Part Two: Overseas visits
Our Committee's overseas visits were as follows:
Visits in Canada
Correctional Service of Canada Headquarters, Ottawa (Visited 26 November 1999)
Kingston Penitentiary (Visited 25 November 1999)
Queen's University, Kingston (Visited 25 November 1999)
Simon Fraser University, Vancouver (Visited 24 November 1999)
Conference: 'Risk Assessment and Risk Management: Implications for the Prevention of Violence' (17-19 November 1999, Vancouver)
Visits in The Netherlands
Henri Van Der Hoeven Clinic, Utrecht (Visited 2 December 1999)
Ministry of Justice, The Hague (Visited 3 December 1999)
Pieter Baan Centre, Utrecht (Visited 1 December 1999)
Veldzicht Clinic, Ommen (Visited 30 November 1999)
Visits in the USA
Sexually Violent Predator Programmes:
Arizona Community Protection and Treatment Centre, Phoenix, Arizona
(Visited 12 November 1999)
Mendota Mental Health Institue, Madison, Wisconsin (Visited 10 November 1999)
Oshkosh Correctional Facility, Madison, Wisonsin (Visited 11 November 1999)
Special Commitment Centre, McNeil Island, Washington (Visited 23 November 1999)
Non-SVP prison:
Patuxent Institution, Maryland (Visited 8 November 1999)
Discussion:
Virginia Law School (Visited 9 November 1999)
CANADA
1. In Canada we met with a number of people in the criminal justice system to talk about the Dangerous Offender Order. We also spoke to Dr Stephen Hart and Dr Vernon Quinsey about their risk assessment tools (HCR-20 and VRAG respectively).
Sentencing
2. The range of available sentences for serious offences in Canada is as follows:
Dangerous offender and long-term offender orders
Dangerous offender order (DOO)
3. DOOs are indeterminate sentences imposed on convicted offenders who pose a serious future risk to the public.
4. This type of legislation has existed in Canada for many years. The DOO in its current form was introduced in 1977.
5. In 1997 the law was changed so that, if the offender fulfils the criteria for a dangerous offender order, the judge cannot impose anything other than an indeterminate sentence. However, the option is still open to the judge to rule that the person does not fulfil the criteria and impose a long-term offender order (LTO) or another determinate sentence.
6. After parole eligibility is reached at seven years, there are biennial reviews of detention. Release from a DOO is rare and most such offenders can expect to end their lives in custody.
Long-term offender order
7. LTO orders ( determinate sentences with a fixed period of up to 10 years community supervision) were introduced in 1997 as a disposal for paedophiles with multiple past offences who could be controlled with appropriate community supervision. To give a LTO sentence the courts have to apply the 'confidence test': that they are 'confident that community supervision at the end of the sentence will protect the public'.
Flagging and information-gathering (British Columbia)
8. British Columbia has begun a system of flagging potential candidates for DO status on their computer system. The goal of the flagging system is to provide Crown Counsel with the necessary information it needs to proceed with a DO application.
9. There are at present 780 provincially flagged individuals. Of these, there are 200 to 400 who, when they commit their next offence, will seriously be considered for DO applications. 320 of the provincially flagged individuals are also flagged nationally: on the basis either that DO proceedings are being seriously considered; the offender is geographically mobile, or the offender is a violent Warrant Expiry Date candidate.
10. For all 780 provincially flagged offenders the following information is kept: current criminal records; a photograph; information on the circumstances of previous serious personal injury offences; court transcripts; parole information; psychiatric assessments and any other available information from Police and prison records. If a DO application is being considered, the complete file is made available to the Crown.
Definition of serious offence
11. In both the DO and LTO provisions a 'serious offence' is defined as being a violent or sexual offence that is punishable by 10 years or more. The offences themselves are not listed. The Courts have interpreted this to mean that all sexual offences are included, as are all offences of personal violence, particularly where a weapon has been used.
12. Doubts have recently been aired over whether, for a DO sentence to be passed, the index offence itself must be 'serious', or whether a past serious offence or an escalation of offences could suffice.
Assessment for DO and LTO orders
13. If an offence is 'serious' as defined in Section 752 of the Criminal Code, and the offender might prove to be a serious repeat offender, the Crown can request the court to order an assessment of the offender to be undertaken.
14. On the order for assessment being made, there is a 60-day period within which the Court-Ordered Assessor must complete his/her report. Practice has been to file the report within 75 days.
15. The report should be a clear opinion in non-legal language on two questions:
1. What is the person's risk of re-offending based on their past conduct?
2. What is the likelihood of successful treatment or control over the period of an appropriate definite sentence or other measurable period?
