Criminogenic needs
The New Zealand Criminogenic Needs Inventory (CNI) was developed by the Department of Corrections as a tool to identify the criminogenic needs of the New Zealand offending population. The CNI complements the Risk of ReConviction models, identifying why offenders are at risk 1. The CNI is offence focused and includes the assessment of psychological needs, responsivity factors linked to offending and the role of culture in the offending period (starting the day before the offence and finishing at the completion of the offence) and the pre-disposing period (six months prior to the offence). The CNI has been validated on New Zealand offenders and has been found reliable and valid when compared with other measures of needs. Criminogenic needs specific to women are not identified and norms for female offenders are not available although the latter was recommended (Coebergh, Bakker, Anstiss, Maynard, & Percy, 2001). "There is no direct evidence that in using [the CNI and RoC*RoI] the risk and needs of women offenders cannot be accurately identified, however, there has also been no adequate development of a specific women's gender risk prediction tool or needs assessment that takes into account any gender specific factors for female offenders" (Department of Corrections, 2003a, p. 4). A recommendation has been made to include personal and emotional adjustment as an additional criminogenic need for women. If empirical data were available supporting the notion that assessing for dynamic factors in a considerable period before the offending (such as the pre-disposing offending period) predicts women's re-offending with greater accuracy, the Pre-Disposing Period Criminogenic Needs Inventory for female offenders would provide valuable information.
The Maori Culture Related Needs 2 (MaCRNs) of the CNI were developed as a partial response to address over-representation of Maori in the criminal justice system (Maynard, Coeberg, Anstiss, Bakker, & Huriwai, 1999) "on the basis that there are specific and unique needs to Maori offenders. These needs are characterised by culture and the place of that culture in New Zealand society and which, if not addressed appropriately, are likely to contribute to an increased risk of re-offending by that individual" (Coebergh, Bakker, Anstiss, Maynard, & Percy, 2001, p. 16). These needs were developed independently from the Criminogenic Needs Inventory and integrated later in the CNI. No studies on Maori female offenders were found, although indications exist that "Maori women are just as likely to have identified MaCRNs as men, particularly in relation to whanau influence to crime and cultural identity" (Department of Corrections, 2003a, p.32). However, this needs further investigation.
In its Maori Strategic Plan the Department of Corrections (2003b) endorsed its support of the Framework for Reducing Maori Offending (FReMO) and commitment to early intervention and prevention by developing "a framework for addressing the needs of Maori women offenders" (p. 12).
Consultation with cultural advisors for this project on female offenders revealed that in general the position and acceptance of Maori female offenders in the whanau and in their community would not change because of the crime they committed. However, it was said that female sexual offenders are possibly a "forgotten" group because of whakama, their shameful actions, similar to how Maori females in prison are perceived. In addition, violent women may not be acknowledged ("It has nothing to do with me. It belongs in that family") in contrast to joyful events embraced by everyone. Reasons given for why Maori in general do not function as they normally might do were connected to the fact that links of accountability became severed due to processes of losing land (no land equals a person without mana), being unable to provide and to protect the whanau (as intrinsic to Maoridom) and loss of connections and responsibilities within hapu and whanau due to urbanisation (M. Neho, personal communication, August 9, 2005; M. Rolleston, personal communication, July 27, 2005).
A small number of studies have identified particular criminogenic needs in violent Australian Aboriginal male offenders, such as unemployment, alcohol abuse and domestic violence (Howells, Day, Byrne, & Byrne, 1999) which affects responsivity to treatment if not taken into account with programme design and delivery. Blanchette and Motiuk's (1997) comparison of Canadian maximum-security women and men revealed that the maximum security female offenders were more likely to be Aboriginal, had a higher suicide risk potential and difficulties in every criminogenic needs domain, significantly more in the marital/family and substance abuse domain. No differences in criminal history were found, except for sex offence history (males). More recent research examining Canadian Aboriginal and Caucasian incarcerated women showed substantial higher needs for Aboriginal women in the domains of substance abuse, employment, marital/family and association/socialization whilst Caucasian women offenders rated consistently lower levels of the seven needs domains (Dell & Boe, 2000). However, similarities existed between the two groups in the overall risk domain and in community functioning, personal/emotional orientation and attitude. Dell and Boe (2000) conclude that "individuals differ due to their racialized experiences but they also resemble one another due to common life experiences" (p. 2).
