasons include fear of being reviewed. pursued by the attacker and post traumatic stress, disorder. Ethnic minority victims may face 2. Domestic violence additional hurdles such as language barriers, concerns about immigration status and cultural a. Reporting pressures against reporting (for example see Choudry 1996: ). Failure ro report 1-4 Another factor influencing reporting decisions It is generally accepted that most domestic violence is not reported to the police. The BCS may be negative experiences when the police were suggests reporting to the police has increased in contacted in the past. Numerous criticisms of recent years: 30% of respondents said they had police responses to domestic violence have been reported domestic violence in 1995 against 20% made in the past, including (Buzawa & Bu*zawa in J981 (Mirrlees-Black, Mayhew & Percy 1996: 1996; 37-38; Grace 1995; 1): 29). Despite this, it was found thac domestic violence is half as likely to be reported to the • failure to attend or, when they do attend, police as muggings (60%), although the gap was that they are slow to respond; narrower for stranger and acquaintance violence (39% and 37% reported respectively). The 1992 • viewing the problem as civil not criminal survey suggested that within this, male victims (reflected in "no-criming" or treating were twice as likely to report as women (40% offences as public order offences rather than against 21 %, Mayhew, Aye Maung & Mirrlees as violence); Black 1993:96). • reluctance to get involved and lack of sympathy for the victim, sometimes However, the victims who disclosed domestic reflected in attempts to reconcile rhe victim violence to interviewers may have been more and perpetrator or to side with the assailant;

178 • low arrest rates, even when the victim has • iorces should liaise with other statutory and been seriously injured; voluntary bodies to ensure a common approach; and that • failure to recognise that the incident ts usually just one in a series. Different officers • there should be a presumption in favour of may respond each time a call is made from arrest (Grace 1995: 1). any one property, and will often not be aware of previous visits by the police. Grace (1995: 53} looked at the impact of this guidance. She found almost all forces had changed To some extent the reasons for this poor response their policies in response. There was also evidence have been historical. For example, in some of improvements: for example, in increased jurisdictions physically punishing your wife was understanding and sympathy for the victims, and not illegal as long as the rod used was no thicker more positive responses including advice and than the husband's thumb. This was reinforced support for the victims. Nevertheless there with ideas of the police function as enforcing appeared to be a gap between policy and practice. public order and the sanctity and privacy of the Despite managers confidence that the new- home, which continue today to some extent. In policies had been successfully filtered down to modern times, this has been translated into views frontline officers, many operational officers were about domestic violence not being "real police unaware of the new policies. Few had received any work'' in a culture which does not value social specific training on domestic violence, although work, instead focusing on crime fighting. In other agencies interviewed (such as women's domestic violence cases, the offender is known, so refuges) were willing to help with such training. relatively little or no detection or investigative skills are necessary. Consequently Buzawa and One example of the discrepancy between policy Buzawa (1996: 37-40) -suggest that arrests for and practice is arrest. Although pro-arrest and domestic violence are not seen as counting: results even mandatory policies have been used in other are equated with arrests and prosecutions. Added countries (such as America) to increase arrest to this is a perception that victims of domestic rates, past evidence has .suggested that the police violence are likely to drop charges. Finally, Buzawa and Buzawa (1996: 40-43) suggest that may be reluctant to arrest even if the complainant domestic calls are viewed as dangerous by officers, asks them to (Smith 1989: 57). One reason given who fear offenders may turn their aggression on has been the perception that domestic violence them. This risk, is they argue, probably victims are likely to withdraw their complaints, overestimated. and that the work in processing the case will have been wasted. Coupled with this perception was a lack of understanding by the police of victim's Encouraging reporting reasons for withdrawing complaints, .such as fear of retaliation by me offender (Edwards 1989: In response to these criticisms, a number of efforts 100-103). Other research suggests that this have been made to improve police responses to becomes a self-fulfilling prophecy: the police may domestic violence. For example, two circulars actually discourage complainants from proceeding were issued by the Home Office (69/1986 & by repeatedly asking them whether they wish to 60/1990), which made a number of continue and allowing a cooling off period to recommendations. At the reporting stage, these think things over (Chambers and Miller 1 983 included that: cited by Smith 1989: 57, and Farragher in Pahl [Ed] 1985: 110-124). Grace's (1995: 20-21) • procedures used for victims of sexual assault research suggests reluctance to arrest still should be applied to domestic violence continues. Few operational officers and managers victims; saw arrest as paramount: instead it was typically seen as the third or fourth priority. • protecting the victim from future risk of violence should be prioritised over There are problems with pro-arrest policies. For attempting reconciliation: example, the National Inter-Agency Working Party on Domestic Violence convened by Victim Support in 1990 (hereafter referred to as "the Victim Support Working Part)"") was against automatic

179 arrest on the grounds rhar domestic violence should by improved responses to reports, referrals to the be treated like other violent crimes, Greater police from other agencies will be minimal. numbers of arrests do not automatically translate into tougher approaches in the rest of the criminal In the past the police have been criticised for justice system, nor do they necessarily deter re- rarely referring complainants to other agencies, offending (Polsby 1992: 250-253; Sherman et al despite evidence that complainants would 1 992: 680-690). Arrest is nor Therefore a panacea. appreciate this assistance (Smith 1989: 53). There However, it does .send a message to offenders and was some evidence in Graces study that forces victims abour the seriousness or such incidents. The were giving information to victims about other Viaim Support Working Party (1992: 2.15) sources of support. For example, some forces had suggested that pro-arrest guidelines could be useful, (or were preparing) leaflets, which included and recommended that this be a matter ihv national contact derails for Victim Support and other policy. sources of help and advice. The Victim Support Working Party recommended that all officers Grace's (1995: 47) interviews with refuge staff should carry a small card or leaflet giving contact suggested one reason for low arrest rates could be details for local support groups, to pass to victims lack of knowledge of law in this area, particularly when the attacker is not present. The Victim civil law. Possible measures to tackle this again Support Working Party also recommended that include training. Less than a quarter of the victims should be reminded subsequently that operational officers in Grace's study said they had they could contact such agencies or that the police received specific training on domestic violence, could do this for them, and that posters and although two-thirds of managers thought that leaflets should be available in police stations special training was needed. However less than (1992:2.11-2.12). half of the operational officers thought this would help. It is not clear why this difference of opinion Victim Support has police representatives on all existed, but it might perhaps reflect greater it's Wai management committees. However, awareness of policy changes among management. Grace (1995: 47-48) also interviewed Another measure would be to issue officers with representatives of some organisations (such as flash cards about their powers and victims rights. womens refuges) who suggested that they had Less than a third of operational officers in Graces very limited contacts with the police. One study had these cards, and only seven managers consequence of this may be th.it police thought their officers had been given one (Grace tnisperceptions about the support available to 1995:25). victims of domestic violence are not challenged. For example, one respondent suggested that the Part of domestic violence officers work (where police tend to think women's refuges only provide they exist) is to refer victims of domestic violence emergency accommodation. to Victim Support, refuges and other possible sources of advice. This has two advantages. First, More recently, Hague Males and. Dear surveyed it helps provide victims with support and advice. multi-agency work on domestic violence in all Secondly, many victims may contact these local authority areas. Their findings suggest that agencies initially rather than the police. police involvement in inter-agency groups has Developing relationships with other statutory and improved since Grace's study. Police Involvement voluntary bodies may encourage referrals from was much higher than that of the probation these agencies to the police. For example, the service, and particularly the CPS and courts. In Victim Support Working Party report (1992: addition, police representatives tended to be from 6.28-6.32) highlights the importance of health senior ranks, which may have assisted policy- professionals such as Accident and Emergency change and signified to other officers the Department staff in identifying domestic violence. importance of treating domestic violence The report stresses the value of guidelines and seriously. The researchers suggested this work training covering issues such as identifying victims could be built into job specifications for some of domestic violence and careful documentation staff and that commitment from senior staff in of injuries. Similarly local hotising authorities these organisations might assist. However, they have a part to play in assisting victims find accommodation. It may be argued though that also say that in some areas the police tend to unless liaison with other agencies is accompanied dominate multi-agency groups and in some areas concerns persisted about the policing of domestic