16. In about half of the cases that go to assessment, the Crown decides to pursue DO proceedings. The Attorney-General has to give consent to this application.
The sentencing stage
17. The sentencing judge is always the original trial Judge.
18. In the sentencing stage, hearsay evidence is admissible and the strict rules of proof do not apply, unless the prisoner denies any part of the Court Ordered Assessor's report, in which case these parts must be proven.
19. The prosecution has to prove the following:
20. Anything in the Assessor's report that is not proven must be subtracted from his/her report. It is after this part of the sentencing hearing that the Court-Ordered Assessor takes the stand.
21. The defence or prosecution are at liberty to call counter-experts.
The prison system
Correctional Service of Canada
22. The Correctional Service of Canada is the integrated custodial and community agency for Federal offenders (those serving more than two years). It is responsible for a total of 14 000 prisoners in custody and 8 000 in the community.
Post-conviction assessment
23. If offenders are convicted of a Schedule 1 (violent or sexual) offence (which accounts for approximately 70% of the Federal population) they are given an Offender Intake Assessment (a comprehensive assessment done by the intake parole officer) and a psychological assessment. The outcome of these assessments can also lead to the application of other specialised assessments. A detailed plan for the prisoner's time in custody, including details of programmes to be supplied, is then formulated, and seems to be subject to relatively little change as the sentence progresses.
Aboriginal offenders
24. A major problem in Canada is the serious over-representation of Aboriginal men and women in the correctional services. Aboriginal people constitute 2% of the Canadian population, but 19-20% of the federally incarcerated population.
25. Aboriginal villages are now working with the CSC to re-integrate offenders into their communities by using traditional healing processes in prison and community settings.
Mentally disordered offenders
26. There are very few secure psychiatric facilities in Canada except in prisons, which means that mentally disordered offenders are frequently given determinate prison sentences. We visited one psychiatric facility within a prison, the regional treatment centre (RTC) at Kingston Penitentiary. It treats a mixture of vulnerable prisoners, those who have developed mental illnesses in prison and some with long-standing major mental illnesses who would have been very unlikely to have been given a prison sentence in this country.
27. The RTC occasionally takes people with a primary diagnosis of personality disorder, but only temporarily, only if they have been causing problems in the mainstream prison system and not with any real expectation of being able to offer them treatment.
28. It was argued at the CSC HQ that a personality disordered person with a psychosis has a better chance of having both his/her criminality and his/her mental illness addressed by being held in a psychiatric hospital within a prison than if he/she was shuttling between a secure hospital and a prison.
Treatment of sex offenders
29. In 1994 there was an overall review of provisions for sex offenders in Canada, and general standards for programmes were set. These give guidance for the length and intensity of programmes, and their aims and design. Within these guidelines each institution is free to design its own programmes.
30. A tripartite system is in operation: the three levels vary in length and intensity. The high intensity programmes attract the most resource input. The people in the high intensity programmes are serious sex offenders who are personality disordered, very recidivistic or have overlying psychiatric issues. National standards suggest that there should be six to eight month programmes in maximum security conditions, with about 15 hours per week of therapy (normally cognitive-behavioural in nature). A study in the Pacific Region has shown that, at the serious end of the scale, untreated offenders had a recidivism rate of 60 to 70% over seven years, while the treated offenders recidivated at a rate of about 30%.
Release into the community
Voluntary sector partnerships
31. The aim is that programmes should be organised so that they take the prisoner seamlessly from the institution into the community. There is a new focus on community programming: in the last three years the prison population in Canada has gone down by 1000 and the number of offenders in the community has risen by the same amount.
32. The CSC works in partnership with voluntary agencies in the community. A major advantage of this type of inter-agency work is that the agencies can carry on with their work past the Warrant Expiry Date.
Warrant Expiry prisoners
33. Federal prisoners are released after two thirds of their sentence unless they are specifically thought to be dangerous, in which case they are kept in for the full length of their sentence (until Warrant Expiry Date) and then released without any supervision. This was designed as a last resort measure, but more and more prisoners are now being kept in until the end of their sentence. Staff feel that they are under pressure not to release people before the end of their sentence because they have to defend their decision in court if the prisoner seriously re-offends.
Recognizance
34. Section 810 of the Criminal Code is a good behaviour bond for those at risk of offending that is applied for to the courts by the police. It is time-limited and specifies certain prohibited behaviours. The maximum length of the bond is one year but it can be renewed. The bond acts in a similar manner to a probation order, and is particularly used for policing Warrant Expiry Date offenders.
