Women in Prison : a report by the Anti–Discrimination Commission Queensland

---

6.1. The classification system
6.1.1 The Corrective Services Act 2000 (Queensland) (CSA)

Prisoners’ classification levels have a direct impact on their quality of life. It affects which prison they are placed in, their level of supervision and privileges, and directly impacts on critical decisions such as when they may be granted early release. The validity of the classification process should be of serious concern for all prisoners, and persons involved in administering prisons.

Like many other jurisdictions, under the CSA in Queensland, prison authorities are required to assign a classification to a prisoner.[67] The Act sets out five levels of classification: maximum, high, medium, low and open security. While the Act does not expressly state the purpose of the classification of a prisoner, the DCS says:

• the purpose of a security classification is to signify the prisoners’ institutional security risk;

• identify the prisoner’s level of escape risk on escort; and

• assist in identifying the most suitable placement options for the prisoner.[68]

The DCS further states that:

• the classification system ensures prisoners who are a high risk, due to their likelihood of attempting to escape, the risk they pose within a correctional services facility and their likelihood of re–offending, are accommodated in a secure facility with appropriate levels of supervision; and that

• prisoners who are assessed as being unlikely to escape, who present a low risk within the corrective services’ facility and pose little or no risk to the community can be accommodated in facilities with less security and supervision.[69]

The CSA presently sets out 12 factors to be taken into account in deciding a prisoner’s classification,[70] however, at the time of writing this report, the DCS is reviewing the Act and re–examining the classification system.[71] The factors include:

• the risk of the prisoner to the community; (s12(3)(a));

• the nature of the offence for which the prisoner has been charged or convicted (s12(3)(b));

• the period of imprisonment the prisoner is serving; (s12(3)(c));

• whether the prisoner has any outstanding charges and the nature of the charges; (s12(3)(d));

• the prisoner’s criminal history; (s12(3)(e));

• the prisoner’s escape history (if any); (s12(3)(f));

• the prisoner’s demonstrated attitude towards the sentence being served; (s12(3)(g));

• the likelihood of the prisoner being deported or extradited, and the prisoner’s demonstrated attitude towards the deportation or extradition; (s12(3)(h));

• the prisoner’s previous conduct in a corrective services’ facility including whether a prisoner has committed an offence or breach of discipline or returned a positive test sample; (s12(3)(i));

• the prisoner’s previous conduct while subject to a community–based order or a post–prison community–based order; (s12(3)(j));

• the prisoner’s medical history (including psychological or psychiatric history) (s12(3)(k)); and

• the likely influence of the prisoner’s family relationships; (s12(3)(l)).

While the classification system under the CSA is the same for men and women, current DCS policy is that no female prisoners are classified as maximum security[72] and all women on remand are classified as high security (CSA s12 (1)).

All women, upon being sentenced and entering prison, are initially housed in the secure area of either BWCC or TCC until they are assessed. An initial security classification must be completed within 14 days of admission, and prisoners serving more than 12 months must be further assessed under the Offender Risk/Needs Inventory (ORNI) [73]within 21 days of admission.[74] Prisoners have a right to seek a reconsideration of their security classification if it has been changed, within seven days of the date of that change.[75] After the initial classification has been assigned the CSA requires that a prisoner’s classification be reviewed at intervals of six months or less.[76]

In December 2005 the DCS advised that:

Further, the DCS states:

Simplification of the classification system is a positive step forward. However, the critical issue is how classification levels are determined. If not done properly, there is a risk that offenders will be sent to facilities with higher than necessary levels of security, which are costly and provide restricted access to correctional programs and activities. Conversely, dangerous offenders might be assigned to lower security facilities with all the implied risks to correctional workers and fellow inmates.

Two basic tests, one for validity, and another for reliability, would determine whether or not the approach in allocating classifications is fair and based on legitimate assessments. The first test would verify that the assessment tool measures what it is intended to measure. The second test would reveal if different users operated the same tool in the same way and came up with consistent results.

