Developing Models To Strengthen Cultural Perspectives

In Order To Reduce Recidivism


Leon Bakker & Kristen Maynard, Department of Corrections, New Zealand




An Overview


Developing models for Maori involves understanding generally who is at risk and why and then determining what is different for Maori.  It is most likely that many of the reasons why offenders commit crime, and what is needed to reduce recidivism, are common to all ethnic groups.  However, there are also likely to be issues unique to specific sub-groups of offenders such as Maori.  This paper will start by briefly discussing what is known and constitutes best practice for offenders generally.  Limitations of the existing models’ application to New Zealand will then be described.  The models we have developed and how we are specifically strengthening cultural perspectives for Maori will then be provided.



Principles of Best Practice


A number of principles have been found to be associated consistently with effective intervention in a correctional setting.  Among the most important of these are the principles of risk, need, and responsivity (e.g. Andrews and Bonta, 1994; Gendreau, 1996).  The risk principle holds that intervention is most effective when targeted towards those individuals who have the greatest risk of further criminal offending.  The needs principle asserts that there are certain aspects of an individual’s functioning – ­substance-abuse, criminal attitudes and associates, and the like - which should be targeted for intervention, to bring about a reduction in down-stream offending (Bonta, 1996).  A key feature of these aspects is their dynamic nature.  This means that they are potentially changeable.  The responsivity principle states that offenders will be most affected by interventions that are matched to their learning style, language, culture etc.



Risk Principle


Targeting those most at risk increases the impact of treatment resources.  Because some appraisal of risk is central to almost every decision that is made about an offender within the criminal justice system, this area has been intensively investigated over the last seventy years.  Research on the prediction of future criminal behavior has increased substantially in volume over the last twenty-five years and these investigations have become progressively more sophisticated, reflecting developments in the field of statistical methodology and also the ability of computers to store and manipulate increasingly large arrays of data. 


Although the various scales that have been developed to predict recidivism are still imperfect, there is a single most significant conclusion that may be drawn from this research.  This is that statistical or actuarial scales derived for the purposes of predicting future criminal risk consistently out-perform the judgements of experts in almost every investigation which has compared these two approaches to risk assessment.  This situation appears to prevail irrespective of the level of experience and professional training of the persons making clinical judgements, and risk scales consistently do better than the judgements of social workers, experienced correctional officers, parole boards, psychologists and psychiatrists. (Gottfredson and Gottfredson, 1988)


In a significant publication, Gottfredson and Gottfredson (1994) reviewed their own work of 30 years in the area of prediction and that of others, and struck a rather pessimistic tone.  They concluded that the best available predictors are still quite poor. Indeed, the most sophisticated statistical measures may produce predictions that are no better than simple ones, and in some respects may be worse.


Even with these limitations, however, predictions made with the use of statistical devices out-perform those made without such help. To this list of problems can be added limitations to existing devices such as a lack of validation, use of risk devices outside of the country they have been developed in, lack of training of staff etc.  There are no risk devices that have been developed and validated specifically for New Zealand offenders.



Development of New Zealand Risk of Re-Conviction (RoC) Models


In 1993 we began developing risk models specific to New Zealand.  Using 130,000 criminal histories obtained from the Law Enforcement System we used logistic regression to develop statistical models based on the criminal history available.  These models predicted a number of events that we were interested in.  Particularly, re-conviction for any offence in the following five years, the risk that such re-conviction would lead to imprisonment, the length of such imprisonment and how serious the offence would be.


Figure 1 provides an indication of the accuracy of the re-conviction model.  The more closely the dots on the graph are to the dashed trend line the more accurate the model.  Clearly the correspondence of the model is very good.  While ethnicity is a variable that is used in this model it is not a particularly important one; removing it reduces the accuracy of the model by only 0.4 percent.





There are limitations to this approach to evaluating risk.  First the models still are not 100% accurate - they do better than human judgement but there is still much room for improvement.  Second the models use “tombstone” predictors; the variables cannot be readily changed.  Such variables as number of previous convictions, length of time spent in prison, are static or unchangeable - these models will not reflect changes in risk due to treatment.  Another limitation of these models is that they do not tell us why an offender is at risk but only who is at risk.  Ideally, we would want to know what puts a person at risk so that they can be given appropriate treatment.  If we want to develop models that help reduce reoffending in general and Maori in particular, we need additional information.