35. Theoretically, if a person refused to sign their behaviour bond, they could be sent back to prison, but this has not yet been tested in a challenge under the Canadian Charter of Rights and Freedoms.
Notification
36. Community notification of an offender's release is normally to particular individuals, groups, or schools etc. In rare cases, notification is community-wide. Notification is being used less as the use of recognizances has increased.
Circles of Support Movement
37. The Circles of Support Movement, started in Toronto but now nationwide, consists of a small group of church and community members who provide active individualised support to an offender on release. It appears to have been very successful in minimising recidivism.
Conference: 'Risk Assessment and Risk Management: Implications for the Prevention of Violence'
17-19 November 1999, Vancouver
38. This conference was addressed by Professor John Monahan, Dr Marnie Rice, Dr Stephen Hart and others, and attended by many of those working at the leading edge of risk assessment and risk management. It had as its main focus the relative accuracy of various approaches to risk assessment, although several of the smaller lectures and workshops leant towards more practical questions of risk management.
Main plenary lectures
Professor John Monahan
39. The opening plenary lecture of this conference was by Professor John Monahan, Professor of Law and Psychology at the University of Virginia. Our Committee also met with him at Virginia Law School (see later section on the USA). Professor Monahan addressed the conference on the subject of the relationship between mental disorder and violence.
40. Professor Monahan observed that the assumption in law that clinicians can measure dangerousness and predict violence does not appear to be backed up by the research evidence. The question is therefore not whether clinical judgement should be used instead of actuarial assessment, but whether actuarial evidence should be used alone or in combination with clinical judgement.
41. He made the following observations about violence and mental disorder:
42. Professor Monahan noted that questions of the generalisability of screening instruments to different patient groups had not yet been fully addressed by researchers.
Dr Marnie Rice
43. Dr Marnie Rice was one of the authors of the Violence Risk Appraisal Guide (VRAG). Our Committee met with another of its authors, Dr Vernon Quinsey, at Queen's University in Kingston. Dr Rice is a proponent of the use of actuarial risk assessment tools to the exclusion of clinical judgement in making such assessments and she addressed the conference on the development of the VRAG.
44. The VRAG is a tool which was based on studies carried out by Dr Rice and colleagues on 618 mentally disordered offenders, all male, from a mixed correctional and mental health institution (Penetanguishene in Canada). The variables most predictive of violence were:
Professor Robert Hare
45. Professor Robert Hare, author of the Psychopathy Checklist (PCL), addressed the conference on psychopathy and its implications for the criminal justice system.
46. He observed that knowledge of the determinants of human behaviour and violence is relatively scanty because of the fearsome complexity of this area, and because until recently there was a lack of powerful research tools with which to investigate them.
47. He said that he thought it unlikely that there would ever be a single unified theory on human violence. However, he indicated that there was now the beginnings of a 'mini theory' of human predatory violence. Such violence can be defined as being instrumental, cold-blooded, predatory and owing more to the individual than to the context.
48. Psychopathy is a specific personality disorder (although Marnie Rice and others with a socio-biological perspective would argue that it is an adaptive strategy to aid reproduction). It is a clinical construct made up of clusters of traits.
49. The PCL was not designed as a measurement of risk, but instead measures the construct of psychopathy, which is only one factor in violence but may be the most important factor in predatory violence.
50. Psychopathy is a very important part of much sexual offending. Psychopathic sex offenders tend to be versatile in their range of offences. The combination of sexual deviance and psychopathy is a particularly powerful predictor of sexual recidivism.
Other lectures
Barbara Hart
51. Ms Hart, who works with the victims of domestic violence, made the case for practical risk management strategies as opposed to theoretical constructs and actuarial tools.
52. She observed that much of risk assessment, when applied to domestic violence, is stating the obvious: 'that batterers recidivate'. The important issues are therefore the identification of men who are violent against their families and the design of strategies to avoid the re-occurrence of the violence.
Deborah Ross, Rudiger Muller-Isbermer, Henrik Belfrage on HCR-20
53. These three presenters had investigated the use of HCR-20 in three different groups (mentally disordered offenders in British Colombia; mentally disordered offenders in Haina, Germany; offenders in maximum security prisons in Sweden). All of their results had shown that the HCR-20 had good predictive power for violent and sexual recidivism.
Martien W.G. Philipse (The Netherlands) and
Caroline Logan (Ashworth and Liverpool University)
54. Both these speakers raised concerns about the generalisability of many risk assessment tools (actuarial tools specifically) outside North America on the following grounds:
55. Something may be a good predictor everywhere, but may be more common in one country than elsewhere. For example, it appears from Professor Cooke's research that only 3% of the Scottish prison population are psychopathic, which is a significantly lower percentage than in North American prisons.