180 violence (Hague, Maios and Dear 1996: l7.1- domestic violence as solely the problem of that 23.3: 52). The research did not examine the unit or those officers. However" there are ways or reasons for this, bur possible explanations might combatting this. For example, in one force in include insufficient feedback to frontline officers Graces study, uniformed officers were attached to and insufficient contacts between frontline the DVU for between three and six months. officers and rhe support agencies. Recording and investigating as for other violent b. Investigation crime Other recommendations in rhe Home Office On the second point, research by Phillips and Circulars mentioned above were mat: Brown (1998: 80, 91) found the charging rare for those arrested for domestic violence (58%) was • forces should establish specialist domestic higher than for all offences (52%). Cautioning violence units or officers: and and NFA ("no further action") rates were similar to rhe average for all offences (13% cautioned and • domestic violence .should be tecorded and 20% NFAed, against an average or 17% and investigated in the same way as other violent 20%). However, for domestic cases of violence offences. against the person, the charging rate was even higher (71% compared to 67% for all violence Specialist domestic violence units/officers against the person). By contrast, Grace (1995: 12- 13) found a 81% charging rate, but because of On the first point, Grace (1995: 5) found just poor recording practices suggests caution when over half of all forces had specialist units dealing considering this finding. The results from another with domestic violence, but only 5 were dedicated data collection exercise within the same study Domestic Violence Units (DVL's): others were suggested a rate of 60%, but the sample size was Family or Child Protection Units. This runs smaller. contrary to the advice of the Victim Support Working Party, which .suggested rhat "in joint As well as the frequency of charges, the nature of units child abuse tends to consume all rhe the charge is Important. Although Phillips and available resources because dealing with it is a Brown did not examine downcharging, previous statutory responsibility". Although there may be research (sec for example Edwards 1 989: 73, and benefits in working with Child Protection Units others - Smith 1989: 43-44) has suggested where children are involved in domestic violence, domestic violence is commonly downgraded to the Working Party recommended that domestic less serious or non-violent offences such as breach violence units should be separate from family or o!" the peace. It was observed above regarding child protection units (1992: 2.21-2.23). sexual offences that there may be legitimate reasons for downgrading some offences, but the A number of problems were found in those areas concern is that on many occasions such where domestic violence officers did exist. For justification is lacking. More recently Home example, the role or domestic violence officers includes keeping victims and uniformed officers Office circular 60/1990 has reminded officers of the range of powers they can use for domestic of case progress. However, Grace found formal procedures to facilitate information exchange violence. However Graces interviews with police tended to be lacking. Other problems included officers suggest domestic violence was dealt with that officers working alone felt their work was not as breach of the peace in the vast majority of cases, given appropriate priority. Excess workloads were even in some very violent cases. About two-thirds also complained about- In some cases, there was of officers saw policing domestic violent as just one .specialist officer, meaning that there was different ro policing other violence, although no cover if the officer was not on shift, took leave most other interviewees did not draw any or was sick. To combat these problems, Grace distinction (1995: 19). (1995: 55) recommended that domestic violence officers should work in pairs. Phillips and Brown also examined what happened after charges were laid regarding police bail. In 60% of cases bail was refused, substantially higher Establishing specialist units or officers does carry a than the 22% rate for charged suspects generally, danger that other officers will compartmentalise probably because of the potential risk to their

181 partners if released on bail (internal note, 1997). victim's evidence. A pilot scheme (The Observer More research is needed on the use of hail 8/2/98) is planned which would attempt to do conditions In demesne violence and the extent of just that: this will involve developing other forms enforcement in the light of evidence of police of evidence such as taping emergency calls and unwillingness to enforce civil injunctions11. raking photographs of the scene, and retraining police officers and CPS staff to consider new c. The decision to prosecute forms of evidence. The second phase of Phillips and Brown's (1998: d. The trial 132, 141 -142) research has examined case progress following forwarding to the CPS. Civil measures Preliminary findings are currently restricted to i. termination by the CPS. These suggest domestic Although the civil courts are beyond the remit of violence cases were much more likely than cases the Working Group, some discussion of civil generally and more likely than other violence to measures may be useful ror several reasons. First, be terminated by the CPS fa termination rate of civil measures may be taken before or in 37% for domestic violence, against 14% for all conjunction with criminal measures and therefore offences and 29% for violence). colour the complainant's experiences of the criminal justice process. Similarly, findings about The main reasons for ending domestic violence personnel in the civil system (such as the police cases tended co be different to those for and legal profession) may be informative about terminating other offences: in half of domestic how they behave in the criminal justice context. violence cases refusal ok. a key witness to give Another reason is that disillusionment with the evidence was given as a reason for not proceeding criminal justice system has led to suggestions that (internal note 1997). The prosecution system the civil system has more to offer victims of relies heavily on the complainant's evidence in domestic violence. Important differences are that domestic violence cases, where there are often no in the civil system the victim has to initiate other witnesses. It is possible to proceed without proceedings, the civil courts have a lower standard the victim's agreement, by compelling the of proof, and civil measures tend to be aimed at reluctant witness to give evidence. However, it is protecting and compensating the victim more questionable whether they would be effective than punishing the offender. Finally, the witnesses in such cases, and this approach is Prorection from Harassment Act 1997 introduces difficult to reconcile with the idea that the law- new measures against harassment that blur the should empower victims rather than add to their boundaries between civil and criminal measures. vulnerability. Consequently it seems rhat there is some reluctance to use this power. However there The measures is very little research on the role of the CPS in prosecuting domestic violence. As the Victim At present there are three main forms of civil Support Working Party (1992: 2.44) has measures (Barron 1990: 14): suggested more research is needed. • undertakings by the alleged offender not to The Victim Support Working Party further assault his/her partner and/or to recommended where witnesses are compelled to leave the shared home by a particular date; give evidence support (such as housing) should be provided to protect them (Victim Support 1992: • protection (non-molestation) orders aimed 2.31). in addition, CPS respondents in Graces at preventing further violence or harassment (1995: 46) study said it was now standard to in an emergency; and require formal written retraction before dropping a case of domestic violence. In the long-run it • exclusion (occupation or ouster) orders to may be necessary to reduce the emphasis on [he remove the offender from the house or to require him (or her) to keep away, usually for a specified period.

'The literature revieiv also failed to find any research on the Undertakings can be made at county courts and use of bail in sexual offences and hate crimes against ethnic and arc signed in court. The advantage for the accused sexual minorities.

182 i.s that no evidence i.s heard, no admission of guilt Deterrents to legal action is required and the application for an injunction i.s usually withdrawn or .adjourned. This may allow Barron interviewed female victims of domestic faster action for the abused partner, bur no powers violence, solicitors dealing with these cases and of arrest can be attached. In principle representatives of the police, probation service, undertakings have the same force as court orders, women's refuges. Victim Support and the but. Barren argues that this does not occur in magistracy in Bristol and Cumbria. Key findings practice (Barron 1990:22). included that solicitors and the courts often expressed the view that women take oui Protection and exclusion orders are available at injunctions for trivial reasons and have not tried both the count}' court and magistrates' courts, hard enough to make che relationship work. This whether or not the partners arc (or have been) is obviously at odds with evidence that victims married. Until recently, at the magistrates' courts have often experienced numerous acts of violence applicants had to be married to the partner they over a long period before they take legal action. are applying for an order against. Other Many solicitors were reportedly reluctant to begin differences included that magistrates' courts could proceedings without also starting divorce not order a partner to keep away from an area proceedings - despite the intention that protection around the home. I his has been changed under orders should be an emergency measure. This the Family Law Act 1996, so both courts have the further ensures that those victims who do proceed same levels of powers (Lord Chancellor's with legal action tend to be those who have Department 1997: 57-58). decided that the relationship is over (Barron 1990:31-34). Powers of arrest can be attached to protection and exclusion orders in certain circumstances (eg if the A number of other deterrents to legal action were court is satisfied actual bodily harm was caused also identified. Some women were dissuade by and it is likely the applicant will reoffend), fear of the cost. This was partly attributable to However, the Victim Support Working Parly lack of knowledge of the legal aid system. report on domestic violence (1992: 3-16) suggests However, there was evidence that some women powers of arrest are not usually applied for. Tn who were eligible for legal aid on the basis of their addition, Barron argues that courts are relucram income had problems getting it for other reasons. to attach this power (1990: 54). In 1996, 31% of In some cases, solicitors failed to apply for it. In injunctions made had powers or arrest attached others, applications for legal aid were rejected: (calculated from Lord Chancellor's Department some solicitors questioned suggested a lack of 1997:'fable 5.9). consistency in these decisions. The length of time taken in assessing income was also a factor. Some The Victim Support Working Party women had to wait several days or weeks for recommended that powers of arrest should be the supposedly emergency action. For those who were norm, and cover all clauses of the injunction not eligible for legal aid, the costs of pursuing a rather than just some as at present (1992; 3-18). case could be prohibitive (Barron 1990: 34-35). The change in the Family Law Act 1 997 requiring Other reasons included cultural and language powers of arrest to be attached where there has barriers. For example, one Sikh woman said that been violence or threats of violence, may go some women who used the legal process were way to address this, however, the Act does stigmatised by her culture. Consequently it could include a caveat allowing exceptions where the be difficult to find a family member or friend court is satisfied that the applicant will be willing to translate for a complainant possessing adequately protected without it (Lord limited English (Barron 1990:48). Chancellor's Department 1997: 58). As Con way (1998: 142) concludes: Seeing a solicitor