Criminogenic Needs and Risk


To address some of the limitations of static risk prediction,  researchers have focused on dynamic or changeable features of offenders linked to recidivism.  Variables such as unemployment, lack of education, lack of financial skills, relationship problems etc. are all examples of dynamic variables that are linked to offending and can be targeted for treatment.  Such instruments cater for both the risk principle and the criminogenic needs principle cited earlier.


An example of an instrument that uses a combination of static and dynamic factors to estimate risk, and which identifies needs that can be treated, is the Level of Service Inventory - Revised (LSI-R) developed by Andrews and Bonta of Canada.  Even though this instrument is widely used and has good correlation with recidivism it too has limitations as a device for New Zealand.


1.         A number of potential criminogenic needs identified by the psychological research

       literature are not measured by the LSI-R.


2.       The LSI-R was developed using a Canadian offender population.  Canadian offenders may differ from their New Zealand counterparts in ways that reduce the accuracy of the LSI-R when used with the local offender population.  There is some evidence that various measures of criminogenic needs have decreased validity when applied to different offender populations (Loza and Simourd, 1994).


3.       The LSI-R does not measure potential criminogenic needs uniquely related to Maori.  This is a major limitation given that such criminogenic needs may help explain the high rates of offending and re-offending by Maori.  The development and validation of Maori culture related needs would greatly enhance the Department’s strategic goal of reducing re-offending among Maori.


4.         The LSI-R measures the severity of criminogenic needs based predominantly upon

estimates of risk.  Because the estimate of risk is so affected by static factors this reduces the LSI-R’s measurement sensitivity to genuine changes in the dynamic dimensions of its criminogenic needs.


5.       The LSI-R, and other established criminogenic need inventories, was developed based upon the assumption that any assessed offender need is automatically criminogenic.  This feature is likely to result in the LSI-R being overly inclusive in its assessment of criminogenic needs.  It may help explain the low to moderate statistical relationship found between the criminogenic needs measured by inventories such as the LSI-R and static risk prediction measures.




Developing a New Zealand Criminogenic Needs instrument


Given the above limitations, we considered it desirable to develop our own Criminogenic Needs instrument which tackled the shortcomings of existing devices and which complemented the risk models that had been developed.  The Criminogenic Needs Inventory (CNI) was developed using solely dynamic needs that were psychological in nature (needs such as offence related emotions and cognitions, violence propensity, risk-taking arousal etc).  Needs are assessed with the underlying premise that events (cognitions, behaviors and emotions) that occur closest to the offence are most important to understanding what caused the offending.  Events further back in time could be seen to predispose the person to committing offences but are not as critical to the offending event itself.


Our initial piloting has been very promising suggesting that the CNI can be used reliably and is well correlated with existing needs and risk measures.  It also appears to be related to reconviction risk in a way that differs from the LSI-R and other current risk needs measures.  Critical to the development of the CNI was the measurement of needs that might be unique to Maori and that might help explain why Maori constitute such a disproportionately large sub-group of offenders.




Maori Culture-Related Needs: Background


Maori constitute almost half of the New Zealand offender population (Spier 1998; New Zealand Community Probation Service 1998).  The negative effects of the process of colonisation and the associated erosion of traditional cultural values offers a broad basis for understanding the possible over-representation of Maori in the criminal justice system.


A recent study (Maynard 1998) identified a number of possible cultural factors that were likely to contribute to the offending behavior of a Maori individual and their ability to seek to modify this behavior.  In order to increase the validity of risk/need prediction amongst Maori, it was identified as imperative in that study, that these cultural differences be adequately recognised and provided for.


Experience in other sectors (such as Health), has also shown that there are specific cultural factors unique to Maori which can influence the effectiveness of treatment (Te Pumanawa Hauora, 1995).  Further, studies undertaken by Maori research groups (namely Te Hoe Nuku Roa at Massey University) have demonstrated that there is a correlation between cultural identity and socio-economic circumstances (Te Hoe Nuku Roa, 1996).