The Auditor General of Canada and the Canadian Public Accounts Committee have both recognised the vital importance of the integrity of the classification instruments for classifying prisoners. It is incumbent on those administering and overseeing the Queensland prison system to ensure similar levels of accountability.

---

6.1.2 Existing process of classifying a prisoner

Generally, a prisoner’s initial security classification is determined and allocated by a sentence management corrections officer (A04 level) at the reception centre of the prison after sentencing. The corrective services officer responsible for making the initial recommendation or classification receives on–the–job training in the methodology of classification.

Depending on the length of sentence, the officer may not have full authority to determine a classification, but may only have power to make a classification or recommendation, which requires final approval by the General Manager. The ADCQ has been informed that the General Manager changes the classification recommendation in about 6% of cases. The Executive Director of Custodial Operations reviews the classification of prisoners sentenced to a term greater than 10 years. The ADCQ has also been informed that with prisoners serving sentences of two years or more, classifications are automatically checked and verified by the DCS Office of Sentence Management to ensure consistency, the absence of facility–based bias and appropriateness to the prisoner’s circumstances.

The ADCQ is advised that the classification documentation is recorded electronically, and that the text field presently on the DCS information system predates the CSA.

Therefore, the factors presently set out in section 12 of the CSA are not fully reflected in classification documentation in the computer system. The ADCQ has been provided with documentation used by the Sentence Management Officer in making a classification recommendation. [78]These documents set out guidelines for officers on how to gather information or weigh the factors set out in section 12(3) of the CSA.

In meetings with corrective services officers, the ADCQ was told that the determination of a prisoner’s classification level is done in the following manner:

• the officer looks at the prisoner’s convictions;

• the officer works through the 12 factors set out in section 12(3) of the CSA. Information is gathered from a range of sources. No score or weighting is assigned to any of these 12 individual factors;

• the guidelines for outstanding charges are considered;

• the correctional officer balances all the aforementioned information, and considers it holistically to determine or recommend a classification level.

The ADCQ is of the view that section 12(3) of the CSA includes factors that may pertain to some minority groups of women (and men) possessing attributes (covered by the ADA) to a greater degree than women (or men) without these attributes. This view has been confirmed by the ADCQ’s assessment of the DCS classification guidelines document.

This potential discriminatory effect should be identified and eliminated in the proposed legislative amendments, and associated guidelines and procedures.

The ADCQ has concerns about the assessment process for the following minority groups:

1. Women with mental health or intellectual disabilities
Women with mental health issues or intellectual disabilities are a unique category within section 12(3)(k) of the CSA, which refers to the medical, psychological and psychiatric history of the prisoner. Even when such a prisoner is assessed suitable for open classification, the classification guidelines state:

Prisoners who are assessed as suitable for an open classification, but on the advice of the medical, psychological or psychiatric staff, require medical and support services which are unavailable in open custody, should be classified as low security and remain in secure custody.

This is prima facie direct discrimination on the basis of disability. The DCS may be able to argue an exemption, but the critical question that would need to be determined separately in each case would be whether the provision of the necessary medical and support services in an open classification facility would constitute an unjustifiable hardship on DCS. The current guideline does not provide for this.

2. Women from culturally and linguistically diverse backgrounds
Women from culturally and linguistically diverse backgrounds are more likely to fall under section 12(3)(h) of the CSA. The classification guidelines state generally that a number of these prisoners are 'unacceptable for progression.' This may be argued to be prima facie discrimination on the basis of race.

3.Indigenous women
Indigenous women, who have a higher recidivism rate than non–Indigenous women, may more frequently fall under classification criteria that result in higher level classifications than other female inmates due to sections 12(3) (e) and (i) of the CSA.[79]

The ADCQ is strongly of the view that classification levels must be supported by research and evidence that is both reliable and relevant. Discriminatory decisions can result from unjustified and unfair assumptions being made about the risk levels of individuals from certain minority groups, which are based on criteria that are unreliable.

