Working with radicalised individuals: insights from a secondary and tertiary prevention program

ABSTRACT

Despite their widespread adoption, little is known about the content and implementation of case-managed programs targeting individuals at risk of radicalisation and known extremists. This study examined data on one state-based case-managed program implemented by police in Australia – referred to as Intervention 01. Drawing on case note information for a sample of clients and interviews with intervention staff and clients, data are presented on the radicalisation risk factors amongst program beneficiaries, client participation and intervention goals, forms of assistance, and family member involvement and support. Five client case studies are described outlining the experiences of program beneficiaries and lessons are highlighted relating to intervention delivery and the disengagement process. This study provides insights into the practical realities of delivering programs to counter violent extremism, focusing on factors that influence client change and facilitate and inhibit client progress. Implications for the implementation and evaluation of programs to assist radicalised individuals are considered.

KEYWORDS:

• Countering violent extremism
• case-managed interventions
• prevention
• evaluation
• program implementation
• disengagement
• deradicalisation

Introduction

Existing scholarship has drawn attention to the need to better understand the content and implementation of programs to counter violent extremism (CVE) (Costa et al., 2021; Horgan et al., 2020). The necessity for evaluation has also been emphasised (Baruch et al., 2018; Costa et al., 2021; Koehler, 2016; Nehlsen et al., 2020). While still small in comparison to other policy areas, there now exists an increasing number of studies that have aimed to assess the effectiveness of CVE interventions (examples include Cherney & Belton 2019; Hecker, 2021; Helmus, & Klein, 2018; Sahgal, & Kimaiyo, 2020; Webber, et al., 2018; Williams et al., 2016). However, the evaluation of CVE schemes is not without its challenges and requires a mix of quantitative and qualitative approaches (e.g. Malet, 2021; Nehlsen et al., 2020). An essential part of assessing the effectiveness of CVE programs is examining the vulnerabilities and risk factors amongst clients and what is done by practitioners to address these factors, including examining the challenges faced by staff and barriers that influence client progression. As Koehler and Fiebig (2019) have argued, we need to know more about the practical realities of delivering CVE interventions. This is particularly important when it comes to CVE schemes concerned with secondary and tertiary prevention, given they require staff to directly engage individuals at risk of radicalisation (secondary in focus) and known extremists or convicted terrorists (tertiary in focus). Case-managed interventions are an example of such an approach and have become common across many Western jurisdictions (Cherney & Belton 2019, 2021). Case-managed programs involve the development of tailored intervention plans and the use of multiple service providers to assist clients. To date, we know very little about the content of case-managed schemes, how they aim to facilitate disengagement and if they are effective.

This study provides results from research on case-managed programs in Australia targeting at-risk individuals and known extremists.¹ The results reported in this study are derived from an evaluation of one multiagency, case-managed intervention implemented by the police in an Australian state – which for confidentiality reasons is referred to as Intervention 01. Based on the collected data sources, a selection of qualitative and quantitative results is provided into the identified radicalisation behaviours amongst clients, their intervention goals, family member participation and support, and what has been done to assist the referred individuals. One of the key contributions of this study is to highlight the types of supports that can make a difference to how client’s progress on case-managed schemes, which relates to the informal support provided by intervention staff and assistance by family members. Likewise, lessons relating to the presence of residual risk and its impact on client progress are also highlighted and how these need to be considered when managing clients. The topic is applicable to broader questions regarding program effectiveness and knowledge on disengagement and reintegration processes.

To provide further insights into the dynamic nature of client progress, a series of client case studies are provided to examine the experiences of program beneficiaries and how disengagement was facilitated. To date, there are limited studies on these issues or research on the experiences of program beneficiaries and how referred individuals progress during CVE interventions (Horgan et al., 2020). Lessons relating to CVE theory and practise are explored in each case study. The results reported here have implications relating to conclusions about the effectiveness of Intervention 01 and similar case-managed schemes, however, the present study has not intended to evaluate the overall effectiveness of Intervention 01.² There are limitations with the data and the fact it is derived from case note information and interviews, raising issues related to bias, representativeness, and the generalisability of the results. The method is not underpinned by an experimental design, therefore, causation between program inputs and client outcomes cannot be demonstrated. The main goal of this paper is to provide data on some of the contextual conditions and factors that influence the potential effectiveness of case-managed schemes.

The next section provides detail on the administration of Intervention 01, the methodology and data sources. Data are then presented on the client sample, their intervention goals and contextual factors such as the involvement of family members. Next, a selection of case studies is provided on five clients, outlining the types of assistance delivered to these beneficiaries, levels of client progress and factors identified as inhibiting and facilitating client change and progress. Contextual data on client assistance and progress is then explored through interviews with intervention staff and clients. Implications of the findings are considered regarding the relevant literature and the task of implementing and evaluating CVE interventions.

A description of Intervention 01

Intervention 01 is part of a national diversionary scheme involving a series of initiatives implemented at the state and territory level in Australia (Cartwright, 2016). A feature of these interventions is the adoption of a case-managed approach to work with individuals. Intervention 01 began as a pilot in 2014, focusing on clients in the community who were identified as at risk of radicalisation due to their behaviours and associations. It was later expanded to support convicted terrorists and radicalised offenders released from prison. Hence Intervention 01 has a focus on secondary and tertiary prevention.