The Canadian criminogenic needs 3 differ slightly from the New Zealand list . "Currently there is support that many of these dynamic risk predictors may be pertinent for the female population" but it is unclear which needs are paramount in terms of community adjustment (Law, 2004, p. 18). The researcher found that the employment and associates variables of the Community Intervention Scale were the strongest predictors of failure for females. Substance abuse and associates predicted violent re-offending significantly.
Blanchette (2002) lists the following commonly cited women-specific criminogenic needs in the personal/emotional domain: low self-esteem, childhood and adulthood personal victimization and self-injury/attempted suicide. However, some factors may not be criminogenic - though considered significant - because of their nebulous relationship with criminality: victimizations, history of abuse, attempts of self-harm, lack of education and employment skills, high rates of depression and mental illness, and dependency on welfare. "There is still a need for predictive and treatment outcome studies to determine the exact nature of the relationship between several need areas and criminality/recidivism for women" (Blanchette, 2001, p. 78). Some evidence supports that women and men have similar criminogenic needs such as substance abuse, family/marital problems, antisocial attitudes and antisocial associates whilst research data on employment/education and community functioning are equivocal. In addition, not mental ability and mental health problems but cognitive deficits such as problem-solving deficits and impulsivity (all considered part of the personal/emotional domain) have been identified as another predominant criminogenic need of female offenders (Blanchette, 2001).
Based on their literature review Moth and Hudson (1999) summarise the criminogenic needs specific to imprisoned female offenders and related to recidivism: being responsible for children (particularly as a single mother), financial problems, limited job skills and opportunities, current clinical depression and drug use and absence of a stable relationship. In young female offenders criminogenic need areas are antisocial peers and attitudes, lack of affiliation with pro-social agencies and people, educational problems, misconduct and minor personality variables.
At least one New Zealand study (Moth & Hudson, 1999) has studied the criminogenic needs of 37 women residing in Christchurch Women's prison 4. More than half of the participants reported low self-esteem, low mood, drug use, unemployment and financial problems in the six months prior to their offending. Many of the women in this study had past or present mental health issues. Sexual offences were rare in this group. The authors compared the scores on the needs areas as identified by the CNIA (Case Needs Identification and Analysis, now Offender Intake Assessment) with Canadian studies of female offenders and concluded that the New Zealand women had higher needs in all areas, but mainly in employment/education, community functioning, substance abuse and attitudes. Further, women with a higher security classification had greater needs in the domains of education/employment, associates and attitudes. Compared to non-offender populations factors such as abuse as a child, behavioural and academic problems at school, early departure from school, long periods of unemployment, dependence on welfare and sources of illegal income and a lack of job skills were overrepresented. Further, accommodation problems and few pro-social connections occurred prior to their offending. The authors also mentioned difficulties in relationships, in particular with problem solving and conflict and establishing intimacy. Higher scores on the Level of Service- Revised (LSI-R) were associated with higher security classifications. The authors hypothesised that "the reporting of low mood, financial problems and low self esteem issues may all be important variables in case formulation approach to understanding the initiating causes of offending. … [T]he consistent use of psycho-active substances and reported mood difficulties may reflect generalised problems in affect regulation that may be central to the offending process" (Moth & Hudson, 1999, p. 64-65). Pascoe (n.d.) studied the link between the identified areas of need and reconviction data for this sample of female offenders. Of the 29 women from the original sample released since March 1999, 10 had been reconvicted. The small number prevented statistical analyses. The majority of reconvictions were for dishonesty offences and the majority resulted in custodial sentences. The number of days out of prison varied between 30 and 777 days. Of the reconvicted women, 60% identified as Maori, responsible for nearly 58% of the reconvictions. "There was no significant difference indicated between the areas of need identified for the original sample and the 29 women who were subsequently released" (Pascoe, n.d., p. 26) and similarly for differences between the original sample and the 10 reconvicted women. Pascoe (n.d.) concludes that which CNI criminogenic needs weigh more heavily for women than for men may be a more relevant question than what female criminogenic needs are.