"'the full potential of the Act will only be Having decided to see a solicitor, Barron reported fulfilled if magistrates make it their business to many women found them intimidating. Although rully acquaint themselves with the true nature there was recognition chat the legal process may and effects of domestic violence and take every be slow, complaints were made that solicitors were opportunity to protect vulnerable victims". slow to start the legal ball rolling. Other problems included lack of explanation about the legal

183 remedies available: although many vicrims had Where orders were made, they were not always heard of injunctions, rhev were nor always aware satisfactory: for example, excessive time (several of the different kinds available. In addition, there weeks) was given to vacate the home. The criteria were discrepancies between the women's for deciding whether to exclude a spouse from the perceptions of the effectiveness of injunctions and marital home include the financial situation of those of the legal profession. Some women had each partner and the needs of any children. No applied for injunctions in rhe past, and single criteria is supposed ro rake precedence. In complained of a lack of awareness or the practice though, Barron suggests there is ineffectiveness of injunctions among the legal reluctance to impose injunctions if a custody profession. One reason for this may be that hearing is planned (1990: 52-53). The Victim injunctions are rarely enforced, and consequently Support Working Party recommended that courts that breaches rarely go to court. This luck of should be made aware of the clangers posed to awareness meant some women were misinformed victims of domestic violence, and must take a about the effectiveness of such measures (1990: tougher approach (1992: 3.8, 3.63). In addition, 36-41,; 89-102,; 1 1 2-114). To counteract this, the according to the report orders are usually only Victim Support Working Parry (1992: 3.40) made for a fixed period, typically three months: recommended that all solicitors wishing to take the complainant then has to reapply for an on this work .should be offered some training, extension. The Working Party recommended although it is not clear who would provide this longer orders or indefinite orders until further training. The report also suggests there is a need notice (1992: 3.10-3-1 I). for more comprehensive emergency out of hours service, which might be met by a rota system Enforcement (1992:3.50). Finally, when orders were made Barron found that Attending court police were unwilling ro enforce them even when they were informed of rhe injunction and powers At court, there were numerous concerns. Lack ot of arrest were attached. Despite the fact that rhe separate waiting areas and luck of special areas in orders were made by a court, they persisted in which complainants cm consult their legal viewing violence as private matter (Barron 1990: advisers were highlighted. Scheduling was also a 16). Barron also found some solicitors were matter of complaint: Barton suggests that the reluctant to pursue action when orders were domestic violence cases were given a low priority, breached (1990:112-114). These findings should so that court dates and times were more likely to be treated cautiously because the samples of be changed. In court, most victims feit that they people interviewed were very small. Nevertheless had a fair hearing, although some were upset their there are some other (albeit small) studies which story was not given enough weight. In some cases, provide further support. For example, Farragher women who had begun new relationships since (in Pahl [Fd] 1985: 110-124) observed twenty-six the break-up of their marriages were blamed for domestic calls to ihe police and found reluctance provoking rhe violence (1990: 42-51). by the police ro intervene even when an injunction had been breached. The Victim In addition, Barron observed char although almost Support Working Parry recommended that arrest all were able to obtain a non-molestation order or for breach should be the norm (1992: 3.IS). undertaking, "anything more than that could be difficult to obtain" (1990: 50). She argued that ii. Criminal measures the courts appeared overly concerned with not removing a man from his home: It was noted above that the Protection from Harassmcnt Act 1997 creates some new measures that blut tiie boundaries between civil and legal "Even in cases where there has been a Ions measures. For example restraining orders, similar history of violence, many lawyers appeared ro to injunctions, will be available in the criminal believe it would be perfectly possible and safe courts to prohibit any further harassment. Where for the woman to return home once she had an injunctions are granted and then breached, the order or an undertaking that her partner would plaintiff will be able to apply for a warrant for the not molest her" (1990: 50). offender's arrest. Breach of an injunction will be a criminal offence, punishable by a maximum of

184 five years imprisonment tried on indictment and non-custodial sentence. It seems plausible that six months tried .summarily. This means it will be lack of knowledge about domestic violence and an arres table offence: in other words the police ideas about [he sanctity of the home continue to will be able to arrest the offender If rhey know have some impact. However, further research is about the injunction and suspect it has been needed both to examine sentencing of domestic breached without requiring the victim to apply for violence zo see if it is really treated so leniently, an arrest warrant. and if this is so, to examine to what extent non- legal factors influence sentencers. Two criminal offences of harassment are also created. The first, where the behaviour is so Implications of sentence for victims threatening the victim fears violence, is punishable by a maximum of five years imprisonment. The A final concern Is the impact of the sentence on second does not require the fear of violence and is the victim. Edwards (1989: 153) describes one punishable by a maximum of six months case where the offender refused to pay the fine imprisonment. In rhe past harassment has been levied: faced with the threat of losing her prosecuted using other criminal laws such as furniture, the victim paid the fine herself. This assault occasioning actual bodily harm, but has suggests that sentencers need to be alert to the been difficult because the law requires proof of possible Impact of sentences on victims: financial intent. The new provisions have a lower standard ties or dependence (such as .shared property, of proof, requiring instead that the conduct must maintenance contributions or child support) may have occurred at least twice and that a reasonable continue even where partners split up. The Victim person would have realised their actions would Support Working Party (1992:2.51-253 citing cause a fear of violence or sense of harassment (for Smith 1989) suggested probation officers' social further discussion see Jason-Lloyd, L 1995: 787- work training, with emphasis on keeping families 790; Wells, C 1997: 463-470).' together, may lead them to send out wrong messages to offenders. It is important to ensure Leniency that safety of women is paramount and the seriousness of offence is not underplayed. This is It has yet to be seen whether these measures will clearly something training might address: ihe address concerns about lenient sentences for Victim Support report also suggests that domestic violence. For example Craig {] 992: 568} Probation Service management should encourage found that of 500 incident reports studied in good practice. 1991, half were no-crimed, half again were prosecuted, and half of those dealt with by a bind e. ...and beyond over. I Jowever, Craig .suggests that this apparent leniency was partly explained by the victims Beyond the trial, there are numerous other issues wishes. She suggests there is anecdotal evidence or concern to victims of domestic violence. Not that the younger often unmarried women are all are of direct concern to the criminal justice more likely to report, support charges and want a system, although they will be very important to custodial sentence. In contrast the older, more the victims of domestic violence. Examples dependent and less socially mobile women with include formally ending the relationship with the children, are more likely to want to maintain ties abusive partner through divorce, access to an with the offender and to resume cohabiting, and adequate income, and negotiating contact with to therefore want less serious sentences. any children (see Victim Support 1992 5.5-5.20 for a discussion of these issues). Two longer-term issues involving the criminal justice system are This presents something of a contradiction briefly considered below: places of safety and though. If sentencing reflects victims wishes and information. younger women are more likely to report and want harsher sentences, this would suggest more Places of safety severe sentences should be imposed than appears to be rhe case. Other possible factors which Craig highlights Include the principle that the offender Securing the victim's safety is a particular concern should only be sentenced for the present crime in domestic violence, when continuing to share a rather than for past record. This has to balanced home with the offender could endanger both the however, with the risk posed to the victim by a victim and the prosecution case. This may also be

185 a concern in sexual offences where the defendant example. Phillips and Brown (1998: xis-xiv) found is or was [he victim's partner or another close Afro-Caribbeans were over-represented among relative living in the same home, and also in some those arrested compared to their representation in racially motivated crimes where the victim's safety the local population. They were more likely than is threatened. However, provision of places of expected to have been arrested following a safety is patchy, and usually met by die voluntary stop/search, and (along with Asians) ro have no sector (for example women's refuges)1". In chapter further action taken against them. This suggests 2 it was explained that accommodation may be the arrests of Afro-Caribbeans may have been provided for intimidated witnesses to assist based on less evidence than that of white suspects. prosecutions against their harassers: if could be However, it might also be related to differences in argued that there is a need for equivalent offence patterns. Phillips and Brown (1998: 29) assistance for victims of domestic violence. suggest Afro-Caribbeans tended to be arrested for some more serious crimes (such as robbery and Information fraud) which might have been more difficult to prove than those white suspects were charged with Despite the emphasis In Home Office Circular (such as public order offences). 60/1990 on keeping victims informed about the offenders' whereabouts, Grace (1995: 24-25) It would therefore, be reasonable to assume that found only 15 police forces had developed some more negative attitudes would be translated into a means of obtaining ail offenders' release dace from greater tendency to under-report, particularly for prisons in their areas. This was usually the racially motivated crime, However, the BCS attending officer or domestic violence officer's suggests ethnic minorities appear more willing to responsibility. This raises a couple of questions: report household offences to the police than firsr, whether this responsibility should be placed whites. The pattern is more complex regarding on the police, and secondly whether formal personal crimes, where racial motivation seems procedures or policy could be used to improve more likely. The BCS suggests only Indians are information. If the responsibility for informing more likely ro report personal crimes than whites, witnesses rests with any agency other than the and that under-reporting is particularly marked Prison .Service, procedures will have to he agreed for Pakistanis. The survey did find that Afro- with them. Caribbeans and Pakistanis were less likely to report racially motivated crimes than other 3. Racially motivated crime crimes, but also found that Indians were more likely to report racially motivated crimes a. Reporting and investigation (Fitzgerald and I laie 1996: 27-35).