At the time of completing this research, Intervention 01 was led and co-ordinated by the police, who established a dedicated unit to case-manage clients. The scheme adopted a multiagency approach in which case management support was provided by several government agencies, e.g. Youth Justice and the Departments of Education and Health. Sources of client referrals into Intervention 01 include state and federal police, Joint Counter Terrorism Teams, the Australian Security Intelligence Organisation and other government agencies – for example, the Department of Education, Juvenile Justice, and Corrections. When an individual is referred to Intervention 01 a client risk and needs assessment is first undertaken, with also a deconfliction review completed to ensure they are not subject to an active police investigation. If assessed as suitable, the potential client is approached (including their parents if they are under 18 years old) and voluntary consent to participate is sought. Once voluntary consent is granted a case management plan is developed that identifies intervention goals and external services to be engaged. At the time of writing, Intervention 01 is now coordinated by a dedicated government unit, with the police providing case-management support. However, it still operates based on a multiagency approach, with referrals coming from the same sources as those listed above.

Data sources and method

Research on Intervention 01 occurred in early 2020 and comprised a mixed-methods approach to evaluate the program. The data sources included the analysis of client case notes and interviews with staff and clients. Both university and police ethical clearance was obtained.

The author was provided access to the case notes of 15 clients. This sample was selected in negotiation with the coordinator of Intervention 01 and was related to accessing information on clients with the largest volume of case note data associated with their participation. Rather than select a random sample, this decision was pragmatically driven to ensure the project had a meaningful amount of data to conduct various forms of analysis. The sample comprised a mix of discharged (n = 7) and active (n = 8) clients. This sample was only a proportion of the overall number of clients managed on the intervention. This sampling process raised the possibility of bias in the data. For example, it potentially skewed the sample towards clients who had been engaged in the program for a long time and had been subject to extensive engagements, resulting in more case note information being recorded. This is in comparison to clients who had been participating for only a few months and who had been subject to a few engagements, and hence would have fewer case notes. However, periods of engagement varied across the sample. The sample might also have been skewed towards clients given higher priority over others due to risks and needs, and thus were monitored and engaged more intensely, resulting in a higher volume of case note data.

The detail and length of the sampled client case notes varied. Some of the case notes for the discharged clients spanned participation periods from twelve months up to nearly three years for one client. For active clients, participation periods spanned three months for one client and up to sixteen months for another. These case notes covered different participation levels from 2015 to the end of 2019. The case notes were de-identified and then provided to the author for analysis, who was the only person granted security clearance to access them.

The client case note data was extracted from the police case management system. This comprised client background information, risk and needs assessments, dated observations about staff engagements with clients, service provider correspondence and feedback relating to client activities and appointments, inputs about staff and client engagements, police contacts, court documents and results of social media checks. These various inputs were entered by the intervention coordinator and staff.

To quantitatively report client social demographic histories and background information, client case notes were analysed to determine a set of case attributes. This included the identification of variables spanning social demographics (e.g. age and gender), historical information (e.g. previous contact with the criminal justice system), referral source into the program, mental health flags, family dynamics, intervention goals and listed indicators of radicalisation. These case attributes were developed to numerically capture various client background information. The threshold for capturing data was a simple dichotomous present/absence measure. If the information on an attribute was mentioned on a single occasion within the case note sources, it was reported as present. This approach was used to develop a quantitative dataset on client variables.³ The case notes were also analysed by the coding of case note information against different metrics of client change relating to compliance with intervention goals, adoption of pro-social activities (e.g. work and education) and behaviours reflecting a shift away from extremism (e.g. distancing from radicalised peers, repudiation of extremist views) and any episodes of non-compliance (e.g. failing to turn up to appointments) and regression relating to re-engagement in extremist-related behaviour and criminal conduct.⁴ These quantitative and qualitative data were used to evaluate the program. As noted, we provide a selection of our results here.

As mentioned, interviews were conducted with staff and clients. This comprised a small number of interviews with intervention staff (n = 3) and clients (n = 2). The interviews aimed to gain further insight about client participation and progress on the intervention and to complement information derived from the case notes. The small number of interviews needs to be kept in mind. It would have been ideal to have interviewed more clients, however few agreed to be interviewed and only two did so out of a total of ten clients that were approached, with some assessed as too high-risk by the intervention coordinator. The interview material was coded thematically relating to identifying the types and levels of supports provided to clients, factors that facilitated or inhibited client progress and examples of successes and setbacks amongst clients.

Results

Sample background and intervention goals

Here a selection of information is provided on the social demographics and intervention characteristics of our sample. This includes demographics, historical information including mental health and criminal history, client intervention goals and the listed behaviours identified as associated with a client’s radicalisation risk that resulted in their referral to Intervention 01. Due to the small sample size, we report both frequencies and percentages and, in some instances, due to gaps in the case note information, data on a client’s background was missing.