56. Something may be a good predictor in one culture but not in another. It appears from studies that some Belgian instruments may not even generalise to The Netherlands.
57. Cultures vary widely in their tolerance of and attitude to drugs, weapons, mental disorder, deviancy, the influence of religion, the acceptability of discussing emotions and particular personality disorders. This can affect the outcome of tests which are based on the norms of one culture.
58. Dr Logan said that there was a pressing need for a risk assessment tool to be developed specifically for Britain based upon British cultural norms.
THE NETHERLANDS
The TBS system
59. The Dutch TBS2 system serves a dual purpose: protection of the public from dangerous offenders, and care/treatment/rehabilitation of such offenders with a view to returning them to society.
60. In essence, TBS candidates are people who have been accused of committing very serious crimes which would receive prison sentences of four years or more (murder, rape, arson and other serious violent offences) but who may have some element of reduced responsibility in relation to the offence because of a mental disorder (a diagnosis which includes personality disorder).
61. Some 30% of TBS patients suffer from some form of psychotic illness, and some 60% of patients have drug and alcohol addiction problems. Most have a personality disorder of some sort. About 25% of TBS patients are sex offenders, 5% of the totality of convicted sex offenders.
62. TBS patients are given a prison sentence which consists of a punishment term in jail (appropriate to their level of culpability) and a TBS order. After serving their prison term they are transferred to a TBS clinic for treatment.
63. The average TBS population in recent times has been approximately 1100-1200.
64. Due to the high staffing levels required in order to manage TBS patients, the cost of the TBS system is some £250 per day per patient. By way of comparison, it costs around £80 per prisoner per day to run a mainstream prison in The Netherlands.
65. In the first five years after release from the TBS system, the recidivism rate for offences generally is 50-60%, but the rate for serious offences only around 20%. About one in six or seven TBS patients relapses into some sort of serious predatory behaviour following release.
Decision to assess
66. The Forensic Psychiatric Service carries out an initial assessment on a remanded prisoner and its report, based on a standardised list of TBS criteria, informs the Examining Magistrate's decision as to whether to have the individual assessed for TBS.
67. If a clinical assessment is undertaken, this is either at a psychiatric hospital or, for people suspected of having personality disorder, the Pieter Baan Centre.
Assessment
68. We visited the Pieter Baan Centre, one of the main facilities where pre-conviction assessments are carried out. Around 225-250 patients per annum are assessed at the Centre: patients undergo a residential assessment programme of approximately seven weeks' duration. The assessment aims to establish whether the patient has a personality disorder and whether or not he/she presents a danger to the public.
69. Twenty-five per cent of admissions to the Centre are found to be suffering from psychotic illness and 75% are found to have personality disorder. Around 60 to 80% have drug or alcohol abuse problems. A small number of cases also have learning disability. Ten per cent of admissions are female. Around 50% of referrals to the Centre are recommended for the TBS system.
Time spent in the prison service
70. The Dutch prison system has not traditionally provided any 'treatment' programmes at all to prisoners (although this approach is now being softened). The general effect for TBS prisoners is that the 'punishment' element of their sentence has to be served before their referral to a TBS clinic, yet for the duration of their time in prison (usually a number of years), they have not received any treatment of the sort provided in TBS clinics. This seems to make the job of rehabilitating them at the TBS clinics rather harder than would otherwise be the case.
71. There is currently an experimental programme in which a TBS programme is being operated in one wing of a prison in Amsterdam. However, it appears that the treatment is not working as well as in TBS clinics, at least partially because the prison is not physically appropriate for TBS treatment.
72. Staff from the Van der Hoeven Clinic also carry out some limited 'pre-therapy' treatment in prisons with people who are preparing for release to the community. However, this does not usually apply to patients who are due to move from prison to a TBS clinic.
Supervision of TBS offenders on release
73. TBS patients are monitored by the TBS system during their phased stages of release as they are gradually allowed more freedom in the community. Once they are fully released by the TBS system, the social work/probation service carry out monitoring for three years.
74. Conditions which could be attached to conditional release might include a requirement to reside in a particular place, or to continue attending for therapy at an outpatient facility.
Outpatient clinics
75. There are six outpatient clinics linked to TBS clinics in The Netherlands. We visited one, the Van der Hoeven clinic day treatment centre. Offenders can be referred to the Centre when they have committed an offence which is not serious enough to justify referral to the TBS system (for example serious assault but not murder). Most of the outpatients at the Centre are on probation, and if they fail to keep their appointments they can be returned to prison.