The classification system has a significant impact upon the day–to–day life of a prisoner. It is therefore of critical importance that the system be managed and operated in a way that maintains high standards of credibility, consistency and accountability. Otherwise the system risks encouraging or tolerating corrupt and incompetent practices. It also risks officers making unlawful assessments based on ill–informed or unarticulated prejudices about gender, race, disability or religious belief.

In spite of our examination and understanding of the existing classification process and assessment criteria, the ADCQ is not confident that the present system has integrity in measuring the prisoner’s likelihood of escape and re–offending, the impact on the community, or the risk the prisoner poses within the corrective services facility.

While some research has been done on how to correctly predict a prisoner’s risk to the community (upon which the ORNI is based), very little has been done on measuring correctional or institutional risk. Given the important consequences that can flow from a prisoner’s classification, the present method of risk assessment appears to be highly subjective, with few levels of quality control to ensure consistency of decision making among officers.

The ADCQ is concerned that there is a strong possibility of systemic discrimination occurring in the classification of female prisoners, particularly, those who are Indigenous. This view is based on:

• a greater proportion of Indigenous women compared with non–Indigenous women who receive high level classifications;

• likely inadequacies and inconsistencies of the on–the–job training of officers performing the risk measuring role; and

• the lack of research justifying the integrity of the risk measuring tool, especially for women and prisoners from minority groups.

The ADCQ urges the DCS to undertake research on the cultural relevance and validity of its existing classification and reclassification tools. Such tools should take into account the structural factors that may result in systemic discrimination against women, particularly Indigenous women, culturally and linguistically diverse women, and women with mental health issues.

'Human rights law requires that assessment tools be responsive to the populations to which they are applied and properly crafted to meet the purpose that they were intended to serve.'
(Canadian Human Rights Commission 2003)

---

6.1.3 Classification and the custodial infrastructure

There are a number of compelling arguments for classifying women prisoners at the lowest level necessary for ensuring the good order and security of prisons, and community safety. The most compelling argument is the psychological and social health of the prisoner, but also the rehabilitation of offenders is also more likely to be facilitated where prisoners have higher levels of responsibility and self control.

Until recently all women prisoners in Queensland were kept in high security prisons and intermingled at all security levels. This was due to the very low numbers of women in the state’s prisons. Only in the last decade have efforts been made to provide open security facilities for women.

Queensland’s use of open custody facilities is one of the lowest in Australia. Of the beds allocated to female prisoners, 25.6% were low/open classification beds. Male prisoners had 17.6% of their beds in low/open custody.[80] This compares to a national average of 27.3% of all prisoners (male and female) across Australia.[81]

The DCS makes it very clear that the present classification system does not directly relate to placement decisions, and the classification system does not match the available custodial infrastructure.[82] There are only five prisons in the state that accommodate women, and the largest of these, BWCC and TCC, are secure facilities.

A prisoner’s classification level does not necessarily reflect the type of facility in which she will be accommodated. Prisoners classified as high, medium or low in BWCC and TCC are all held in secure facilities with similar security measures and levels of supervision. Therefore, many female prisoners with low risk profiles are being accommodated in secure facilities with inappropriately high levels of security and supervision. Not only is this an inefficient use of resources, it results in them being placed behind high barbed wire or electrified fences, and being subjected to a heavy presence of uniformed officers, the risk of placement in solitary confinement as a punishment, rigidly enforced rules, lack of independence in day–to–day living, and a sharp demarcation between staff and inmates.

Institutionalisation of women prisoners to this level does not assist in their rehabilitation. It does not adequately prepare them with the necessary skills to live independently, especially where the issues that could lead to re–offending have been addressed.

The DCS has recently raised the issue as to whether there are benefits in aligning the classification system with the types of correctional centres.[83] It has been suggested that a prisoner with a secure classification would be accommodated in a secure custody facility, whereas an open custody prisoner would be managed in any custody facility, including an open custody facility, to suit the prisoner’s management and risk needs. The ADCQ is most concerned that the DCS considers this an appropriate management strategy of prisoners, particularly female prisoners.