Demographics

Based on information from the case notes, the average age of our client sample was 22 years old, with just under half (n = 7) being between the ages of 14 and 17. Of our total number of clients, 86.7% (n = 13) were male. The majority of clients were born in Australia (80%, n = 12). Table 1 shows the level of education identified in the case notes based on the highest recorded level the individual had completed or was enrolled in when referred to Intervention 01. Many clients had not completed their final year of high school in Australia (i.e. Year 12). This potentially points to a wider problem of disengagement from the formal educational system amongst radicalised individuals, which has been noted as an issue in the Australian context (Cherney et al., 2020) and also amongst radicalised individuals engaged by European CVE programs (e.g. Hecker, 2021). However, studies have shown that violent extremism is not necessarily associated with educational status, and should not be understood as an individual-level risk factor for radicalisation (Brockhoff et al., 2015; Krueger & Maleckova, 2003; Wolfowicz et al., 2020). Others have argued that poor educational achievements create vulnerabilities during the early phases of radicalisation by generating social isolation and promoting low self-esteem, thus leading individuals to seek an alternative sense of status and meaning through the exploration of, and engagement with, extremist beliefs, propaganda and groups (Reiter et al., 2021). Recent studies have shown that educational levels can act as a protective factor against radicalisation (Lösel et al., 2018; Wolfowicz et al., 2020). Hence, there is some logic that many intervention goals and services associated with clients in our sample aimed to provide access to educational opportunities (see below).

Table 1. Education level amongst sample.

	Frequency	Percentage
Year 9	1	6.7
Year 10	5	33.3
Year 11	1	6.7
Year 12	5	33.3
Missing	3	20

Radicalisation behaviours

Table 2 outlines the types of behaviours listed in the case notes as resulting in a client’s referral and were recorded by staff as influencing their radicalisation. The various categories listed in Table 2 are those that were stated in the case notes as the primary reason for referral and the associated risks for radicalisation. In some cases, written conclusions from risk assessments were also included that were derived from a case management tool called RADAR, which has been used in the Australian context to guide risk and need assessments of individuals referred into case-managed programs such as Intervention 01.⁵ Based on data in the case notes the influence of a client’s associates was identified as a key source of radicalisation. The internet and social media were also common sources for engaging with extremist materials and views and led to individuals being referred into the program.

Table 2. Listed radicalised behaviours evident across sample.

	Frequency	Per cent
Radical associates	7	46.7
Viewing, downloading, possessing extremist material	7	46.7
Online interaction with extremist material, including posting, conversations and distribution	7	46.7
Involvement in violence/threatening violence*	7	46.7
Changed behaviours in school environment	2	13.3
Criminal history related to radical behaviours	2	13.3
Attempted to travel overseas	2	13.3
Convicted terrorist	2	13.3
Returned terror suspect	1	6.7
Increased devotion to religion	1	6.7

Note: N = 15.

*This does not include a general ‘history of violence’; it has to be clearly stated violence was related to radicalisation.

Some of the behaviours listed in Table 2 might be seen as problematic and troubling, given they influenced client referrals and assumptions about radicalisation risk. For example, a criticism could be that some categories in Table 2 reflect confusion between what is an indicator of radicalised behaviour, i.e. an observable behaviour that has been established as having a relationship with radicalisation to violent extremism (e.g. having radicalised associates, see LaFree & Schwarzenbach, 2021), and a simple descriptor that does not have such a relationship (e.g. a conviction for terrorism is not necessarily associated with ongoing involvement in violent extremism, see Renard, 2020).⁶ However, just as academics have faced challenges in identifying the causal connection across various factors and individual background characteristics associated with radicalisation, so to have practitioners. These challenges relate to making decisions about what constitutes a risk factor that should be an issue of concern given its relationship to the potential involvement and re-engagement in violence, compared to a vulnerability that on its own has little effect, but does so when there is the accumulation of complex needs resulting from various deficits and vulnerabilities (Al-Attar, 2020; Gill et al., 2021; Reiter et al., 2021; Cherney et al., 2021).

The characterisation of the behaviours listed in Table 2 as troubling is because it could be argued that some present a broad spectrum of potential indicators, such as changed behaviour in the school environment, which any number of teenagers might express.⁷ Based on details in the case notes, some of these behaviours included defying teachers, using profane language towards students and teachers, criticising students and teachers as ‘kafir’ and suddenly refusing to attend classes where females were present. Some of these changes were only associated with Muslim clients. A linked concern is using increased devotion to religion as an indicator of radicalisation risk, given that the Australian government advice is that it should not be seen as such.⁸ Again, this description was associated with Muslim clients in our sample. It would be inappropriate to conclude that the inclusion of these behaviours reflected bias in the types of indicators of radicalisation to violent extremism that referral agencies or police were utilising at the time, and which were underpinning their judgements and decisions about risk, and who should or should not be referred and engaged. The author cannot make such as judgement given the project did not specifically collect data on such decision-making processes.