Failure to report According to the BCS the most common reasons for not reporting crime generally are that the Xumerous studies have shown that Afro- police could have done nothing, the police would Caribbeans and to a lesser extent Asians cend to not have been interested and that the incident was have more negative attitudes towards the police too trivial. These were particularly salient for than whites. For example, BCS data suggests a racial incidents, bur dislike or fear of1 the police large proportion of Afro-Caribbeans believe the was not a strong concern for any victims police do not treat everyone fairly, and in (Fitzgerald andYlale 1996: 35-36). The latter particular that they do nor treat minorities equally finding is surprising: as Smith (1994: 1090) (Fitzgerald and Hale 1996: 29). observes, it has been well-established that racial prejudice within the police is common, and not This has been explained by ethnic minorities simply confined to the junior ranks (see for wider experiences of policing, particularly example Graef 1990: 1 17-144: Reiner 1991:204- evidence that Afro-Caribbeans are stopped and 210). arrested by the police mote often than whites. Fot However racial prejudice is not the only influence on police behaviour. Although allegations ate made about individual cases of misconduct, Places of safety for victims of racially motivated crimes and male victims appear less common than refuges for female victims research is needed on whether prejudice routinely of domestic and sexual violence. influences police treatment of ethnic minority

186 complainants. Certainly the BCS suggests rhar ''the police perceived another categorv ol minorirv victims are less satisfied wirh police incidents as the real racial incidents, even response than whites, particularly regarding though they were rarely reported (or recorded) racially motivated crime (Fitzgerald and Hale as such. These were inter-racial incidents where 1996:37-38). More specifically, I lesse & the victim and the suspect were from different McCiilchrist, reporting on a council inquiry in ethnic groups...'blacks on Asians and Asians on London argue there are six main complaints about blacks' was seen as the main racial problem". police investigations (in Hesse. Rai, Bennett & McGilchrist Eds 1992:70): Sampson and Phillips 0998: 129-132) observed a catch-22 situation. What the police and other • "The police do not treat racial harassment agencies viewed as minor incidents had a .seriously (ie as a crime). cumulative effect on die victims. Harassment was experienced as a continuous process, and no • Where the police do respond there is no distinction was made by victims between repeated follow-up. harassment and "criminal' incidents. This difference in perceptions (and a lack of language • The police do not take action against interpreting facilities) meant victims were perpetrators. dissatisfied with the police response when they reported. As a result many racial incidents were • The response of the police is not unreported or not reported immediately encouraging. afterwards. However, the police in rum complained there was little or nothing they could • The police do not provide support. do unless the incident was reported immediately. Another disincentive to action suggested by some • The police treat the victims as the problem'". other respondents was the fear of a backlash by the white people on the estate if racial harassment This is given some .support by Sampson and was given a higher profile. Phillips (1998: 129-132) research on racial incidents on an East London estate. They report Further research is necessary to show how typical that the police and other agencies viewed racial these findings are. Other factors which may incidents as non-criminal, and were disinclined to discourage reporting in racially motivated crimes intervene: (or indeed any crimes where ethnic minorities are involved) include cultural barriers, and concern "Despite the presentation or detailed data on that their immigrarion status will be questioned repeat racial victimisation...there was a (see Chigwada-Bailey 1 997: 37 for an example of consistent denial or undermining of the extern a case in which this occurred). This raises a related of the problem..Attempts were made to question of whether the greater risk of challenge victims' accounts of racial incidents, victimisation among ethnic minorities is partly either by suggesting that the incident was not due to offenders playing on such fears, bur no racial, or by shifting the emphasis to blaming research was found on this subject. the victim for not reporting immediately..Sometimes accusations were Improving responses and relations made about families Fabricating incidents so they could be rehoused quickly." Measures to address some of these problems may be relatively simple, for example in ensuring that a Sibbit (1997: 25-27) similarly suggests that the network of interpreters is available to overcome police categorised racial harassment and violence language barriers. Dealing with cultural barriers, reported by victims into as neighbour disputes, for example, in encouraging reporting by Asian "not really racial" incidents, verbal abuse and women, and challenging racism within the police nuisance behaviour. None were seen as worthy of are more difficult. Six areas are discussed below: police attention. Serious attacks where victims pro-arrest policies, investigation, recruitment from perceived racial motivation were forwarded to ethnic minorities, retention, public consultation CID and seen as serious crimes such as attempted and multi-agency cooperation. murder rather than as racial incidents, and recorded them accordingly. Instead: • Pro-arrest policies.

187 As with domestic violence, arrest policies could be low. However more recently Maynard and Read considered. There are two aspects 10 consider: (199": 7, emphasis added) found "racially arrest in racially motivated crime, and arrest las motivated incidents are rather more likely to be well as stop and searches) of ethnic minorities. cleared up than non-raeially motivated incidents, Al chough the Litter is beyond the scope of this but arc rather less likely to result in charge or report, it seems likely chat will contribute to caution". The introduction of ethnic monitoring ethnic minority perceptions or policing. On the in police forces from 1996 should provide much former, Sampson & Phillips' study of racial more information about how suspects and attacks on an Fast London Estate suggested arrest offenders are treated according to race. Fitzgerald in racially motivated crime is low: only about and Sibbit (1997: xiii) suggest monitoring may thirty arrests were made for the almost four improve relations between the police and ethnic hundred racial incidents that came to police minorities generally: both by increasing police attention in the borough in 1990 (1992: 12). awareness of how they treat ethnic minorities and Phillips & Brown (1998: 38) found only 14 by encouraging greater dialogue between them. suspects in their sample of over 4,000 detainees Nevertheless more information is needed about were arrested for incidents officers said were complainants' and victims experiences. perceived as racially motivated. This suggests that pro-arrest policies might be an option for racially * Recruitment motivated crime. If has also been suggested that increased Nevertheless the limitations of this approach in recruitment of police- officers from ethnic domestic violence cases were noted above. In minorities would help improve relations. Yet addition, there may be some problems particular despite a number of national recruitment to racial harassment and violence where the campaigns since the mid 1970s and various identity of the perpetrator may be less obvious. recommendations on improving recruitment in Sibbit (199": 92, 95-96) found little effort was the 1 980s, the number of ethnic minority recruits made to identify suspects in cither of die areas she remains low. In 1996/7 just 2% of new recruits studied. She attributed this in part to lack of appointed in England and Wales were from ethnic information/intelligence and a failure to make minorities, compared to 3-7% in 1994/5. This effective use of the information that did exist: for compares to 6% of the general population example cases were filed by victim rather than (liMIC 1997, Appendices 7 & 8). perpetrator and information on perpetrators was not collated or linked to other reports. A more Research has suggested various reasons for ethnic general factors was officers' sense that racial minorities reluctance to join, including racism harassment was not really a criminal matter. This from the public, abuse from colleagues and in turn appeared to be associated with the perceptions that to do so would be joining the involvement of a large number of children under enemy {see Smith 1994: 1096). 1 Ioldaway (1991 the age of criminal responsibility ". Sibbit cited by Smith 1994: 1096) has suggested that comments "the police appear to be disempowered positive action by individual forces to recruit by the notion the that only action they can take is ethnic minorities has actually been limited, and prosecution". The emphasis on prosecutions in failed to make good use of community or race performance targets may comribute towards this. relations staff. I fowever, he argues the major reason for lack of success in recruiting ethnic • Investigation minorities has been failure by police management to tackle racism within the police. When arrests are made, it has been suggested that police investigations are sometimes less thorough • Retention than when white complainants are involved. The Home Affairs Committee (1989: 13) has One reflection of this is that retention or ethnic suggested the clear-up rate for racial incidents is minority recruits is also a problem. Both the