76. The Centre also undertakes work to avoid offence escalation with those who have committed mild sex offences . The Clinic works with the police to identify such individuals and to bring them into the Centre.
TBS and ECHR
77. The TBS system has not, to date, encountered any challenges in relation to ECHR. A key element of its compliance is the judicial review available to offenders in the TBS system every two years.
Length of stay in TBS
78. The average length of stay in the TBS system is increasing (it is now seven years). There may be two reasons why this is happening. The first relates to the increasingly serious types of offences which offenders in the TBS system have committed. The second is that diagnoses of patients in the TBS system are now more complex than hitherto.
'Treatability' of personality disorder
79. The general view amongst those we met was that personality disorder cannot be cured, but that the inclination of personality disordered offenders to offend can be treated, by addressing the particular internal conflicts within the individual patient, which cause the inappropriate/offending behaviour.
80. Twenty to 30 TBS patients have proven extremely resistant to any treatment and are unlikely ever to be released - most of this latter group are resident at the Veldzicht Clinic.
UNITED STATES OF AMERICA
81. In the USA, we visited a number of States in which Sexually Violent Predator (SVP) Legislation operates. We also visited the Patuxent Institution in Maryland, which is a prison with an emphasis on treatment programmes, and had a discussion with Professor John Monahan and others at Virginia Law School.
States operating Sexually Violent Predator programmes
82. SVP legislation of one sort or another is currently in force in 14 States. It is a civil commitment (i.e. the person is detained under mental health law) because the process is theoretically patterned on the mental health model (treatment must be provided and there is an attempt to get away from the punitive element of a custodial sentence) and, more importantly, because this avoids the legal challenges of double jeopardy and ex post facto detention.
83. Thirteen of the States operate this legislation post-incarceration (a person is considered for a SVP order towards the end of their punitive sentence). The exception is North Dakota, where the legislation allows the person to be committed from the community.
84. Washington was the first state to introduce civil commitment of sex offenders. The order was created in response to specific public demand within the State. Legislators we talked to in Washington State seemed rather surprised by the enthusiastic uptake of this idea across the States.
85. We visited Washington, Arizona and Wisconsin to discuss their various SVP programmes.
Procedure for commitment
86. In Arizona, a person who is due to be released within three months can be referred to the County Attorney. On the basis of an assessment of the records (and an interview, if the person consents), the Attorney decides whether to petition for commitment of the individual as a SVP. This is then reviewed by a Superior Court judge, to determine if probable cause exists to believe that the individual is a SVP. If a finding of probable cause is made, the individual is detained for an evaluation by an expert appointed by the court. (The State and the defendant can also appoint their own experts.)
87. Washington State's procedure is similar. Actuarial risk assessment tools are used to give a probability of re-offending. The results given by the actuarial tools are never clinically over-ridden (although it is not technically prohibited, and clinical judgement is used in addition to actuarial tools to make decisions on release). There is then a Probable Cause hearing where the offender's risk of re-offending must be judged to be in excess of 50% for them to be detained.
88. In Madison, the first stage is a paper screening using an assessment tool which is evaluated by a committee within the Department of Correctional Services. If the person is considered 'likely' to qualify as an SVP, a clinical evaluator does a more complete assessment. This then leads to a Probable Cause hearing. If 'probable cause' is found, the person is sent for detention at Mendota, where an evaluation for trial is done. This then goes to court and the judge decides if the person is to be committed. Ninety-five per cent of the time, the evaluation by Mendota is decisive. The process is not totally actuarial. They do look at whether people participated in treatment and, where they received any benefit, what people said about their situation. They also look at idiosyncratic factors (such as one recent referral who was aged 95 and so felt to be of lower risk).
89. The North Dakota power of civil commitment without having to be in prison apparently means (we did not visit North Dakota) that a person with previous sexual convictions can be sent to the SVP programme on arrest for any sexual offence. In effect, the prosecutor who has a strong case would prosecute under the criminal law, but if he/she has a weaker case, he/she has the facility to use the SVP legislation to detain the person civilly for an indefinite term.
Criteria for commitment
90. The criteria for commitment were relatively similar amongst those States that we visited, although phrased in slightly different ways. All States (except North Dakota) require that the person be within a few months of release from his/her determinate prison sentence, and they also require that the person be a danger and be suffering from a mental disorder of some sort (usually broadly defined).