However, these issues above do raise an important point relating to the thresholds set for referral into case-managed schemes. If set so low and framed by broad criteria then the danger is that it can have a net widening effect, drawing individuals into the program who do not constitute an actual risk and leading to government or police intrusion into their lives and potentially the lives of family members. The labelling and stigmatising effect of this can be extremely negative, particularly for the young (Thomas, 2016). This is also applicable when risk factors underpinning assessments are used that only have a weak association with radicalisation to violent extremism. These various concerns have been raised in the context of the UK Prevent strategy and its case-managed scheme called Channel, which is related to the referral of youth who are ‘acting out’ as a natural phase of youth development and whose behaviour is the result of complex social vulnerabilities and needs e.g. developmental disorders such as autism spectrum disorders and multiple mental health needs (Aked et al., 2021; Lowe & Bennett, 2021; Moseley, 2021). The lesson here is to have clear and specific evidence-based criteria around referrals into case-managed programs, and risk assessment processes that take account of multiple risks and the unique needs of individuals. One of the key tasks of the unit now responsible for coordinating Intervention 01 has been to develop clear guidelines for referral and initiate closer consultations with partner agencies to minimise the possibility of inappropriate referrals. This includes undertaking an eligibility case conference with the referring agency to discuss the appropriateness of the referral and any proposed intervention.

Historical information and client intervention goals

Individuals within our sample manifested background characteristics associated with violent extremist cohorts and individuals targeted by interventions in jurisdictions abroad, including previous criminal history and violence (Heide & Schuurman, 2018; LaFree & Schwarzenbach, 2021). For example, six clients (40%) had a criminal history, either being charged and/or convicted of an offence. Three clients (20%) had been incarcerated as adults and two (13.3%) incarcerated as juveniles. Seven (46.7%) were perpetrators of violence and two (13.3%) were victims of violence.

While the causal relationship between mental illness and involvement in violent extremism is contested, mental illness can present a vulnerability when coupled with other social deficits and cognitive dysfunctions (Al-Attar, 2020; Gill et al., 2021). There is no space here to explore further this relationship or debate. Our data captured mental illness as either officially diagnosed by a professional or indicated as a suspected mental illness that had not been diagnosed. Five clients (33.3%) were diagnosed with a mental illness and three (20%) were suspected of having a mental illness. Four clients (26.7%) had a history of drug and/or alcohol use and three (20%) had a history of trauma. These characteristics have also been shown to be present amongst radicalised individuals targeted by interventions in Europe. For example, Heide and Schuurman's (2018) evaluation of a Dutch program targeting released extremists from prison, reported mental health and psychiatric problems amongst clients. Hecker's (2021) assessment of a French program targeting radicalised individuals also reported clients with histories of violence, criminality, psychological disorders and addiction.

Table 3 illustrates the main client intervention goals identified across our sample. It should be noted there is an overlap across intervention goals, as a single client may have several goals. The main goal was to engage clients in pro-social activities and build an alternate sense of purpose through work and education. Scholars have argued that such practices are an important part of fostering disengagement amongst radicalised individuals by promoting alternative pathways and a new sense of identity (Chernov-Hwang, 2018; Jensen et al., 2020; Marsden, 2017; Cherney et al., 2021).

Table 3. Client intervention goals^a.

	Frequency	Percentage
Securing and maintaining employment	11	78.6
Assessment/assistance of mental health needs^b	10	71.4
Establish pro-social activity	7	50.0
Obtain license/driving lessons	6	42.9
Educational needs	5	35.7
Religious and/or cultural guidance	5	35.7
Develop coping skills/support plan	4	28.6
Assist with living arrangements	3	21.4
Family assistance	3	21.4
Peer support programs	2	14.3
Appearance modifications	2	14.3
Family engagement	1	7.1
Reintegration to society or develop exit strategy	1	7.1
Financial assistance	1	7.1
Address re-offending risk	1	7.1

^aN = 14; One client’s intervention goals were not reported so excluded from analysis.

^bOne count refers to the assistance of mental health needs to a client’s wife.

Social and family environment and support

A majority of the sample (n = 10, 66.7%) were reported as having experienced some form of social isolation from peers, friends or family. Such isolation presents a contextual risk factor for radicalisation and can act as a barrier to individuals desisting and disengaging from violent extremism (Barrelle, 2015; Chernov-Hwang, 2018; Lösel et al., 2018). The role of family members and the family environment has also been recognised as important in a person’s pathway into and away from extremism (Chernov-Hwang, 2018; Marsden, 2017; Schewe, & Koehler, 2021). Tables 4 and 5 outline different family environmental factors and support systems and the level of involvement by family members in the program. This is an important issue given the importance that family members (especially parents) can play in the deradicalisation process and in supporting the progress of a client (El-Amraoui & Ducol, 2019; Schewe, & Koehler, 2021; Sikkens, et al., 2017). Amongst our sample, some clients experienced an unstable family environment. However, there was a high level of support amongst family members relating to their child’s involvement in the intervention, including family members actively assisting their son or daughter in attending appointments or enrolling in applicable courses and engaging with intervention staff about their progression in the program. Schewe and Koehler (2021) point out that, family members can become involved in CVE interventions due to a desire and sense of moral obligation to prevent future trauma and suffering. For parents whose child has radicalised and even left to fight overseas and died or been arrested by authorities, this can be particularly strong due to a sense of parental shame, failure and guilt (Gerrand & Grossman, 2018; Schewe & Koehler, 2021). Providing support to one’s son or daughter by being actively involved in a case-managed program is a possible way of dealing with such emotion and fulfils the sense of parental duty to create a better future for one’s child and help them avoid any future contact with the authorities (Siennick, 2011). Hence, while some clients amongst our sample experienced conflict with their parents, others did not and had a supportive relationship. This can be the outcome of their child being referred into Intervention 01, which spurred a parent to help their son or daughter achieve positive change through providing social support (Cullen, 1994).