This is not to suggest that children were the only perpetrators This excludes the Metropolitan Police, where 7.4% of those identified. Most were between 11 and 18, but ages ranged from appointed were from ethnic minorities in 96/7 and 5.4% in 5 to 80 years old (Sibbit 1997: 631 1994/5, compared to an ethnic minority population of 20.2%

188 Home Affairs Commirtee in 1989, and more minorities (both ethnic and sexual) may help recently HMIC have emphasised the importance improve relations with them. However, the of retention. In one force inspected, ilie wastage approach most areas have adopted (Police rate for ethnic minority officers was 10% against Community Consultative Groups or PCCGs, less Than 3% for whites. Reasons suggested for which have open meetings at least quarterly) has poor retention included feeling unsupported by tended to be poor at reaching these groups. There management, demonstrated by lack of are alternative ways of trying to reach these groups intervention following racist behaviour or though. Examples include special consultative language (1997: 272-2.75). "Police officers groups for minorities, or holding normal PCCG ...cannot be expected to behave in ways which will meetings in different locations such as temples or enhance relations with minority populations in mosques (Elliott & Nicholls 1996: 9-1 1 & 42- rhe public at large if they fail to treat colleagues 55). HMIC (1997: 38) has recommended the from the same groups equitably"' (1997: 3.73). creation of procedures to ensure all decision- making routinely considers the implications for HMIC has made a number of recommendations race and community relations: improved public to assist police forces in over-coming these consultation could perhaps be a part of this problems, including (199™: 4.1-4.20): process.

• monitoring recruitment, retention and • Multi-agency cooperation. career development of start from minority backgrounds; Finally, multi-agency cooperation may assist in a number or ways such as encouraging referrals, • considering sensitivity to race {Mid increasing support to victims and improving community) relations in recruitment, responses to prevent further victimisation. Sibbit promotion, srafi appraisal and staff (1997: 98) found schools could have a greater role deployment; to play in preventing and addressing harassment. I lousing authorities also have a particular interest • inclusion or race and community relations in tackling racial harassment. Love and Kirby in training, focusing on dealing real-lire (1994: 17-20, 25-28) found they used various situations; and approaches, included inserting clauses in tenancy agreements forbidding racial harassment and • ensuring that policies and practices make violence (reported by 61% of respondents) and clear that any expression of racial (or other) interviewing and warning known perpetrators prejudice is completely unacceptable. (80%). Some also reported initiating repossession proceedings and arranging a priority transfer for If racism is as deeply embedded within the police victims !54% and 75%). Just over half (55%) as suggested, it seems questionable whether were involved in multi-agency groups, and almost increasing ethnic minority representation in the all of these involved the police. police will be sufficient to change police culture and relations with members of the public. The main role reported of these multi-agency 'I adding racism needs to be a central parr of Any groups was coordination, with less than half efforts to improve relations wirh ethnic involved in providing support to victims' or minorities. prevention. A number of benefits were recorded: for example almost all reported increased • Improved public consultation. understanding between agencies. Two-fifths thought cases were being dealt with more Other possible measures to improve relations effectively and more than a third though that between the police and ethnic minorities include more cases were coming to light as a result of the improved public consultation. The importance of group. However, problems such as lack of consulting the public generally but particularly cooperation of some agencies, lack of resources with ethnic minorities was recognised In die Scarman report which followed inner-city riots in the early 1980s. This led to the creation of a statutory duty on police authorities to consult the 'There are a number of possible objections to special consultation fora for minorities. Both these and the counter- public on local policing. Consultation with arguments are explained in Elliott & Nicholls (1997: Table 8.1).

189 and agreeing the roles of c.ich agency were imprisonment - under rhc new provisions this reported (40%). A fifth also reported problems in would rise to seven years where there was some agreeing a definition of racia! harassment. racial motive. Under the proposals the standard of proof of racial motivation would be lower in these Sibbii (1997: 100} also looked at multi-agency cases than for the new offences of racial groups in the two areas slid studied. A number of harassment and racial violence. problems were observed. In the first group, rhe.se included a tendency for discussions to focus on • Recruiting and retaining ethnic minorities, the truthfulness of allegations, failure to identify especially among the judiciary and the perpetrators and a lack of police action. In the magistracy second, victim's dissatisfaction with the police and the heavy representation of the police contributed As with the police, there is concern that under- to a strained atmosphere: "it was felt, certainly representation of ethnic minorities continues where the police were present, members of the elsewhere in the justice system. For example, there public would feel reluctant to attend, let alone are no ethnic minority Lords of Appeal. Lords voice their views''. Overcoming such obstacles justices of Appeal, I ligh Court judges or Deputy may be a slow and lengthy process. High Court judges. Hthnic minorities are under- represented at every level of the judiciary and b. The decision to prosecute and the trial magistracy, accounting for less than 1% of circuit judges, less than 2% of recorders, district judges, Even less literature was found on the experiences stipendiary magistrates, and less than 4% of of racial harassment (and other ethnic minority) acting stipendiary magistrates for example (The victims of the subsequent stages in rhc criminal Guardian 25/2/98). Interestingly Home Office justice system. Nevertheless rwo possible areas for figures (1997b: 31-33) suggest that under- action can be identified. represcn ration is not a problem affecting solicitors (6%), barristers (8%), the CPS (8%) or\he • Highlighting racial motivation as an probation service (8%). However, this docs not aggravating facror mean thai these groups have no problems. Concerns about career prospects and about bias in At present, guidance issued to the courts pre-sentence reports written by die probation recommends that racial motivation should be service (see Chigwada-Bailcy 1997: 49-61) are treated as an aggravating factor (1 lome Office two cases in point"'. 1997c: 7-9). However Ruddick {1993: 5) suggests that prosecutors do nor always highlight racial motivation as an aggravating factor. This is The dominance of whites in the judiciary might supported by CPS figures which suggest that this be one reason for the apparent failure to routinely is so in about 1 5%. The CPS report also suggests treat racial motivation us an aggravating factor. that racial motivation does not always result in Whether or not this the case, the recruiting more longer sentences upon conviction. The court people trom ethnic minorities to the bar and stated that sentence had been increased as a result appointing more judges from ethnic minorities in only a fifth of those cases where prosecutors might help reassure both witnesses and defendants drew attention to racial motivation. It seems from ethnic minorities about the fairness of the plausible that in some cases the sentence was courts. At the same time, though, it has to be increased but the court omitted to state this fact acknowledged that some people from ethnic or it was not recorded. Nevertheless this suggests minorities may themselves resent suggestions of racial motivation is not always seen as an positive discrimination. The experience of the aggravating factor in sentencing. police also suggests that retaining people from ethnic minorities is just as important: training could be one element of this (see for example Government plans (Home Office 1997c: 7-9) to require the courts to view evidence of racial motivation as an aggravating factor may address this. Up to an extra two years could be imposed tor some offences where there is evidence of racial :Sibbit (1997. 43) says that probation officers have a motivation. For example, malicious wounding responsibility to challenge racial motivation, but found that this usually carries a maximum penalty of five years opportunity was rarely grasped in either of the two areas she studied.