In addition, a number of evaluations (on predominantly tikanga Maori based initiatives and programmes) have been undertaken or contracted by the Department of Corrections.  Although these evaluations are predominantly ‘process’ and ‘formative’ orientated rather than focused on ‘outcomes’, they are consistent in highlighting that a tikanga or kaupapa Maori approach to programme content and delivery can have positive effects on Maori offenders.  It appears that all of these programmes focus (to some extent) on addressing specific cultural needs such as:


·       Fostering a positive cultural identity,


·       Creating an environment which enables individuals to feel both safe and good about being Maori; and


·       Promoting whakawhanaungatanga.


These key aspects of programme intervention have provided early indications of success in changing Maori offenders attitudes and behaviours, promoting pro-social lifestyle changes, and increasing their receptiveness to other rehabilitative programmes.  These are also wholly consistent with the identification of specific and unique Maori culture-related need.




Maori Culture-Related Need (“MACRN”)


Maori culture-related need has developed on the basis that there are unique needs specific to Maori offenders which 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.  It recognises that there are certain factors or needs additional to generic offending needs, which require recognition and incorporation into risk/need assessment tools in order to effectively identify and address offending and re-offending by Maori.


The Department of Corrections is currently developing a New Zealand based need assessment tool - Criminogenic Needs Inventory (“CNI”).  A significant component of this tool will be focussing on assessing the culture-related needs of Maori offenders and how these are linked to offending behaviour.  Four potential MaCRN’s have been identified so far.  These are:


·       Cultural identity;


·       Cultural tension;


·       Whanau; and


·       Whakawhanaungatanga.


These have yet to be refined, and may be modified in the piloting phases of the CNI. The measurement of each MACRN may also require ongoing refinement.


Further, determining the extent and severity of each MACRN and recognising the interrelated nature of these needs, is necessary to ensure that a more complete picture is formed of the severity and complexity of the Maori individual’s offending behaviour and how best to go about addressing this.



MACRN 1: Cultural Identity


Promoting and enhancing a positive pro-social cultural identity has been identified as an important and relevant culture-related need by a number of sectors and research groups. Research undertaken and reported in 1996 by Te Hoe Nuku Roa at Massey University, demonstrated the particular significance of cultural identity as a policy variable.  Amongst its conclusions the report found that Maori generally wished to retain a Maori cultural identity and that there was a relationship between a secure identity and other areas of social and economic well-being.  Access to Maori institutions and knowledge was also recognised as a key determinant for Maori well-being (Te Hoe Nuku Roa 1996).


Further, based on an analysis of the information gathered, the report suggested that a secure cultural identity has advantages far beyond cultural affirmation.  It may for example afford some protection against poor health, it is more likely to be associated with educational participation and with positive employment profiles.  The corollary is that reduced access to the Maori world may be associated with social and economic disadvantage (Durie 1996).


Raising a Maori offender’s understanding and appreciation of Maori culture and promoting a positive pro-social Maori cultural identity could give that offender enough self-esteem and self-awareness to work toward changing their offending behaviour.


Consequently, it is imperative to assess whether the offender has a positive pro-social perception about being Maori or whether their perception about being Maori is negative and/or anti-social.  Once this is identified, the extent to which this perception impacts on their functioning can then be examined.  This would assist in identifying whether cultural identity is a ‘need’ that if addressed, is likely to contribute to a change in offending behaviour.



MACRN 2: Cultural Tension


It appears that acculturation factors such as issues associated with colonisation, and situations where offenders have felt that being Maori has created some barriers or difficulties for them in their personal, social or occupational life are likely to impact significantly on their overall well-being and could also increase the risk of Maori indulging in criminal activity. If one regularly receives the message that one’s culture, language, and identity are unacceptable, the impact on one’s sense of security and self-esteem will be negative (Berry 1994; cited in Maynard 1998).  This in turn could lead to a greater risk of offending than those secure in their culture and identity.