Table 4. Family environment.

	Frequency	Per cent
Single parent, separated* and not living together	7	46.7
Parents married and living together	5	33.3
Client married and not living at home	1	6.7
Family environment unknown	2	13.3

*separated includes parents who are divorced.

Table 5. Family involvement in intervention.

	Frequency	Per cent
Evidence of support and involvement	8	53.3
Minimal or limited support and involvement	3	20.0
No evidence of support or involvement	4	26.7

Individual case studies of client progress and forms of assistance

A selection of case studies is provided in this section. The detail of each case is based on the coding of case note information to capture forms of progression and regression and is also informed by information obtained from staff interviews about our client sample and interviews with two clients (i.e. interviews with Client 88 and Client 94). The coding of the case note data was based on recording intervention activities and cataloguing factors that inhibited or facilitated progress, which is the focus of the five case studies. Due to space limitations, we provide case studies on only five of the fifteen clients in our sample and there are limitations on the background information that can be provided regarding each client. These five were selected because they reflected variability across our client sample relating to assistance and progress.

Client 003

Client 003 was a teenage male born in Lebanon and enrolled in Year 9 (third year of high school in Australia). During his participation, he had an excessively high number of school absences. This client received educational departmental permission that he no longer had to attend high school. He was from a single parent home and was the oldest of three siblings. He had mental health issues and was socially isolated due to having few friends. He was referred to Intervention 01 due to his excessive viewing of Islamist extremist content online and was stopped by authorities from travelling overseas. Client 003 had social and psychological vulnerabilities, and these coupled with an unstable family environment and his social isolation, raised his risk of radicalisation given his online activities (Bjørgo & Carlsson, 2005; Borum & Patterson, 2019; Lösel et al., 2018). These accumulated vulnerabilities can be particularly pronounced for youth, which manifest in specific risks relating to online engagement of extremist content (Borum & Patterson, 2019; Cherney et al., 2020). Disengagement from mainstream institutions such as work and education meant that going online was one way for Client 003 to seek the social connections he lacked.

Hence Client 003’s intervention goals were focused on providing social connections and re-engaging in mainstream activities, as well as addressing mental health and psychological needs. Helping to generate a more stable family environment was also an objective. This included exploring employment options and helping Client 003 transition to stable employment, improving his health and fitness, completing mental health assessments, providing psychological support, participating in Police Citizen Youth Club programs and Islamic religious guidance, and providing support to the client’s mother that would assist her with parenting.

Client 003 consistently attended his psychological appointments, participated in a youth wilderness camp, and undertook training to help transition him to employment. He did not engage in any religious guidance or mentoring due to a lack of interest. Setbacks that Client 003 encountered included having to look after his younger siblings due to his mother leaving the family home for a period to see her family in Lebanon and to marry a man overseas. During this time, his compliance with the program dropped. Here we see family instability and parental absence influencing his progress.

Client 003 also displayed low motivation due to mental health issues, therefore, his consistent participation fluctuated overtime. This was common for many clients in our sample. This was also exacerbated by Client 003’s social isolation and ongoing family instability, contributing to levels of non-compliance and setbacks. However, engagement in activities to build a sense of identity/achievement and make new friends (e.g. participation in a youth camp) and high levels of informal peer support by staff (e.g. one-on-one conversations, help with attending appointments) resulted in spikes in positive progress and change relating to the display of pro-social activities and a more positive outlook about his future. Therefore, generating social connections through specific activities or the engagements of intervention staff can provide opportunities for individuals to disengage. Client 003 was eventually discharged from the program.

Client 66

Client 66 was a 27-year-old male, born in Australia, who had a history of involvement in white supremacist groups and was incarcerated as an adult for a non-terrorist related offence. He was referred to Intervention 01 after being released into the community on a parole order. He lived with his mother and sister during his time of participation, who expressed a desire to help him reintegrate, find work and break away from his associates and connections with far-right groups. Researchers have noted the relevance of these factors in facilitating disengagement amongst individuals who were members of white supremacist/far-right groups (Jensen et al., 2020). His intervention goals were to attend his psychological appointments, improve his physical health, identify work and educational options, improve his family engagement, identify housing options so he could live independently and remove his tattoos that displayed far-right insignia.

Client 66 did attend his psychological appointments and his overall relationship with his mother and sister improved during his involvement in the program. However, there were hindrances to his progress. This included the impact of his mental health functioning on his day-to-day coping, which compounded his ability to consistently engage in the intervention. This observation has also been made for other types of case-managed programs dealing with individuals who manifest complex mental health needs, substance abuse and forms of criminality (Rapp, 1998). There were also instances of regression in which Client 66 re-engaged with extremism related material. During his participation period, Client 66 was attacked at a train station, allegedly by a member of a right-wing group to which he once belonged. Many of these same barriers have been identified by research looking at the process of extremist disengagement amongst white supremacists more generally outside the intervention space (e.g. Kruglanski et al., 2019). Therefore, mental health problems can compound the difficulties that some clients may experience on a case-managed program, which requires a commitment to a routine of appointments and a functional capacity to engage service providers and staff. Like other clients documented here, intervention staff provided Client 66 with forms of informal support, including driving him to his appointments. Client 66 was subject to a parole violation and returned to prison during his participation.