190 Chigwada-Bailcy 1997: 63. Maynard and Read • denial of anonymity (suggested 1997:25-27). sensationalist media reporting sometimes encouraged by judges expressions of 4. Hate crimes against sexual minorities disgust);

Again, very little literature was found on sexual • interpretation of the law - for example, minorities experiences of the criminal justice extending the idea implicit in the Sexual system despite anecdotal evidence or Offences Act that homosexuality is discriminacory treatment. Certainly, forma! unlawful; policies on treating sexual minorities equally appear much less common than those for other groups such a_s ethnic minorities. As noted above, • differential .sentencing: Galloway argues that .some behaviour by sexual minorities is prosecution rates, convictions and sentences criminalised and hostility cowards sexual are ail higher for homosexuals rhan for minorities has been dubbed "the last acceptable equivalent offences by heterosexuals. prejudice". Galloway (1983: 102-124) suggests According to Galloway, under half of those seduction theory, that young men ate seduced by convicted of sexual intercourse with a girl of older homosexuals, encourages homosexuality 12, but more than 90% of men convicted itself IO be seen as corrupt. Galloway argues that for sex with 1 5 year old boys are this contributes cowards discrimination in the imprisoned. courts, in six further areas: • civil law: for example, Galloway argues that adopting, getting child custody or having • leniency towards queer-bashers (which may access to their children are ail more difficult sometimes be linked to the gay panic for homosexuals. defence, see box); Galloway recommends three main approaches to The gay panic defence (Toolis 1997: 36- combat discrimination;

1. Educating, Toolis describes the "gay panic" defence, whereby people accused of (often very brutal) Galloway argues that both Initial training and killings claim that they were victims of unwanted sexual advance. He gives several refresher courses should be used to combat examples of how the defence has been used prejudice, particularly for operational officers in successfully, and suggests that some lawyers have Vice Squads. Refresher training may be tried to establish "homosexual panic" as a particularly important: Galloway's observes thai medical condition, although the medical studies on race relations training have suggested profession had not previously identified it as racial prejudice initially falls but soon returns to a such. similar level. More specifically Mason and Palmer (1996: 95-99) recommend training for crown prosecutors and judges to tackle homophobia and Mason and Palmer (1996: 95-7) report than encourage them to challenge gay panic defences. Colin Richardson, from the publication Gay They argue homophobia should be treated as an limes, has collected information on gay murders aggravating factor nor a mitigating one. At a from newspapers and information passed to broader level, they recommend that the him. These records suggest that the homosexual Department for Education and Employment panic defence was used in 15 of the 137 cases should issue guidance on the needs of gay, lesbian recorded between 1986 and 1996. However the and bisexual pupils for example on homophobic outcome of these cases is not reported, and it is bullying in schools and colleges^". questionable whether this sample is representative.

• discounting homosexual testimony (eg in agent provocateur cases, where the prosecution evidence is based on police They also argue that s28 of the Local Government Act, which testimony); prohibits local authorities from intentionally promoting homosexuality be repealed.

191 2. Monitoring and protesting However, this is nor common practice and as with recruiting ethnic minority officers, there may be Galloway suggests monitoring poliee action, problems with retention. There has been some reporting abuses of police power and lodging research on the experiences of sexual minorities formal complaints will be more effective than who join the police. For example, Burke (1994a, education in the short-term. Although some 219-227) interviewed approaching fortv police groups already do this, Galloway .suggests this officers from sexual minorities. "The sample was could be extended further. quite small and self-selected, so it the representativeness of Burkes findings is open to 3. Legislating to end legal discrimination and question. Nevertheless, they are interesting: change policing. • None of those interviewed had mentioned Recent proposals to change the homosexual age of their sexual orientation during the selection consent were noted above. Mason and Palmer process even those the vast majority knew (1996: 99) argue the offence of gross indecency they were gay, lesbian or bisexual when they should be reviewed and possibly replaced with a joined; public .sex offence tor both heterosexuals and homosexuals. Thev also suggest consideration • Most (approaching nine out of ten) believed should be given to creating a test of they would not have been selected if they proportionality requiring a reasonable relationship had done so; between the level of provocation and retaliation-8, and requiring judges to direct juries to consider • About two-thirds believed that the police this. In addition to legislation. Galloway were either slightly or much more recommends community policing, in which homophobic than the rest of society; greater contact is made with people being policed, and consultation with local gay groups (see • Approaching four-fifth:, believed an openly above). homosexual officer would not have the same career prospects as a heterosexual officer; Requiring police forces to follow community policing through legislation seems problematic. At • A quarter felt they had been discriminated present the style of local policing is seen as an against in some way. operational matter decided by chief constables, and community policing is just one of a number One step towards improving police attitudes of competing approaches (such as problem towards sexual minorities might then be to start oriented policing and zero tolerance policing). with equal opportunities policies within the police. Burke (1994a: 227) found less than half of Nevertheless, greater consultation and contacts all forces in England and Wales had extended between the police and sexual minorities could their equal opportunities policies to cover sexual perhaps be encouraged either through guidance, minorities, although the number may have or through the national policing objectives set increased since then. annually by the Home Secretary (on the latter see Mason and Palmer 1996: 99). There is some Establishing gay liaison officers or specialist units evidence of efforts to improve consultation, such is another approach. For example in one division as lesbian and gay consultative groups (see Elliott in Northamptonshire, a specialist unit has been &.'. Nicholls 1997: 42-47), national conferences on set up to cover homphobic incidents as well as policing lesbian and gay communities and the domestic violence and racial incidents (HMIC creation of a Lesbian and Gay Police Association. 1997: 3-25). Although the research on domestic violence suggests there may be problems with Some forces have also attempted to attract recruits specialist units, this development does suggest from sexual minorities by placing advertisements that more importance is beginning to be attached in the gay press (The Guardian 26/1 1/97). to protecting sexual minorities from homophobic attacks.

According to Mason and Palmer there was a precedent for this Finally, the creation of force policies on policing before the 1957 Homicide Act. homophobic incidents (recommended by HMIC

192 1997: 4.19) and disseminating good practice recent years crime prevention hjt. begun to focus could also he considered. The endorsement of a on repeat victimisation. The risk of victimisation national charter for good practice and tends to be highest after an initial offence, so consuluuinn to rid the police oi anti-gay prejudice intervention immediately after the initial offence by the Association or Chief Police Officers (The may help reduce the risk of repeat victimisation Guardian 26/1 1/97) may assist rhis process: (see ror example I'arrdl and Pease 1993). A couple Mason and Palmer (1996: 99) recommend that all of points are worth noting about the impact of forces adopt the charter. Similarly, Mason and this finding to dare. Palmer (1996: 100) argue rhat local authorities and housing associations should develop policies First, although all parts of the criminal justice on hate crimes against sexual minorities, covering system have a role to play in repeat crime tenancy agreements, injunctions and prevention, the impact has been felt most by the repossessions. police. Tackling repeat victimisation has been made a priority for the police nationally through the national objectives set annually by the I lome Conclusion Secretary. Systems arc being developed to help the police identify repeat victimisation and respond Since the 1970s the rise of feminist criminology more appropriately. In some areas for example, has made a major contribution in highlighting attending officers will be provided with details problems in criminal justice responses to sexual about previous calls from the same address. offences and domestic violence. Efforts have been Historically police command and control systems made to improve responses to both offences as a have not been designed to include such features. result, with some effect. For example, police so progress varies. Since the pohce are usually the treatment of rape is widely acknowledged to have first point of contact with the criminal justice improved, and has been attributed with system and the research indicates quick contributing to the rise in the number of rapes intervention is needed, their contribution is reported to the police. Despite this, the literature particularly important. However, repeat victims reviewed suggests there is still evidence of room experiences and contacts with the rest of the for improvement. criminal justice .system have generally been neglected. Topics for further consideration include Victims of hate crimes against eihnic and sexual the recognition of repeat victimisation and minorities have not received as much attention, whether responses should vary accordingly. although anecdotal evidence suggests criminal justice responses could be improved. Research is Secondly, most repeat crime prevention needed both to measure the extent of these forms programmes have focused on burglary, with the of victimisation and to chart the experiences of aim of reducing opportunities ("situational" crime these groups when they come into contact with prevention as opposed to ''social" crime the criminal justice system. Although complaints prevention, which seeks to reduce propensities to have led to some efforts to improve responses commit crime through education for example). A (eg.widening the definition of racial incidents, Jew have targeted other crimes such as domestic and ethnic and sexual minority recruitment violence and racial incidents. For example, one of drives), concerns continue. the measures used in domestic violence has been the loan of attack alarms, to notify police if help is A number of possible measures to improve the needed and enable a rapid response (Lloyd, Farrell position of all four groups have been highlighted & Pease 1994). It seems strange that programmes by the literature review (summarised in Table 4). have not been more common for victims of those In choosing between them their relative costs will offences most commonly associated with repeat need to be considered. It should also be borne in victimisation, such as domestic violence. Targeting mind that to be confident that any new measures such vulnerable groups might help maximise the have had the desired results some evaluation of value of such programmes. The implications of their effectiveness will be needed. repeat victimisation tor the criminal justice system may have yet to be fully realised. However, rather than simply responding to crime the longer-term aim must be to prevent crime. In