The keeping of open classification prisoners in a secure facility is not best practice. All efforts should be made to ensure the open classification prisoners are accommodated and remain in open facilities, even when they are in need of medical or other services.

Rather than aligning prison classification with infrastructure, an emphasis should be placed on all women’s correctional facilities being situated and designed to enhance community interaction, with an aim of successfully returning the prisoner to the community. Female prisoners should be located as close as possible to their homes and families. Only a very small minority of women prisoners are seriously violent or predatory. The majority of women prisoners can be appropriately managed in facilities that are based on community living, with prison regimes and practices to encourage positive supportive interaction between staff and residents. The highest priority should be given to the interests of children in determining the placement of mothers serving full–time sentences.

DCS has already developed a modest but highly effective and innovative model that could be expanded and made available to many more female prisoners. The Warwick Women’s Work Camp appears to have achieved commendable results in rehabilitating women prisoners with the assistance of dedicated and talented staff who manage programs and activities that promote self–esteem, personal skills including living and inter–personal skills, teamwork, and a variety of job skills. The combined efforts of the women in performing a range of community work in this medium sized rural centre, appear to be highly valued, worthwhile, and appreciated within the Warwick community.[84]

In December 2005 the DCS advised ADCQ that:

in principle support has been given to the establishment of a women’s camp in North Queensland and an additional camp in South East Queensland.

The successful elements of the WWWC model could be analysed and developed to fit a broader range of women prisoners, including those less physically able or who have children residing with them in prison.

---

6.1.4 Indigenous prisoners and the security classification system

It has been asserted that Aboriginal women are disproportionately classified as high security prisoners, and that the majority of Aboriginal women in prison are kept in the high security prisons.[85]

Both male and female Indigenous prisoners generally have a higher security classification than non–Indigenous offenders.

Table 5: Prisoners by security classification as at 30 June 2005

Prisoners by security classification as at 30 June 2005

Security classification	Indig– enous	MALE Non Indig– enous	Total	Indig– enous	FEMALE Non Indig– enous	Total	TOTAL
High security facilities
High	459 (25%)	1375 (74.9%)	1834	40 (33%)	80 (66.6%)	120	1954
Medium	466 (27%)	1210 (72%)	1676	27 (27.5%)	71 (71.4%)	98	1774
Low	46 (14.4%)	218 (82.5%)	264	5 (23.8%)	16 (76%)	21	285
Open	72 (22.8%)	243 (77%)	315	5 (22.7%)	17 (77%)	22	337
Unclassified	42 (28.7%)	104 (71%)	146	6 (35%)	11 (64%)	17	163
Total high security	1085 (25.6%)	3150 (74.3%)	4325	83 (29.8%)	195 (70%)	278	4513
Low security facilities
Low	3 (10.7%)	25 (89%)	28	5 (45%)	6 (54%)	11	39
Open	148 (20%)	585 (79%)	733	8 (11%)	64 (88%)	72	805
Total low security	151 (19.8%)	610 (80%)	761	13 (15.6%)	70 (84%)	83	844
TOTAL	1236 (24.7%)	3760 (75.2%)	4996	96 (26.5%)	265 (73.4%)	361	5357

Source : Department of Corrective Services Annual Report 2004–05 Table 4 page 45

In June 2005, Indigenous women constituted 33% of all female inmates classified as high security in Queensland, 27.5% of all those classified as medium security, 23.8% of those classified as low security and 22.7% classified as open security, while being held in high security facilities. Indigenous women constituted just 15.6% of those in low security facilities. At the same time, Indigenous women represented 26.5% of the total female prison population. Similar statistics for the period 2003–04 confirm that this is a recurrent pattern,[86] and that Indigenous women are more likely to be held in secure custody, and are over–represented at higher classification levels.

The ADCQ acknowledges that the nature of the offence and the length of the sentence are factors to be considered in determining classification, and that the types of offences, for which Indigenous women are incarcerated, have a distinctly different profile from those of non–Indigenous women. The ADCQ is concerned, however, that a large proportion of Indigenous female prisoners are classified as high security, with few classified at the lower levels.