Client 88

Client 88 was an active client when this research was conducted and was one of the client interviewees. Client 88 is an Australian female Muslim, who was charged with a terrorist-related offence as a teenager. She was charged along with a group of older males. Her mother and father were separated, and she was a victim of physical violence. During her participation in Intervention 01, it was reported she had good support from her immediate and extended family. Her intervention goals were to engage employment services and receive mental health support through counselling. Her participation was characterised by a high level of compliance, and she engaged with the required mental health and psychological services. Her offending behaviour was strongly influenced by her associates (suffering a romantic teenage crush on a radicalised male at the time) and pursuing a sense of identity as a teenager, which made her vulnerable to being influenced by older men. Similar processes have been noted in cases of young female Muslims who have radicalised (Alonso & Delgado, 2020; Bloom & Lokmanoglu, 2020). However, it has been emphasised that women are not simply passive participants in their radicalisation, or the victims of being romantically duped or brainwashed (Nuraniyah, 2018; Pearson & Winterbotham, 2017). Current scholarship on gender perspectives in CVE has emphasised the need to promote agency and resilience (Pearson, Winterbotham, & Brown, 2020). The latter includes a capacity to self-reflect about one’s decisions and actions (Crane et al., 2019). Some of these elements are evident in the case of Client 88.

For instance, intervention staff and Client 88 reported that being arrested and the psychological and informal support provided by staff, offered the space to allow for cognitive disengagement from her radicalised beliefs. This included beginning to see many of her radicalised associates as hypocrites and being flabbergasted by some of the beliefs she held at the time, and statements she made about her family ‘not being true Muslims’. This self-reflection for Client 88 included developing an increasing awareness of the negative influence of her previous associates and how they aimed to exploit her teenage vulnerabilities. Client 88 reported she had to ‘detox from my religion’ and stopped attending a Mosque, choosing to pray at home instead.

Client 38

Client 38 was a nineteen-year-old Australian male Muslim who was arrested overseas trying to enter Syria. It is alleged that he intended to join the Islamic State. He was returned to Australia and was engaged by intervention staff and consented to participate in the program. His father expressed support for his son’s participation in the intervention. A concerned expressed by his father was that his son experienced social isolation due to concerns about him being publicly identified by members of the community given his actions overseas. Client 38’s intervention goals included changing his surname, being provided psychological assistance, attaining his driver’s licence, and registering with an employment agency. There was observed progress in his day-to-day coping, and he engaged a psychologist and registered with an employment agency. No major hindrances to his progress on the intervention were observed and there was a high level of compliance, with staff reporting that Client 38’s participation was consistent. There was reported evidence that Client 38 acknowledged his past support for Islamist groups such as ISIS as misplaced and ill-informed. However, some questions were raised about the sincerity of Client 38’s participation in the program, with it emerging he was still engaging with radicalised individuals online. This fact raises the possibility that there was some level of disguised compliance occurring, referring to radicalised individuals deceiving program staff about their deradicalisation (Acheson, 2020; Weaver, 2021).⁹ The risk of such deception is a problem that confronts many behavioural change programs and case-managed interventions (Camilleri et al., 2020; Phillips et al., 2021). Client 38 was later arrested for a terrorist-related offence outside the observed participation period when this research was conducted. It was not evident from any of the evidence accessed that Client 38 was actively deceiving intervention staff and that his reengagement might have been caused by other factors, such as contact with previous radicalised associates. Either way, this case points to the challenge of assessing the genuineness of participation and identifying client activities and engagements that are authentic demonstrations of behavioural change.

Client 94

Client 94 was an active client when this research was conducted. He was a 19-year-old male Muslim, referred because he was charged as a juvenile for terrorism-related offences and had been released into the community. Data indicated that his peer network influence his radicalisation, who were all actively engaging with ISIS-related material online. His activities included producing extremist propaganda and he attempted to travel overseas, purportedly to join Islamic State, and to recruit members for ISIS. His pathway echoed a common pattern amongst youth who have been radicalised in Australia and abroad relating to exposure to radicalised associates and their differential reinforcement of extremist views and behaviours (Alonso & Delgado, 2020; Neve et al., 2020; Cherney et al., 2020). His radicalisation began in high school and he referred to the fact that engaging ISIS material and espousing their views ‘was the thing to do at the time … everyone was doing it’. While in detention he engaged theologically with an Imam, read Islamic texts, began to gain insights about his ideological beliefs, and completed his final year of high school. When released from detention his movements and associations were restricted and he was the subject of police monitoring. During his participation in Intervention 01, he lived with his mother and father and expressed guilt and regret about the impact of his behaviour and offending on his family, particularly the stress it caused his parents. He also reported that his extremist interpretation of Islam and support for extremist groups were misinformed and mainly influenced by his previous associates.