193

Section 5: Conclusion

The definition of vulnerable and intimidated 4. These included domestic violence and racially witnesses adopted will be important in motivated crime. Hare crimes against sexual determining whether those in most need or .special minorities could also be included; however, due to assistance and support receive it. The literature the lack of literature on this subject they were not review focused on three main groups: intimidated considered in detail. witnesses, those with mental and physical disabilities and illnesses, and victims of special Numerous possible measures to improve the offences. These groups were chosen because there situation of vulnerable witnesses were identified seemed to be some consensus that these three within the report. These require varying levels of s;roups should be considered vulnerable. Some intervention: from simply providing witnesses others who might also he seen as vulnerable were with leafless providing useful information to excluded. For example, child witnesses were not changing the law and redefining the examined because covering this sizeable literature responsibilities of the various criminal justice would have been impractical given the time agencies. available. The exclusion oi some other groups (for example, the elderly) was due to the lack of If is also apparent that the different groups of literature found on their experiences as witnesses. vulnerable witnesses discussed have varying needs. Of course, the Review may identify different Therefore some measures are specific to particular groups of vulnerable witnesses to those considered groups, such as the provision of female doctors for within this report. rape victims. Despite this there are also some areas of common ground, so some measures may be of Witness intimidation, examined in section 2, value for more than one category of vulnerable undermines both public confidence in the witnesses. For example, pre-trial preparation has criminal justice system and its effectiveness. To been raised for all three groups examined. Table 5 recognise how serious it is, witness intimidation summarises die measures derailed and the groups became a criminal offence in its own right in who might benefit from them. 1994. However, very little material was found on how witness intimidation is being tackled, both The cost implications of the various measures here and abroad. One reason may be the need for discussed will vary greatly. Some measures will be security (that is, ro stay one step ahead of [he relatively cost neutral, such as clearing the court's offenders). Another reason may be general neglect public gallery ro prevent witness intimidation. In of the issue of witness intimidation, which is now addition, those cases where measures can be used beginning to be redressed. for more than one group may be more cost- effective. However, the costs of special measures There is some evidence that witnesses with may increase over time. It the measures are disabilities and illnesses, examined in section 3, effective, it seems plausible that repotting rates may be more vulnerable to crime than other will increase. This may increase the total costs of witnesses, for example because of greater the special measures, as more witnesses take dependence on others. 1 hey may also be advantage of them, it would also increase the vulnerable in relation to the criminal justice workload of the various criminal justice agencies system. For example, people with disabilities or concerned. Nevertheless there may be economies illnesses may find acting as a witness particularly of scale at some point. upsetting or difficult. A number of other considerations can be Some of the definitions of vulnerable witnesses identified concerning the use of special measures discussed in section 1 mention special offences, but for vulnerable witnesses. Many of these are largely only sex offences have been specified. Other practical concerns: possible special offences were examined in section

195 • how easy it is io determine vulnerability • whether use of the measures recommended using the definition; will be monitored or evaluated.

• whether all \vi en esses or just some (eg Other considerations concern the implications of victims) should count as vulnerable; particular measures for justice. For example, granting witness anonymity has been considered • who should decide whether the witness for all three groups examined by the report. 1 he meets the definition or vulnerability; argumenrs in favour of this approach are that it would reduce trauma for the witness and help • whether and how rhe witness's views should prevent intimidation. However, this raises issues be taken into account in defining about the defendant's ability to defend vulnerability; him/herself, and whether juries might draw adverse inferences. • whether particular measures .should be granted as a right, whether there should be Similar concerns are raised by the idea that the an assumption in favour of their provision, defendant's right to cross-examine witnesses or whether they should be provided personally (in the absence of legal representation) completely ac the discretion of rhe various should be curtailed in certain cases (eg. in rape agencies; cases).

• who should have responsibility for providing Finally, it has been observed that .some of the each of the measures. I his might be given to measures discussed may improve a witnesses specialist officers or units in each part of the performance in court, I his raises the issue of criminal justice system, building on the whether witnesses should have a right of appeal specialist units that already exist in police against decisions about whether they are forces for example ; vulnerable, or concerning the provision of particular measures. • whether one agency should be given the task of coordinating the various measures; and

Kent County Constabulary already has a Vulnerable Victim Coordinator (Law Society Mental Health and Disability Sub- Committee submission to the Review, 11/9/97).

196

2oo bibliography

Adler, Z (1987) Rape on Trial London: Rourledge revised PACK codes of practice Home Office and Kegan Paul Research Study 174. London: Home Office

Adler, Z. (1992) "Male victims of Sexual assault - Burke, M (1994) "Prejudice and Discrimination: legal issues" in G.C. Mczey & M.B.King (Eds) The Case of the Gay Police Officer" in British Male Victims of Sexual Assault Oxford: Oxrord journal of Criminology 34, 2: 192-203 University Press Burke, M (1994b) "'Homosexuality as Deviance: Advisory Group on Video Recorded Evidence The Case of the (Jay Police Officer"' in The Police (1989) Report of the Advisory Group on Video- Journal, July: 219-229 Recorded Evidence London: i iome Office Burke, RA (1996) A Matter of justice: Lesbians American Bar Association 0 981) "How To Do It" and Gay Men in Law Enforcement- London/New Suggestions for Implementing the ABA York: Routledge. Victim/Witness Intimidation Recommendations Washington: ABA Burquest, R., Farrell, G & Pease, K (1992) "Lessons from Schools" in Policing, 8, 148-1 55 American Bar Association (1981) Reducing Victim/Witness Intimidation: A Package Buzawat, ES & Buzawa. CG (1996) Domestic Washington: ABA Violence: The Criminal justice Response California, USA: Sage Barron.J (J 990) Not Worth [he Pjper..? The Effectiveness of Legal Protection for Women and Casey,C (1993) "Twist and Shout" in Police Children Experiencing Domestic Violence Bristol: Review 28 May 28-29 Women's Aid Federation Ltd Cervi, B (1992) "The time has come: vulnerable1 Brereton, D (1997) "How Different are Rape adults disclosing sexual abuse are likely to be Trials? A Comparison of the Cross-Examination rejected by the justice system" in Community Of Complainants in Rape and Assault Trials'1 in Care, 25 June, 921: 14-15 Brirish journal of Criminology 37. 2: 242-261 Chandler, j & Lait.D (1996) An analysis of the Brown, B., Burman.M & Jamieson. L (1992) treatment of children as witnesses in the Crown Sexual History and Sexual Character Evidence in Court London: Victim Support -Scottish Sexual Offence Trials Central Research Unit Papers. Edinburgh: Scottish Office Choudry, S (1996) Pakistani Women's Experience of Domestic Violence in Great Britain Home Brown, D (1995) RA.C.E. Ten years on: A Office Research and Statistics Directorate Review of the Literature Home Office Research Research Findings Xo 43 London: I lome Office Study No 155 London: Home Office Research and Statistics Directorate Chigwad-Bailey, R 91997) Black Women's Experiences of C'nminal Justice: A Discourse on Brown, L., Christie;R &c Morris, D (1990) Disadvantage Winchester: Waterside Press Families of Murder Victims Project: Final Report London: Victim Support Clarke. M (1994a) "GMP team providing 'lifeline' for witnesses'' in Police Review, 1 July, Bucke, T & Brown, D (1997) In police custody: 12-13. police powers and suspects' rights under the

201 Clarke, M (1994b) "Witnesses in fear' in Police Crime Suivey Unpublished report for the Home- Review 6 May. 16-17 Office Research and Statistics Directorate.

Cohen, P (1994) "Bearing Witness" in Elliott. R & Nicholls J (1996) Its Good to Talk: Community Care, 30 April, 20-21 Lessons in Public Consultation and Feedback Police Research Series Paper 22 London: Home Con way, H (1998) '•Protecting Victims of Office Domestic Violence - Parr IV Family Law Act 1996" in Justice of the Peace. 162,21 Fcb, 139- Farrell.G., Jones, G & Pease,K (1993) "Witness 142 and Juror Protecrion: Using Available Technology" in The Magistrate, September. 131-132. Craig. Y (1992) '"Domestic Violence and Magistrates' Courts" in Justice of the Peace, Spet Farrell, G & Pease, K (1993) Once Bitten Iwice 5, 568-570 Bitten: Repeat Victimisation and Its Implications for Crime Prevention Police Research Group Crawford, A., Jones, "{'., WoodhoiLse, T & Young, Crime Prevention Series Paper 46 London: Home J f 1990) Second Islington Crime Survey Office Middlesex: Middlesex Polytechnic Centre for Criminology Farrell, G, Phillips, C & Pease, K (1995) "Like "Faking Candy: Why docs Repeat Victimization Crown Prosecution Service (1995) CPS Poiicv for Occur?" in British Journal of Criminology, 35, 3: Prosecuting Cases of Domestic Violence London: 384-399. CPS Fennell, P (1 993) "Vulnerable Witnesses" in Law Davies, G & Westcott, H (1992) Gazette 90/21. 2 June, 21-24 ''Videotcchnology and the Child Witness" in (Eds) Dent, Ii & Flin, R Children as Witnesses Fitzgerald.M & Hale, C (1996} Ethnic Chicbester; John Wiley and Som Minorities: victimisation and racial harassment. Findings from the 1988 and 1992 British Crime Davis, G (1997) The Admissability and Surveys Home Office Research Stud}' 154. Sufficiency of Evidence in Child Abuse London: Home Office Research and Statistics Prosecutions: Interim Report Unpublished Report Directorate for the ! lome Office Fitzgerald, M & Sibbii, R (1997) Ethnic Deloitte & Touche (1996] Review oi'ihe Use of monitoring in police forces: a beginning. Home Medical Specialists by Police Forces London: Office Research Study 173. London; Home i fome Office Police Research Group Office Research and Statistics Directorate.