91. Arizona's law provides for the civil commitment of individuals convicted or found guilty except insane of a sexually violent offence, who have a mental disorder that makes the person likely to engage in acts of sexual violence. Mental disorder includes paraphilia or personality disorders predisposing the person to commit sexual acts to such a degree as to render the person dangerous.
92. Candidates for commitment under Washington's SVP legislation are those who have been convicted of a serious sexual offence, have an Axis 1 or 2 disorder and are within six months of the end of their prison sentence. The legislation was deliberately drafted so that the decision on whether to commit an offender as an SVP would be based on the likelihood of re-offending rather than the diagnostic label. Non-sexual offenders are not covered by Washington's legislation, but are instead given fixed periods of supervision.
93. There are four criteria which need to be fulfilled for someone to be committed as an SVP in Wisconsin:
1. A relevant offence.
2. A mental disorder.
3. A high risk of re-offending. The statute uses the term 'substantial probability'. There has been considerable argument over the meaning of this. Judicial authority now says that it means 'much more likely than not' - in other words, the threshold is more than 51%. Other States use 'likely', which could be a 51% probability of committing an offence. Some States would take account of 'seriousness', so that a person might be detained on, for example, a 30% probability of committing a very violent crime.
4. Being within 90 days of release from prison.
Facilities
94. The Arizona Community Treatment and Protection facility has two sets of buildings. One is a secure facility. The other is the Alamo complex, which is the 'halfway house' - the first stage in the 'less restrictive alternative' programme of graduated release into the community, which is an important part of the treatment in Arizona.
95. The SVP service in Wisconsin appears to have been developed at speed from scratch, following the passing of the SVP legislation. It was developed from the programme at the Minnesota Psychopathic Personality Treatment Centre at Moose Lake, Minnesota. Wisconsin has two State hospitals, of which Mendota is one, and, at Oshkosh, a State prison which is administered by the Division of Care and Treatment for Mentally Ill Offenders. Mendota and Oshkosh manage the programmes for sexually violent persons, under Wisconsin's SVP legislation.
96. In Washington, the SVP facility is the Special Commitment Centre (SCC), a Department of Social Health Service's Mental Health treatment facility. It is situated just offshore, on McNeil Island, and is within the walls of a different prison.
Patient profile
97. The age range of people in the Arizona facility is 18-78, with an average age of 45. Around three quarters of the residents are diagnosed as paedophiles.
98. In Wisconsin, the most common diagnosis of the inmates is also paedophilia (over 50%), followed by one third with general personality disorder and 10% with paraphilia. Ninety-five per cent of Wisconsin's SVPs come from prison, and 5% from State hospitals.
99. The SCC in Washington is split evenly between residents who have committed paedophile offences and those who have committed rape against adults.
Numbers of patients
100. All of the places that we visited indicated that numbers of inmates are rising significantly faster than predicted, which is causing problems with accommodation, programme provision and staffing. It was notable that they were all undertaking building programmes for new, larger facilities when we visited them in November.
101. However, it was commented in Arizona and Washington that, in the longer term, it is likely that fewer people will come into the SVP programme than at present, because sentences for sex offences have lengthened dramatically in recent years. In some States, there is also the effect of the 'two strikes' law to be considered, which means that a person will automatically receive a 'natural life' sentence for their second sexually violent offence.
Treatment programmes
102. All of the States we visited based their treatment programmes on a thorough intake assessment and regular assessments throughout the treatment programmes. There was plentiful use of techniques such as penile plethysmography and polygraphy throughout. An individualised treatment programme is then drawn up.
103. Treatment in Phoenix is a six-stage programme based on research which suggests that there is no 'cure' for sex offending, but that there is effective treatment for many, if not all, offenders. The programme uses a relapse prevention model, using a cognitive behavioural therapy approach, and with the lead taken by psychologists.
104. The Washington SCC's standard programme is an extremely detailed four-year treatment programme, essentially cognitive behavioural in nature, with detailed attainment goals for promotion towards release. No-one has yet completed it, and, in fact, only about 30% of the total population is actually participating in the full programme.
105. Wisconsin's programme appeared to consist of a number of education modules, including sex education and some cognitive skills, and some therapy models such as victim empathy, conflict resolution etc. The programme is evolving, as the numbers of patients is rising almost exponentially.
Discharge and community supervision
106. The States are all subject to the principle of the least restrictive alternative, which means that if, on appeal, the judge agrees that an offender could be held in lesser security, this must be done.
107. In Wisconsin 11 people had left the programme, all on supervised release. They were all released by judges against the recommendation of the Mendota Institute. On release from Wisconsin's facilities, there is an extremely long list of conditions with which the person has to comply indefinitely.