Client 94’s intervention goals were to secure employment, receive psychological support and address risks of reoffending, particularly relating to exposure to radicalised networks and individuals in the community. He enrolled in a trade course and secured work through his own initiative and contacts. The main assistance provided by Intervention 01 was the referral to a psychologist and financial assistance to pay for this service, which Client 94 regarded as ‘helpful in getting me use to being out and back in the community’. Intervention staff voiced some concerns about how superficially Client 94 sometimes engaged in the intervention, and that he frequently displayed a lack of insight regarding his offending behaviour and the risks presented by previous associates and peers who lived in the area and that he might encounter. Previous studies have raised this latter issue by arguing that intervention work to promote disengagement can be undermined by pre-existing community and peer networks that reinforce extremist norms (Lewis & Marsden, 2021; Weggemans & De Graaf, 2017; Cherney et al., 2021). This was raised as one of the main issues of concern by intervention staff relating to this client’s ongoing risk of re-engagement.

Practitioner and client experiences and insights

In this section, results from the interviews with three staff and two clients are summarised.

The individualised nature of the intervention was emphasised by staff interviewees, with each stating it was difficult to describe a typical type of client engagement. While there was consistency in the types of services offered (i.e. psychological, vocational, educational and lifestyle support), each client was described as different, therefore it was difficult to state what individual success looked like. Two of the interviewed case managers emphasised that it often took some time to build rapport and trust with clients, who were often suspicious of police. An initial process of rapport building was important to help clients ‘realise we are there to help them’, and was regarded as a milestone given the background of some clients and their negative history of contact with the police. This echoes similar practitioner experience with programs targeting radicalised offenders overseas, with the need to ‘break the ice’ and build rapport noted by Hecker's (2021) evaluation of the French PAIRS program targeting jihadists.

Staff also acknowledged that some clients were ‘very high need’, and often a high level of informal/peer support was required, which was time-consuming. While this might reflect a certain level of client dependence on staff, it does accord with reported experience and research across the health and correctional context that staff engagement of clients in case-managed programs needs to be underpinned by supportive, responsive styles, which can generate a commitment and motivation to change amongst clients (Camilleri et al., 2020; Trotter, 2015; Hachtel et al., 2019). In the context of Intervention 01, this included transporting clients to appointments, having regular ‘catch-ups’, participating with clients in sporting activities and having conversations about day-to-day concerns and future aspirations, family relationships, financial issues, ideas around employment, current and past associates, and ideological beliefs. For the latter, this included attitudes towards extremist groups and their ideology. Staff also emphasised that these informal engagements included engaging with family members, such as a client’s mother or father, who might have specific needs (e.g. health and financial) influencing the degree to which they could support their child. On some occasions, family members were also referred to service providers. This informal support was seen by one client interviewee as particularly valuable and assisted them to ‘keep on track and motivated’. The financial assistance provided by the program to see a psychologist was also acknowledged by both client interviewees as particularly helpful, given the costs associated with such services that they stated remained outside of their financial ability to pay.

The interviews highlighted the challenges that clients can face relating to their progression throughout the intervention, and which could have a bearing on extremist-related risk factors. For some clients, this included fluctuations in their mental health and general motivation, which can influence their participation and compliance levels relating to their intervention goals. One example cited was that consistently turning up to appointments for some clients was challenging due to a decline in their general motivation to ‘get out of bed’. This would also affect their ability to commit to a routine. For such clients, the attraction and influence of social media presented a risk of re-engagement with extremist content and individuals online. For clients who had been charged for a terrorism offence or had some connection to a terrorist plot or known radicalised individuals or groups in the local community, another set of challenges existed. One related to themselves and their family members being shunned by the local community. Such experience of community rejection and exclusion has been documented by previous studies looking at individual disengagement and the experiences of family members and wives of convicted terrorists (Chernov-Hwang, 2018; Gerrand & Grossman, 2018; Rufaedah, & Putra, 2018). Both client interviewees stated they had experienced such community rejection and exclusion, given their offending behaviours were terrorism related. Another challenge referred to encountering previous radicalised associates or co-offenders, which could occur inadvertently due to them living in the same area as a client. This presented a risk of a client being identified and singled out by these individuals who might attempt to reassert an influence over them. The broader literature has emphasised this as a key threat to reintegration and ongoing desistance (Chernov-Hwang, 2018; Kruglanski et al., 2019).

The challenging nature of a client’s family context and relationships was also raised as an issue. As mentioned, some parents were supportive of their child’s involvement in the intervention, and were instrumental in ensuring their child consistently participated, such as driving them to appointments. However other parents had limited capacity and were in conflict with their child and had little control over their behaviour. Thus, looking at how parental involvement can be enhanced was acknowledged by staff as an issue requiring further consideration. Increasing the involvement and participation of family members has been identified as an important feature of CVE best practice (Schewe, & Koehler, 2021); however, one staff member noted that family members might be the source of a client’s radicalisation risk.

Risks to individual client progression and disengagement were raised by interviewees and can be divided into two main areas. For those clients with few social networks or family supports, mental health triggers and life stressors presented risks for re-engaging with extremist’s settings and groups. This might, for example, include re-engaging online. For others, it was reconnecting with previous radicalised associates, or contact with persons who had links to extremism or radicalised individuals in Australia or overseas. Some clients were credited as having ceased contact with these networks, however, for others, it was an ongoing risk that needed to be acknowledged by the clients themselves and monitored by staff.

