Desistance & Offender Engagement

Key sources

Desistance research can be seen as one part (the part most closely associated with the rehabilitative work of Probation[1]) of a wider ‘criminal careers’ approach. Longitudinal studies, following up individuals over the course of their lives, and using statistical modelling techniques, have been one well-established method[2] to investigate onset, duration and desistance. More recently qualitative studies and narrative analysis of offenders’ stories have drawn parallels with findings from the mental health field about recovery from depression, and prompted new theorising about effective intervention.

The ‘good lives model’ (GLM)[3] is one theoretical approach to desistance. As well as recent research, it draws on experience in sex offender treatment, and wider psychological theories of motivation. GLM has gained popularity following implementation problems found with the ‘what works’ approach to interventions. Andrews and Bonta et al. (2011)[4] have recently pointed out that its current popularity may have more to do with implementation failures of the Risk Needs Responsivity (RNR) model – to which it is sometimes counterposed – than to theoretical differences.

Defining ‘key’ sources is difficult for this diffuse[5] and sometimes densely theoretical topic. There are some key empirical studies:

  • The Cambridge Study in Delinquent Behaviour[6], associated with the work of David Farrington and others
  • Maruna’s Liverpool Desistance Study[7] a narrative analysis of offenders’ accounts of their lives
  • Farrall’s study of 199 probationers[8], followed up over 4 interviews since the late 1990s
  • Rex’s 2002 research into probationer perceptions’ of offender-officer relationships[9]

Reviews and discussions of key theoretical ideas and findings can be found in[10]:

  • Piquero, Farrington and Blumstein (2007) Key issues in Criminal Career Research[11]: New Analyses of the Cambridge Study in Delinquent Development[12]
  • Ward and Maruna (2007) Rehabilitation[13]
  • McNeill (2006) A desistance paradigm for offender management[14]; McNeill (2009) Towards effective practice in offender supervision[15]; and McNeill and Weaver (2010) Changing Lives? Desistance Research and Offender Management[16]
  • Porporino (2010) Bringing sense and sensibility to corrections: From programmes to ‘fix’ offenders to services to support desistance[17]
  • Kazemian (2007) Desistance from Crime: Theoretical, empirical, methodological, and policy considerations[18]
  • Andrews and Bonta et al. (2011) The Risk-Need-Responsivity (RNR) Model: Does adding the Good Lives Model Contribute to Effective Crime Prevention?[19]

NOMS’ recent change of focus towards ‘offender engagement’ is supported by a collection of research summaries and commentaries, and a series of ‘Research Bulletins’:

  • NOMS, 1st edition (2010) An initial summary of evidence gathered to prove the hypothesis that the one-to-one relationship between the offender and practitioner can be a powerful means of changing behaviour, and therefore reducing re-offending[20]
  • NOMS Offender Engagement Research Bulletins[21]

The following websites provide access to NOMS documents and other professional development materials:

  • Essex Probation Trust Resources
  • Midlands Probation Training Consortium ‘Thinking in Practice’ seminars –
  • Discovering Desistance ESRC Knowledge Exchange Project –

Key  points

1)    Desistance research complements other approaches. It is in the same tradition of renewed belief in the potential effectiveness of rehabilitation as the risk-needs-responsivity-based ‘what works’ movement which followed the ‘nothing works’ era. Desistance research is oriented towards reducing reoffending, although from a more positive recognition of offender strengths than RNR, which has been described as a deficit-oriented approach. While some commentators suggest an opposition with ‘what works’ research[22], others see it as already presaged in much of the ‘what works’ literature[23].

2)    Desistance research draws on the ‘criminal careers paradigm’. This means that it concerns changes that happen to individuals through their lives. However, desistance research focuses on the aspect most relevant to the work of Probation – stopping offending, or offending less often, or less seriously. There is a well-known finding that crime peaks in the early teens and drops off after that (the ‘crime-age curve’). The desistance perspective maintains that we need to understand this process in order to help offenders towards giving up crime. McNeill describes this as complementary to ‘what works’ evaluation in the same way as developmental educational psychology is complementary to evaluation of teaching methods.

3)    Limitations of risk-based approach[24] The following points have been made:

  • R-N-R[25] approach is grounded in the idea of risk to society, and has been said to neglect wider service values.
  • By neglecting broader human needs (as opposed to criminogenic need), it neglects motivation and identity as key elements in change.
  • It focuses on general responsivity, while neglecting individual responsivity.