Contemporary New Zealand society has developed primarily from Western European based norms, despite the fact that Maori are recognised as the tangata whenua of this country.  Consequently, Maori culture has been generally compromised and discouraged in the process of colonisation and it is therefore likely that a number of stressors and tensions associated with differences in cultural values and beliefs between Maori and non-Maori has developed.  Further, the lack of positive coping skills for dealing with such tension, may promote maladaptive responses which could include the development of related cognitions and behavioural patterns that are likely to increase the individual’s risk of re-offending. It is important to identify and effectively assess the nature and extent of the cultural tension and the offender’s capability of recognising and pro-socially addressing these where they impact on personal, social and/or occupational functioning.


MACRN 3: Whanau


An appreciation of the collective nature of Maori culture will lead to a better understanding of the uniqueness and importance of whanau to Maori and it’s possible contribution (whether positive or negative) to offending behaviour.


Metge (1976) argued that the concept of whanau and its underlying values is discernibly different to Pakeha perceptions of family and family values (Hirini 1997).  For example, Durie (1984) stated that the Pakeha ideal of being able to stand alone and be independent is actually an unhealthy position from a Maori perspective where intra-dependence is encouraged.  The relationship of an offender to their whanau is crucially important to assess.  An exclusive focus on factors directly related to individual offending without examining concurrent public and community practices, will not of itself reduce offending levels as it doesn’t address factors which reinforce that person’s offending behaviour.  Individual obligations must be balanced against the prevailing value of collective responsibility (Hamilton Community Corrections Office 1997).


Information about the strength or lack of whanau support is likely to indicate predisposing, precipitating, perpetuating or protecting factors related to the development and maintenance of offending behaviour and attitudes.  Where the offender has a strong, healthy relationship with their whanau, it is important to assess whether contact with their whanau has been maintained.  If not, it is then necessary to assess whether lack of contact has negatively impacted upon the functioning of the offender. Where anti-social behaviours are prevalent within the offenders whanau, it is imperative to assess whether these behaviours have predisposed, precipitated or maintained offending behaviour.


In both situations attempts should be made to access members of the extended whanau or significant others in order to strengthen, reconstruct and enhance positive family relationships and support.  Interventions which access community support are important for instilling traditional pro-social whanau values.  There is also a need to accommodate family participation within interventions where appropriate.




MACRN 4 Whakawhanaungatanga


Clearly related to the concept of whanau is whakawhanaungatanga which essentially generates a sense of family cohesion and co-operation and is a specific form of support for Maori.  It also generates observable behavioural processes through which whanau functioning is promoted and enhanced.  It will continue to be a concept of substantial value to Maori in terms of their own social well-being and development (Hirini 1997).


When assessing the relationship an offender has with their whanau, the focus is on establishing the nature of the relationship and the consequent impact the whanau may have had on, or the lack of support that has contributed to, the offending behaviour of that individual.  Conversely, whakawhanaungatanga, focuses on identifying a variety of relationships an offender has with a group of people who they consider to be like a whanau to them and how these relationships may contribute to offending behaviour.  The group to which the individual associates with, does not necessarily have to be blood connected.


Focusing on whakawhanaungatanga and the nature of social relationships an individual forms with others, is important in that there appears to be an inclination for Maori (as a distinctly collective culture) to avoid isolation by seeking to form relationships with groups of people that they can relate to.  As such, there appears to be a strong need for Maori offenders to seek membership to a larger group which will provide that individual with a sense of belonging and collective responsibility.  For example, against the background of not aligning with mainstream society, becoming a member of a gang could provide a sense of importance and belonging that an individual may feel they lack in their lives.  Whakawhanaungatanga is therefore perceived as a unique need specific to Maori.  It further offers a broad explanation as to why many Maori offenders tend to form associations with anti-social gangs and develop anti-social behaviour where whanau support is lacking.


It is essential therefore to identify the nature of the relationship between the offender and their whanau, the impact the whanau may have on the offending behaviour of the individual (whether positive or negative), and the kinds of associates the offender tends to gravitate toward.  This would provide an indication of how these relationships are likely to contribute toward offending behaviour and could also identify important areas of need to address, such as strengthening whanau links, facilitating support for the whanau, and ensuring that the necessary pro-social support is in place for the offender.