108. In Washington the first people approved by the SCC for release were to leave within six months of our visit. However, four offenders have already been released, all contrary to the SCC's recommendation. People who are released from the SCC are also subject to stringent controls in the community, although this community supervision is time limited.
109. Phoenix appeared to have the most co-ordinated release programme. The three levels of custody, a halfway house and discharge into the community were operated by a single agency. No-one had yet progressed to the community, but residents of the halfway house had an active community presence. Most had jobs or college placements, but were subject to intensive monitoring and continued participation in treatment programmes.
Community notification
110. Arizona has a version of 'Megan's Law', which means that sex offenders being released into the community are notified to the State police, who then notify the neighbourhood. The police hold public hearings, and these are often run on local TV. There are also Internet sites identifying sex offenders.
111. Wisconsin and Washington also have strict community notification laws which must be met. Washington claims to be a leader in the field of controlled community notification - their notification takes place by way of public meetings but the aim is to explain the controls that surround SVPs on release. However, doubt was cast on the effectiveness of this approach at the meeting with State prosecutors in British Columbia.
Other issues raised at SVP facilities
Patients' rights
112. One particular problem faced by staff in Wisconsin and Washington is that of the rights of people held under SVP legislation. As SVPs they are technically detained under the States' respective Mental Health Codes, which ensures them certain privileges not given to those in prison. They are therefore permitted unrestricted access to mail and phone calls, which has proven extremely problematic.
113. Wisconsin is looking at ways to restrict patients' rights to a level more appropriate to sexual offenders with personality disorders rather than to mentally ill people. This would involve having two different mental health classifications, one being people with 'victim dependent disorders', and another being the 'vulnerable mentally ill'.
Treatment refusers
114. Non-compliance with treatment is a serious problem in all facilities, as treatment refusers tend to be the most psychopathic of patients and disrupt the facility and other patients.
115. In Wisconsin they have separated out what they call 'behaviourally problematic treatment refusers' from the rest of the people on the SVP programme. They are looking at trying to provide incentives for treatment, by scaling down the privileges of treatment refusers.
116. In Washington, 40% of residents are not in treatment. They are offered some other, educational, activities but recently had to be separated off from the rest of the residents because of concerted group attempts to sabotage the programme.
117. In Arizona, inmates are entitled to refuse treatment, but those who are in treatments get the best facilities and more privileges. Inmates who refuse are asked again every 60 days. About 20 or 30% refuse treatment or drop out.
Qualification as SVP
118. In Wisconsin it was notable that people who have committed paedophile offences will almost automatically qualify as SVPs since this offence is taken to be in and of itself conclusive of three of the four necessary factors that diagnose someone as an SVP -namely, it is a relevent index offence, it indicates the presence of mental disorder (since they are prepared to diagnose paedophilia as a mental disorder), and there will normally be a high score for risk.
Mental health
119. One problem faced by the SCC in Washington is that, because it is technically a psychiatric hospital (although a prison to all intents and purposes), people who genuinely have mental illnesses are not permitted to go to a more appropriate psychiatric facility for treatment. There are therefore three patients who are floridly psychotic and 24 other people on psychotropic medication at the SCC. This is cause for some concern amongst staff. Some 33% of residents have been removed from the mainstream treatment programme and are now in the Special Needs Programme because they have learning or emotional deficits or mental health problems.
Problems on release
120. There are at least three agencies involved in every case of release from Wisconsin's facilities, and there are considerable issues of co-ordination and accountability. There are also severe problems in finding accommodation and employment for these individuals.
Maryland - Patuxent Institution
121. The Patuxent Institution is a prison, but it operates at one remove from other prison and mental health institutions in Maryland. It has its own legislative base and its own parole procedure. The model appears to be unique in the US (other treatment-based forensic services are either in mental health institutions or based around the sexual predator legislation). The 750 people in Patuxent (of whom 90 are women) are part of a total prisoner population of 23 000 in Maryland. Eighty-five per cent of the inmates in the institution have been sentenced for violent crime, with others in for crimes such as large-scale drug dealing.
122. The population served by the Institution is highly selective. Apart from young people, who can be sent there against their will, prisoners must volunteer to come. It would appear that there is considerable demand for places, at least partly because the facilities are somewhat better than in other prisons.
123. To be eligible for the Institution, prisoners must have an intellectual impairment or 'emotional imbalance'. This appears to be a slightly broader term than personality disorder, but it was said that most of the prisoners would have severe personality disorders, mostly ASPD. Because the programme at Patuxent lasts a long time (normally 7 to 12 years), prisoners must be serving a long sentence and have a significant part of this remaining.