4)    The ‘Good Lives Model’:[26]

  • was developed from the research insights from Maruna’s desistance study, and Ward’s theory rooted in sex offender treatment practice and ‘positive psychology’;
  • recognises the strengths of risk-based approach, including overall evidence of its effectiveness in reducing reoffending;
  • builds on the problems and issues found in ‘what works’ implementation studies;
  • sees targeting of risk as necessary, but not sufficient;
  • advocates an individualised approach to assessment and treatment;
  • stresses the possibility of ‘redemption’.
  • GLM suggests:
    • harnessing natural motivation towards legitimate ‘approach goals’[27] in the context of an overall pro-social life-plan
    • going beyond tackling risk factors, towards a ‘holistic reconstruction of the self’
    • basing intervention on offenders’ strengths
    • basing intervention in understanding aetiology[28] of offending
    • balancing focus on offender goods and happiness, with management of risk.
    • interventions that combine programmes with work on family ties and associates, as well as advocacy to access resources and promote inclusion, are likely to be most effective.

5)    Motivating offenders. Offenders share an essential humanity that is motivated towards primary human goals of personal autonomy; relatedness to others; and competence[29]. Interventions should focus on pro-social ways of meeting desired goals, rather than addressing criminogenic need. The officer-offender relationship (the ‘working alliance) is a potential tool[30] in effecting change through engagement of motivation. Compliance will depend on offender’s perception of legitimacy, which in turn rests on engagement and relationships.[31]

6)    Listening to offenders is key to understanding why they give up crime. There have been some qualitative studies providing new insights into: ways of talking about desistance (‘redemption narratives’)[32]; offender motivation, officer-offender relationships, and the zigzag nature of giving up offending[33]

7)    Building social as well as human capital Theorists have differed in the emphasis they place on individual psychological factors (including the stories they tell about themselves), and the wider social context.
8)    Criticisms of desistance approach

  • Theory

o   The concept of desistance is not well defined[34]

o   GLM makes several proposals for practice, but these are not based in a fully developed theory[35].

  • Evidence base

o   The desistance approach focuses on developmental process rather than intervention effectiveness[36].

o   Specific hypotheses need more empirical support. While qualitative approaches have provided valuable new insights, or supported practitioner experience, issues of cause and effect – and how widely findings can be applied – remain.

o   Cost-benefit analysis can only be based on quantitative outcome data[37].

  • Specifically, there is a lack of current evidence on:[38]

o   diversity in views of ‘the good life’, and the general applicability of ‘human goods’;

o   which offenders really need ‘holistic reconstruction of the self’;

o   whether anti-social ‘good lives plans’ might work quite well for some offenders, and thus work to support offending; and

o   whether an individual psychology focussed approach (which is common to both R-N-R and GLM) is sufficient without wider social intervention.

  • Challenges for case management practice

o   Finding the resources to intervene on many levels

o   Individualising treatment could lead to loss of focus and overall service effectiveness.

o   Problem of balancing individualised treatment with structured systematic intervention.

o   Over enthusiastic implementation of a new wave of programmes could face the same obstacles as ‘what works’ programmes.

9)    NOMS Offender Engagement Programme

  • Sometimes presented as a research project to test the effectiveness of one-to-one supervision, which draws on desistance literature[39]
  • Driven by the need to develop and share best practice, as well as the coalition government aim to reduce central bureaucracy[40] and measure outcomes
  • Central feature is practitioner professional development and support, to underpin greater professional autonomy
  • Key projects are: SEED (Skills for Effective Engagement and Development); RSM (Reflective Supervision Model); and SPA (Sentence Planning Approach)
  • OEP is supported by a programme of research, including local and small-scale work reported in its Research Bulletins[41]

In the pipeline

External research commissioned by NOMS OEP is being conducted by the University of Sheffield and by a partnership between Leicester University and the Institute for Criminal Policy Research (ICPR), now at Birkbeck College, University of London.

The Offender Engagement Project at the University of Leicester is using the Offender Management Feedback Questionnaire (combined with a version for Offender Managers) to investigate relationships between:

  • offenders’ and probation  practitioners’ views measured by OMFQ
  • OMFQ and intermediate measures of success
  • OMFQ and re-offending.

The OMSAS cohort survey of offenders in the community may offer useful findings on offender experiences and opportunities to statistically model how different combinations of intervention combine to predict outcomes. The research comprises a survey conducted by NatCen and a dataset, including case management information, analysed by Matrix Knowledge. A report is due in April 2013.

The ESRC funded Desistance Knowledge Exchange Project (Discovering Desistance)[42]  is developing a documentary and a series of practitioner conferences.