The ADCQ and other observers have noted that many offences committed by Indigenous women, especially those living in remote Indigenous communities in Far North Queensland, occur in dysfunctional domestic situations that can be fraught with violence towards women and children. Women who commit violent crimes in these situations, often may have been long term victims of violence from male members of their households. Once these women are removed from those situations of domestic violence, they pose an extremely low level risk of escaping or re–offending.

While the risk of self–harm may be high and warrant careful management and supervision within a correctional facility, the weighting given to the type of offence and length of sentence for Indigenous prisoners, could result in an unjustifiable classification when assessing the likelihood of escape or re–offending. The integrity of the classification system for Indigenous women must be able to be justified against credible research to show that accurate and reliable measures are being used. To date, the integrity of the current classification system as a measuring tool has not been demonstrated to the ADCQ.

---

6.1.5 Remand inmates

Women on remand may spend a considerable period of time in prison waiting for the case against them to proceed to a final hearing by the appropriate court. Some of these women may eventually be found not guilty of the offence for which they have been charged. It is unacceptable that any woman must automatically be classified and managed as a high security prisoner, if this assessment based on all the usual criteria is not warranted. Remand prisoners are not assessed under the ORNI, and do not get access to programs or training. This is detrimental for any prisoners who may be held in custody on remand for a significant period of time.

In December 2005 the DCS advised the ADCQ:

that new legislation proposed for 2006 will allow unsentenced prisoners who are remanded in custody to be classified according to the security risk they pose.[87]

The ADCQ supports this important positive change in the management of remand prisoners.

---

6.1.6 Classification and access to transfers, community work orders, release to work, home detention and parole

A prisoner’s classification is also considered in any request to transfer from one facility to another or for deciding which prisoners will be granted supervised community release such as release–to–work, home detention or parole. The classification decision therefore impacts on a range of other critical decisions concerning a prisoner while incarcerated. The integrity of the classification system as a measure of risk, and the reliability and consistency of its application as a measuring tool, are extremely important issues for a prisoner.

---

Endnotes

67. CSA s 12.
68. DCS Submission to Women in Prison Review (10 September 2004) 10.
69. 'Prisoner classification consultation paper' above n 35, 9.
70. CSA s 12(4).
71. DCS Submission to Women in Prison Review (14 December 2005) 8.
72. Queensland Department of Corrective Services, 'Maximum Security orders' Department of Corrective Services Procedures — Offender Management version 00 (23 July 2004).
73. The ORNI is discussed in detail at 6.2.2 below.
74. Queensland — Department of Corrective Services 'Assessment' Department of Corrective Services Procedures — Offender Management version 02 (28 May 2003).
75. Corrective Services Regulation 2001 (Qld) s 4.
76. CSA s 12(4).
77. DCS Submission to Women in Prison Report (14 December 2005) 8.
78. Queensland Department of Corrective Services, above n 73; DCS 'Review' DCS Procedure — Offender Management (version 02)24 January 2003; DCS Classification guidelines; DCS Form — Initial Security Assessment Guidelines (version 02).
79. Discussed more fully at 6.1.4 below.
80. DCS Submission to Women in Prison Review, (14 December 2005) 2.
81. 'Prisoner classification consultation paper' above n 35.
82. Ibid 14 – 17.
83. Ibid 15.
84. DCS Submission to Women in Prison Review (14 December 2005) 11.
85. Sisters Inside Inc, Submission to Women in Prison Review (June 2004).
86. In June 2004 , 36% of high security, 30% of medium security, 12% of low security and 21% of open security prisoners female prisoners being held in a secure prison were Indigenous. At the same date, Indigenous women comprised 27.85% of the total female prison population. At the same date no Indigenous women were held in community custody, when 27 non–Indigenous women were being held in community custody. DCS Annual Report 2003–04.
87. DCS,Submission to Women in Prison Review (14 December 2005) 8.

 

Back to contents for chapter 6 - Human rights and the corrective service systems

Table of Contents for Women in Prison report