Dixon, M (1995) "Special Witnesses: What's So Freeman, M (1997) "Violence Against Women: Special About Them?" in Brief (Journal of the Does the Legal System Provide Solutions or itself West Australian Law Society) 22, 5: 5-8 Constitute the Problem?' in British Journal of Law and Society 7, 215-263 Dobash, RP & Dobash, RF. (1979) Violence Against Wives New York: Free Press Calvin, J. & Polk, K (1983) "Attrition in Case Processing: Is Rape Unique?" in Journal of Dobash, RP, Dobash, RE, Wilson. M & Daly. M Research in Crime and Delinquency, 1, 126-154 (1992) 'The Myth of Sexual Symmetry in Marital Violence" in Social Problems 39, 1,71-91 Grace, S (1995) Policing Domestic Violence in the 1990s Home Office Research Study 139 Dobash, RP (1996) Research evaluation of London: Home Office Research and Planning programmes for violent men Central Research Unit Unit Edinburgh: Scottish Office Grace, S., Lloyd, C & Smith ,L (1992) Rape: from Dowds. L & Budd, T (1997) Victim and witness recording to conviction Research and Planning intimidation: findings from the 1994 British Unit Paper 71 London: Home Office

202 Graef, R (1990) Talking Blues: The Police in their Home Office (1 996d) The Victim's Charter: A Own Words London: Collins Harvill Statement of Service Standards for Victims of Crime London: Home Office Graham. Mil (1985) Witness Intimidation: The Law's Response Connecticut, USA: Quorum Home Office (1 997) Criminal Statistics England and Wales 1996 London: Home Office Harris,J. (1997) The Processing or Rape Crises by rhe Criminal Justice System: Interim Report. Home Office (19Th) Race and the Criminal Unpublished. London: Home Office. Justice System - A Home Office Publication Under Section 95 or the Criminal Justice Act Hague, G., Malos, F & Dear, W {1996) Multi- 1991 London: I Iome Office Communications Agency Work and Domestic Violence: A National Directorate Study of Inter-Agency Initiatives Bristol: Polity Press Home Office (1997c) Racial Violence and Harrassment: A Consultation Document London: Healey, KM (1995) Victim and Witness Home Office Intimidacion: New Developments and Emerging Responses US Department of Justice: National Iller. M (1992) "A fair trial for rape victims" in Institute of Justice Law Society's Gazette, 89, 9:2

Herck, GM & Berrill, KT [Eds] (1992) Hare Jackson, H & Borthwick, S (1993) "Witness Crimes: Confronting Violence Against Lesbians Protection"' in Community ('are, 9 Sepr: 20-21 and Gay Men California: Sage Jason-Lloyd, L (1995) "The Protection from HMIC (1997) Winning the Race: Policing Plural Harassment Act 1997 - A Commentary'' in Justice Communities. IIMIC Thematic Inspection of the Peace and Local Governement Law, Aug Report on Police and Commnity Race Relations 16,787-790 1996/97 London: Home Office Jenkins, C (1997) "Women encouraged to report HMSO (1 994) Criminal justice and Public Order rapes at hospital" in Police Review, 5/3/97. Act 1994: Chapter 33 London: HMSO Law Commission (1995) Evidence in Criminal Home Office (198™) Criminal Statistics England Proceedings: 1 Ieai-say and Related Topics Law- and Wales 1986 London: Home Office Commission Consultation Paper No 138 London: HMSO Home Office (1989) Home Affairs Committee First Report: Racial Attacks and Harassment Law Commission (1997) Evidence in Criminal London: HMSO Proceedings: Hearsay and Related Topics Law Commission Report \o 245 London: The Home Office (1990) The Government Reply to Stationery Office the First Report from che Home Affairs Committee Session 1989-90 Racial Attacks and Law Reform Commission of Western AuSiralia Harassment London: HMSO (1990) Discussion Paper on Evidence of Children and Other Vulnerable Witnesses Perth: Law Home Office (1996b) Domestic Violence Reform Commision of Western Australia Factsheet Home Office Research and Statistics Directorate Issue 4 November London: Home Law Reform Commission of Western Australia Office (1991) Report on Evidence of Children and Other Vulnerable Witnesses Project No 87 Perth: Home Office (1996) Statement of National Law Reform Commision of Western Australia Standards on Witness Care in the Criminal justice System London: Home Office Lees, S (1996) Carnal Knowledge: Rape on Trial London: Penguin

203 Levin. J & McDevirr.J (1993) Hare Crimes: The Crime Detection and Prevention Series Paper 55. Rising Tide of Bigotry and Bloodshed New York: London: Home Office Plenum Maynard, W (1995) "Witness intimidation: LipovskyJA (1994) ''The Impact of Court on strategies tor prevention' in Criminal Justice Children: Research Findings and Practical Matters No 22, Winter 1995/6, 17. Recommendations'' in journal of Interpersonal Violence, 9,2; 238-2)7 Maynard, W & Read, T (1997) Policing Racially Morivared Incidents Police Research Group Littlechiid, B (1993) '"Vulnerable Groups Under Crime Detection and Prevention Series Paper 84. P.A.C.E. 1984 - Social Workers as Unwitting London: Home Office Oppressors?" in AJJUST Magazine May, 1.5-16 McDonald. E (1994) "Her Sexuality as Indicative Lloyd, S., Farrcll.G & PeascK (1994) Preventing of his Innocence: Tlie Operation of New Repeated Domestic Violence: A Demonstration Zealand's "Rape Shield" Provision"' in Criminal Project on Merseyside Police Research Group Law journal 18: 321-336 Crime Prevention Series Paper 49. London: Home Office Meldrurn, J (1 980) Attacks on Gay People: A Report of the Commission on Discrimination of London Rape Crisis Centre (1984) Sexual the Campaign for Homosexual Kquaiiry London: Violence: The Reality tor Women London: The CHE Women's Press Mencap (1997) Barriers to Justice: A Mencap Lord Chancellor's Department (1997) Judicial study into how the Criminal Justice System treats Statistics: Annual Report 1996 London: people with learning disabilities London: Royal I LVISO/Governmern Statistical Service Society for Mentally I handicapped Children and Adults Love, A., & Kirby.K (1994) Racial Incidents in Council Housing: The Local Authority Response Miller, GR & Fontes, NT. (1979) Videotape on Department or the Environment. London: Trial: A View from the jury Box California: Sage UMSO Mirrlees-Black, C, Mayhew, P & Percy, A (1996) Mason, A & Palmer, A (1996) Queerhashing: A The 1996 British ('rime Survey: England and national survey ofhate crimes against lesbians and Wajes Statistical Bulletin Issue 19/96. London: gay men London: Stonewall Home Office Research and Statistics Directorate.

Mayhevv, P., Ayc-Maung.N & Mirrlecs-Black Morley, R & Mu[lender, A (1994) Preventing (1993) The 1992 British Crime Survey Home Domestic Violence to Women Police Research Office Research Study 132 London: } lome Office Group Crime Prevention Series Paper 48 London: Research and Planning Unit Home Office

Mayhew, P & van Dijk, J (1997) Criminal Murray, K (1997) Preparing Child Witnesses for Victimisation In Eleven Industrialised Countries: Court: A Review of Literature and Research Key findings from the 1996 International Crime Edinburgh: Scottish Office Home Department Victims- Survey Netherlands: Ministry of Justice Central Research Unit. WDOC (Copies also available from London: Home Office) Myers, j & Paxson, M (1992) "Admissibility of Rape Trauma Syndrome in Legal Proceedings'" in Mayhew, P & White, P (1997) The International Violence Update. 2,7: 3-7 Crime Victimisation Survey Research Findings No.57 Research and Statistics Directorate National Schizophrenia Fellowship (1996) 'The London: Home Office Appropriate Adult' and the mentally vulnerable suspect NSE: London Maynard, W (1994) Witness Intimidation: Strategies for Prevention Police Research Group

204