  • [1] Criminal careers research is as much (if not more often) about onset and persistence, as desistance.
  • [2] For example, Laub and Sampson (2003), following up the work of the Glueck’s from the 1950s, or the work of Farrington and others on the Cambridge Study in Delinquent Development, which began in 1961, with a series of reports from the mid 1980s.
  • [3] Ward and Maruna (2007)
  • [4]
  • [5] In its broadest sense, desistance research could be seen as just another term for effectiveness research, or ‘what works’. Use of the term ‘desistance’ as opposed to reoffending, signals a more positive approach, as well as a concern for process, practice, and the realities of offenders’ lives.
  • [6] The dataset for this study has supported a large number of published analyses, and is available for researchers through the US Inter-University Consortium for Political and Social Research;jsessionid=CF1A49DD72CF33CC3D67D2CFD65963CD  or the UK Data Archive . It was a prospective longitudinal survey of 411 South London boys from a largely white working class area, beginning in 1961-2 when they were 8-9 years old and continuing to age 40. Data is based on interviews with the boys, their families and teachers, as well as case information including convictions.
  • [7] See Maruna and Porter et al. (2004)   and Maruna (2004)
  • [8] Farrall (2002) and Farrall and Calverly (2006)
  • [9]
  • [10] This is a selection. There are many other relevant sources, some of which are listed in the spreadsheet accompanying this Briefing.
  • [11] Farrington and others use the term ‘development and life-course criminology’ (DLC). This covers a wider range of concerns than intervention to support desistance.
  • [12] Farrington provides an overview of criminal careers research, analysis of 5 different trajectories of onset, persistence and desistance, and an agenda for future research. Further analysis is expected in Farrington’s forthcoming book.
  • [13]
  • [14] The argument is presented more briefly in a CJScotland paper:
  • [15]
  • [16]
  • [17] Porporino reviews the ‘what works’ evidence and scans desistance theory for promising new directions. This is a chapter in Brayford et al. (eds.) (forthcoming) What Else Works?Creative Work with offenders  This book also contains a chapter by Weaver and McNeill on desistance theory.
  • [18] Kazemian’s account of the state of current research highlights a range of unresolved problems, such as the lack of an agreed definition of desistance, and the need to replicate findings on different groups.
  • [19]
  • [20]
  • [21]
  • [22] See for example the editorial to Probation Journal 56: 2
  • [23] See Porporino (2010) and Andrews and Bonta (2011)
  • [24] These points, and the account of GLM (good lives model) draws heavily on McNeill’s (2009) account, which in turn draws on Ward and Maruna (2007).
  • [25] The ‘risk-needs-responsivity’ model is mainly associated with the work  of Canadian correctional psychologists, such as Andrews and Bonta (2003). As implemented in England and Wales, it became a highly centralised approach to assessment and intervention, which was (arguably) partially successful.
  • [26] Again, this is based on McNeill’s account, particularly McNeill (2009).
  • [27] Mann et al. (2004) reported the greater effectiveness of ‘approach goals’ compared to ‘avoidance goals’ in work with sex offenders. However, these terms have wider application in the developing field of ‘positive psychology’, which concerns optimising health and happiness, as opposed to the frequent focus of psychology professions of illness and deficit. The work of Martin Seligman (known for the idea of ‘learned helplessness’) has been influential. Positive psychology owes much to ‘humanistic psychology’ and the Abraham Maslow’s well-known concept of the hierarchy of human needs.
  • [28] Aetiology means either causes or reasons, and is a term widely used in medicine. Preferring to use the term ‘aetiology’ over ’cause’ signals a certain position on the age-old line of  tension within the human sciences between the machine metaphor (which prefers causes) and purposive or agency models (which prefer reasons).
  • [29] This is one of the key tenets of the good lives model, developed in Ward and Maruna (2007) Rehabilitation.
  • [30] Ansbro (2008)  provides another account of the mechanisms whereby relationships might affect change. Her account is rooted in attachment theory.
  • [31] Farrell (2002) found that the supervisor-probationer relationship was less important than other relationships. Overcoming obstacles to change seemed to be more closely associated with offender’s own motivation and supportive social context, than with the relationship with the officer. This suggests that a sense of self-efficacy is closely connected with change.
  • [32] Maruna presents a systematic content analysis of modes of causal attribution in offenders’ talk about their lives.
  • [33] See Farrall (2002) and Rex (2002)
  • [34] See Kazemian (2007).
  • [35] See McNeill’s account.
  • [36] However, the OEP aims to assess the effectiveness of interventions that incorporate learning from desistance research (e.g. SEED).
  • [37] However, even for quantitative studies, clear and agreed methodologies of benefit-cost analysis have proved elusive.
  • [38] This list is derived from McNeill (2009).
  • [39] NOMS Offender Engagement Programme aims test the hypothesis that the one to one relationship between the offender and supervisor can be a powerful vehicle for change. See Offender Engagement Programme Overview, 2011 . NOMS Business Plan 2011-12 (p.12) suggests that the impact of the OEP  on  reducing re-offending is already known.
  • [40] The much quoted statistic that probation staff spend only 24% of their time in direct contact with offenders appears to arise from an unpublished source.
  • [41]
  • [42]