Given the ground-breaking nature of measuring such needs it was not possible to include them in the initial pilot.  However, a reliability study including the MACRNs is planned to start in August.  There is evidence to suggest that there are a number of additional culture-related factors, that could contribute to the offending behaviour of Maori and their ability to seek to modify this behaviour.  This suggests that in addition to generic needs, Maori offenders require different types and approaches to interventions that also address a number of specific needs.


Consequently Maori culture-related need has emerged from this and will form a substantial component of the new needs assessment tool currently being developed by the Department of Corrections in New Zealand.  These additional unique Maori culture-related factors offer a broad explanation of a predisposition towards and maintenance of socially unacceptable behaviour and attitudes by Maori offenders.  Effectively recognising and addressing these areas of need will arguably assist with behavioural change.


Promoting and enhancing a positive pro-social cultural identity has been identified as an important and relevant culture-related need by a number of research groups, Department of Corrections’ evaluations, and other Government sectors.  This particular “need” does not only provide an effective target for treatment but also provides an indication of the types of interventions that could prove to be most effective for that individual.


It appears that acculturation factors such as issues associated with colonisation, and strategies employed to deal with cultural tensions, could impact significantly on the overall well-being of Maori offenders and are likely to contribute to offending behaviour.  In addition, the collective nature of Maori culture has been highlighted to illustrate the importance of whanau and whakawhanaungatanga to Maori in general and the influence it is likely to have on offending behaviour.  Including whanau and/or other support persons into interventions is considered necessary in order to enhance the effectiveness of interventions and to promote and sustain behavioural change.




Andrews, D.A. and Bonta, J. L. (1995) The Level of Service Inventory - Revised.  Users Manual MultiHealth Systems New York


Bakker, L., O’Malley, A.J. and Riley, D. (1999) Risk of Reconviction: Statistical Models which predict four types of re-offending. Department of Corrections NZ


Coebergh, B., Bakker, L., Anstiss, B., Maynard, K. and Percy, S. A Seein “I” to the Future: The

Criminogenic Needs Inventory (CNI) Department of Corrections NZ in press


Durie, Mason. (1996).  Maori Cultural Identity and its implications for Mental Health Services Department of Maori Studies, Massey University, Palmerston North.


Gottfredson M.R., and Gottfredson D.M. (1988).  Decision Making In Criminal Justice.  N Y, Plenum Press


GottfredsonS.D., and GottfredsonD.M. (1994). Behavioral prediction and the problem of incapacitation.  Criminology, 32, 441-474.


Hamilton Community Corrections Office. (1997).  The Community Programme Sentence: The Use of Sentencing Affirmatives for Maori Offenders - Prepared for Te Puni Kokiri.  Unpublished paper, Department of Corrections, Wellington.


Hirini, Paul. (1997).  He tirohanga ki te Ara Whanaungatanga: Looking at the Whanaungatanga concept. Chapter in Habermann, G. (Ed.), Looking back and moving forward. 50 years of New Zealand Psychology.  New Zealand Psychological Society, Wellington, 97,82-90.


Maynard, Kristen. (1998).  Whaia te Maaramatanga - Te Wahanga Tuatahi – An Analysis of the Cultural and Gender Appropriateness of the Risk Needs Assessment Tool Used by Community Probation Service.  Unpublished paper, Department of Corrections, Wellington.


New Zealand Community Probation Service.(1998). Re-offending over l2 month period after termination and Ethnicity of New Starts July-August 1998, as at 15 September 1998.  Unpublished paper, Department of Corrections, Wellington.


Spier, Philip. (1998). Conviction and Sentencing of Offenders in New Zealand: 1988 to l997.  Ministry of Justice, Wellington.


Te Hoe Nuku Roa. (1996).  Maori profiles - An integrated approach to policy and planning - A Report Prepared for Te Puni Kokiri.  Department of Maori Studies, Massey University, Palmerston North.


Te Pumanawa Hauora. (1995). Guidelines for purchasing personal Mental Health Services for Maori.  Department of Maori Studies, Massey University, Palmerston North.