Byrne and Howells (2000) claim that female offenders have general needs that mainly relate to psychiatric and psychological problems (hence adequate psychiatric screening on entry is paramount) but note the tendency to pathologise female offending. Further, some non-criminogenic needs in male offenders may be criminogenic needs in female offenders (Hannah-Moffat, 1999; Hart, 2000) associated with women's backgrounds, life circumstances and different experiences of environmental, situational, political, cultural and social factors. Hannah-Moffat (1999) claims that 'the recent redefinition of needs as risks in the correctional sphere emerges from a desire to improve predictive capacities for both male and female prisoners" (p. 84). The hybrid term risk/need is in vogue and implies that both terms are indistinguishable. The correctional management of women classified as "high need" is little different from women classified as "high risk". Characteristics of women previously considered needs 5 such as history of abuse, history of self-injury, single motherhood, mental health concerns and dependency on welfare have become criminogenic factors, "risk factors that can predict recidivism" (p. 86), justifying specific interventions and management strategies by prison staff. Hannah-Moffat (1999) argues that risk in terms of institutional adjustment, escape and public safety is gendered. She reminds the reader that behaviour in prison and outside is not strongly linked as many assume (Shaw, 1991 and Loucks, 1995 as cited in Hannah-Moffat, 1999).
The problem of substance abuse - in itself a complex concept related to severity, frequency, context and time frame in relation to the offending - among female offenders is serious (Blanchette, 1997; Byrne & Howells, 2000) as are the high incidence rates of abuse (Byrne & Howells, 2000). Dowden and Blanchette (1999) report that substance abuse impacts on a female offender's ability "to make rational prosocial choices and likely contributes to certain at-risk individuals" criminal behaviour" (p. 9). The women they studied had multiple criminogenic needs compared with non-substance abusers (replicated in a later study by Jones, 2004) and addiction treatment appeared to have a positive effect on recidivism figures, providing "optimistic preliminary evidence for the effectiveness of institutional substance abuse treatment for female offenders" (Dowden & Blanchette, 1999, p. 30). Their findings also indicate that more substance abusers had violent offences in their criminal history or index offences. "Perhaps, violence in women offenders may be mediated by a substance abuse problem" (Dowden & Blanchette, 1999, p. 31). The same authors note that substance abusing female offenders had more needs in a variety of problem areas (not all related to the substance abuse), such as poor stress management, low frustration tolerance and thrill-seeking behaviour, which offers an additional explanation for the significant more violent offences by substance abusers. Byrne and Howells (2000) suggest that offending behaviour could be a product of PTSD or of the substance abuse as a coping mechanism. They report that "there is evidence that treatment of abuse sequelae can reduce reoffending" (Byrne & Howells, 2000, p. 5). In HoItfreter and Morash's 2003 study substance abuse was found to be very strongly connected to high risk of re-offending, whilst emotional stability/mental health and criminal associates had a moderate connection. Blanchette (1997) found in her study of 182 federal female offenders (58% designated as violent and 42% as non-violent) that non-violent female offenders tended to have more criminal associates than violent female offenders, contradicting earlier research. Overall, violent female offenders presented higher levels of need than non-violent female offenders.
Zaplin (1998) noted that there is no definite causal link between child abuse and female crime but clinicians have noticed that the majority of female offenders have a long history of child abuse. Chesney-Lind and Pasko (2004) argue that the link between female offending and women's (childhood) victimization experiences in particular with respect to race and gender is increasingly clear (also Blanchette, 2001). McLellan, Farabee and Crouch (1997) studied the relative victimization of 1030 adult male inmates and 500 adult female prisoners in Texas. Their findings support the suggestion that women react more with self-blame and depression to victimizations than males, that this continues from childhood into adulthood and hence they become more vulnerable to substance abuse. "The severity of substance misuse and problems associated with it are stronger predictors of female rates of criminal activity than male rates" (McLellan, Farabee, & Crouch, 1997, p. 455) but problematic drug use was less predictive of violent crimes. Koons, Burrow, Morash and Bynum (1997) also point to the mediating role of victimization experiences by women, leading to mental health problems and substance abuse and further affecting criminality (see also Morash, Bynum, & Koons, 1998). Loucks and Zamble (2000) found that drug abuse did not correlate significantly with general recidivism and only weakly with violent offending. However, substance abuse in the family (in particular by the father) "is important in the prediction of recidivism" (Loucks & Zamble, 2000, p. 35).