Treatment
124. Patuxent follows a multi-disciplinary team approach involving Remediation Management Teams (RMTs), comprising one psychologist, a social worker, a prison officer and a psychiatrist for each prisoner. Each prisoner has an individually based programme, with an annual review. Essentially, the programme would seem to be a mixture of some psycho-dynamic group work and a selection from a range of individual treatment programmes.
125. The assessment process is directed towards screening out people with high psychopathy scores, who are viewed as likely to get worse through treatment and also as dangerous to the system.
126. There is an annual review of each case by a Board, which consists of four staff from the facility and five lay members. There is a Conditional Release Programme which works on a series of levels. The last stage is parole to the community, which is also decided upon by the Board.
Information availability
127. Staff at Patuxent claim to have access to a fairly full range of information about offenders' backgrounds. It is acknowledged that the crime for which a person is convicted may often be reduced as a result of plea-bargaining. However, staff have access to police reports and other information, and staff tend to assume that the person did what they were charged with, not what they were convicted of. Also, through participation in the programmes, other information about past behaviour often emerges.
Recidivism
128. No-one has been re-arrested in the last three years who has been released into the community. There is a 7% return for parole violations, mostly drug related. This compares with a 42% re-arrest rate for the mainstream services. This is unlikely to be due solely to the efficacy of treatment, but due also to the long time served, the close level of supervision and perhaps the self-selecting nature of the group.
129. Maryland State has a community-based Sex Offender Programme, which involves weekly group work, which seems to be quite successful in maintaining people in the community. This has been going for 25 years and recidivism is 8% for any offence and 3% for sexual offences.
Sex Offenders' Register
130. The Sex Offenders' Register has been in operation since 1994. It is held by the police, and is public, thus vigilante behaviour and victimisation cause considerable problems for offenders and the parole system.
SVP in Maryland
131. There was an attempt to introduce a SVP-type indefinite commitment in Maryland, but this was defeated in the legislature. Part of the reason was that such offenders, if committed civilly, would have had to be housed in a secure psychiatric facility, when psychiatric intervention seemed to legislators to be inappropriate for people with personality disorders, and the State's forensic psychiatry service was already severely over-stretched.
1. Virginia Law School: Discussion
132. In addition to visiting facilities in the USA, we also had a most useful discussion at Virginia Law School with three experts in psychology and the law: Professors John Monahan, Richard Bonnie and Steven Morse. The following topics were part of the wide-ranging discussion.
Risk assessment
133. Risk assessment has greatly improved because of the introduction of structured instruments for risk. In North America, the VRAG is probably the 'gold standard'. There is an active debate between those who advocate an actuarial approach modified by clinical judgement, and those who support a purely actuarial approach.
134. Because many factors which indicate risk do not change, except age, it is difficult for offenders to show that risk has reduced. However, there is evidence that close community supervision can be effective.
Dangerous offenders legislation
135. If a special sentence is required, it was suggested that it should be in the criminal justice system. There is no strong evidence that treatment of the most dangerous group will reduce risk, so the purpose of detention is incapacitation.
136. Nevertheless, there is an ethical duty to help people so detained, provided there are plausible treatments. There is work that can be done with sexual and violent offenders, particularly on specific areas of concern, such as substance abuse.
137. There are strong arguments for 'getting the sentence right' at the start, rather than introducing a new disposal at the end of a sentence. Most factors relevant to risk assessment could be known at the time of sentencing.
Personality disorder and detention
138. Detention on the basis of risk was to be preferred to detention on the basis of a diagnosis of personality disorder. Many high risk individuals were not personality disordered, and many personality disordered people were not high risk.
Sexual predator legislation
139. There was concern that civil commitment for 'sexually violent predators' has the potential to do great damage to the public mental health system by dumping sex offenders into mental hospitals. This creates a risk to people who are genuinely mentally ill.
140. Sexual predator legislation tends not to be constrained by costs, because of political concern, and so considerable amounts of resources are being eaten up by it. In general, it tends to be introduced following a single high profile case.
Megan's Law
141. Sex Offender Notification Legislation ('Megan's Law') is widespread throughout the States. It tends to work on a model of levels of risk. High risk will involve everyone in the community being notified. Medium risk will involve child care agencies being notified, and low risk will involve the police only being notified. There is a danger of over-estimation of risk, because of the potential for criticism if anything goes wrong. There does not appear to be any research evidence that the law protects the public, and some concern that it leads to vigilante behaviour.