In Dowden and Andrews" (1999) meta-analysis interpersonal criminogenic needs (in particular family process variables but also antisocial associates) were the strongest predictors of treatment success for female offenders and not substance abuse and basic education skills as identified in previous studies. "Personal and interpersonal noncriminogenic needs were not related to treatment outcome" (Dowden & Andrews, 1999, p. 449) but associated with recidivism increases (!). Holtfreter and Morash (2003) comment that the 1999 Dowden & Andrews study did "not provide evidence that there were no gender-related differences in specific needs that programs addressed for women and for men" (p. 140). The authors investigated this further with 402 female felony offenders. They included the LSI-R because "previous work … indicates the LSI is a highly reliable measure of risk for recidivism for women offenders" (p. 144). They found three clusters: low need/average risk (cluster 1), high need/high risk (cluster 2) and average need/low risk (cluster 3), arguing that women with lower risk should not be excluded from programmes as "they may benefit most" (p. 152) from programmes that target specific needs. In stating this, the authors question the risk principle in which higher risk offenders should receive the most intensive treatment.
Dowden and Andrews (1999) report that it remains unclear whether past victimization and self-esteem issues are criminogenic or noncriminogenic needs for female offenders whilst they "are promising targets for change" (Dowden & Andrews, 1999, p. 449; Bonta, Pang, & Wallace-Capretta, 1995). Lowenkamp, Holsinger and Latessa (2001) investigated the role of abuse in risk prediction with female offenders but had to conclude that childhood abuse did not significantly impact on criminogenic risk and actuarial risk of re-offending (i.e., re-imprisonment). "The risk factors for men and women remain the same, however, the form by which these factors are measured may differ" (Lowenkamp, Holsinger, & Latessa, 2001, p. 560). They used the LSI-R and confirmed its validity for both females and males. A cut-off score of 12 on the LSI appears to predict recidivism for female offenders (Howells, 2000; Moth & Hudson, 1999).
The question has been put forward whether it is adequate to classify female offender needs into criminogenic and noncriminogenic for correctional treatment (Monster & Micucci, 2005). In a study of 27 incarcerated women in a Canadian facility, Monster and Micucci (2005) identified education, vocational training and specialised programmes in respect of substance abuse, financial management and sex offending as the major criminogenic needs. The non-criminogenic needs (access to better health care and maintaining more frequent and positive contacts with their families and significant others) were perceived as more important by the inmates than their criminogenic needs, which supports the philosophical principles of the enhancement model.
Concluding, the Criminogenic Needs Inventory (CNI) does not include criminogenic needs specific to women. Based on international research the suggestion has been made to include "personal and emotional adjustment" (including low self esteem, childhood and adulthood personal victimisation and self-injury/attempted suicide) as an additional criminogenic need for women. It is hypothesised that Maori female offenders may have more and higher needs in several need domains than New Zealand European female offenders similar to overseas studies with Aboriginal female offenders.
Further studies are needed to determine the exact relationship between need areas and offending/recidivism for women. Some evidence supports that men and women share some criminogenic needs but research data on other criminogenic needs and non-criminogenic needs are equivocal. There is a tendency to pathologise female offending and psychiatric and psychological problems have been found to be significant in female offenders, but it remains unclear whether such problems are criminogenic for women. From the literature it is inferred that females struggle with anger or poor coping skills in interactions and conflicts.
Substance abuse (an adaptive mechanism for coping with PTSD and victimisation experiences) is recognised as a serious problem among female offenders. There is evidence that substance abuse treatment is effective in reducing re-offending although there is inconsistency among researchers whether substance abuse (particularly drug abuse) predicts general offending, recidivism and violent offending in females. Collectively, research has shown that females, who offended violently, belong to an ethnic minority or have a higher security classification and have higher levels of need than their counterparts.
The terms risk and need have become indistinguishable terms in research literature on female offenders and the question has been put forward whether it is adequate to classify female offender needs into criminogenic and non-criminogenic. It appears important to include women's perception of which needs they consider important as they may differ significantly from their identified criminogenic needs. From the literature review, treatment appears to be tailored to women's needs rather than their risk.
1 New Zealand criminogenic needs combined for both offending and pre-disposing period are: alcohol and drug, criminal associates, lifestyle balance, violence propensity, relationships, risk-taking arousal, offence related cognitions and emotions, gambling, sexual arousal, psychiatric disorder and organic disorder.
2 Maori Culture Related Needs are: limited or lack of whanau contact, whanau-related stress, whanau social influence to crime, whakawhanaunga, cultural tension and cultural identity.
3 The most pertinent criminogenic needs used for risk prediction in (Canadian) male offenders are: attitudes, education/employment, substance abuse, family/marital relationships, associates/social interaction, community functioning and personal/emotional orientation.
4 For a profile, see page 11-12.
5 That these characteristics are needs is questionable. It is suggested that for women, needs such as safety, secure attachment, stability, coping and the like are implied in these characteristics.