A challenge in assessing the effectiveness of a program such as Intervention 01 is gauging the genuineness of client participation. This requires discerning genuine commitment from superficial or apathetic engagement, which on face value might reflect compliance with the intervention and a desire to change but is largely perfunctory. This issue potentially existed in relation to Client 38 above and is applicable to the problem of disguised compliance. Staff were asked about this challenge and it was acknowledged that some clients may ‘go through the motions’, however how proactive they were being in making changes in their lives was one way of assessing genuine from superficial engagement. Hence a client was not waiting to be recommended, or offered a service, or had to be asked to reflect on their past behaviours that led to their referral but were themselves suggesting and initiating such actions. Another important strategy was to ensure progress was being monitored and assessed using various information sources triangulated across different services and activities.

Discussion and conclusion

Clients amongst our sample were characterised by various vulnerabilities that manifested in behavioural risks linked to radicalisation and violent extremism. These risks included online and offline behaviours and associations. There is no space here for an in-depth discussion about the causal connection amongst these risks and vulnerabilities or debates about risk factors for radicalisation. The data does raise the issue regarding the types of indicators underpinning referral processes and the assumptions of radicalisation risk utilised by case-managed programs. Consideration must be given by staff and program designers to the implications of referral for clients and their families or partners. There needs to be a focus on minimising the possibility of inappropriate referrals by developing clear guidelines and undertaking consultations with referring agencies and sources. Likewise ensuring services and intervention goals address the unique vulnerabilities and needs of clients is central to tailoring forms of assistance.

The main strategy to address the vulnerabilities and needs evident amongst our client sample was to provide alternative pathways and to promote pro-social activities. For those included in the client case studies, the focus was on strengthening protective factors against re-engagement (i.e. psychological functioning, family stability, pro-social networks, work and education). The data illustrates that informal help by staff and parents was important for supporting the disengagement processes that Intervention 01 was aiming to facilitate. The reason informal support is significant is because while services might be offered and new opportunities afforded to clients to engage in work and education and form new social connections, generating the types of prosocial change essential to disengagement must be embedded in the routines of case-managed clients. Informal support is vital to this process and establishes the conditions necessary for ongoing and intense engagement, helping to generate client change in the long term (Cherney & Belton 2021). These contextual conditions can determine success; however, they can be hard to quantify for the purpose of program evaluation.

Many other lessons are evident from the client data and each case study. Regarding factors that hinder client progress on case-managed programs, the data illustrated that mental health functioning, family instability and the influence of associates could have an impact. These findings point to the need for policymakers and practitioners to consider the issue of residual risk. In Australia, the concept of residual risk has mainly been raised in the post-sentence context relating to whether convicted terrorists present risks of reoffending and should be detained beyond their release date, or intensely monitored and supervised once released (Ripperger, 2021). It is also relevant to the intervention space more generally. However, it should not be framed as a client presenting an ongoing terrorist threat that can be solved by state monitoring. For clients on case-managed interventions, it is about understanding the setbacks and barriers they face and the vulnerabilities they exhibit. In the context of program implementation and client progression (and ultimate success), participation for some clients in case-managed CVE programs will be characterised by setbacks and fluctuating success. Hence some clients may present as having residual risk, which can be exacerbated by factors or events outside of an intervention to completely control (e.g. family breakdown; re-emergence of radicalised associates). This requires long-term monitoring and support, and any setbacks or client failures should not spell disaster for a program, given these will be inevitable (Horgan et al., 2020). Identifying the reasons why should be a priority for intervention staff and program funders.

The data here showed that client success in case-managed programs such as Intervention 01, can be influenced by several factors, which practitioners need to monitor and consider in their risk assessment processes. These include fluctuations in client motivation, sudden declines in mental health, struggles with day-to-day coping, levels of family support especially if a client is young, and family and relationship breakdowns. There is also the ongoing influence and (re) emergence of radicalised associates and networks, the inability to cope with setbacks, social isolation and integration challenges if a client has spent time in custody. These factors are not inherently unique to the intervention context and have been highlighted by studies examining the deradicalisation and disengagement process more generally (e.g. Altier et al., 2021; Chernov-Hwang, 2018; Jensen et al., 2020; Marsden, 2017). The implication is that intervention plans and goals will need to change based on fluctuating progress and shifting circumstances. There are also problems arising from superficial engagement in the context of assessing client change and their desire to disengage. While disguised compliance might be an ongoing risk it should not be assumed that it will always occur. Robust case-management processes, consistent engagement and collecting data from multiple sources on client participation and progress can help to distinguish whether clients are genuine about their motivations to change.

The data reported here provides insights into a little-understood CVE approach focused on individuals at risk of radicalisation and known extremists/terrorists. The present study illustrated the vulnerabilities to radicalisation amongst clients, the types of services offered to address them and the contextual conditions that influence client progress and ultimately program success. These conditions may differ for interventions operating in non-Western contexts, or countries that have a long history of terrorism related to local insurgencies or political groups. The data reported here do not provide insights into the practice of ideological re-education (e.g. instruction relating to religious education and awareness pertaining to Muslims who have radicalised), or the influence of counter-narratives on client change. Intervention 01 appears to encourage ideological change indirectly through promoting alternative pathways and identities. Whether this is effective in the long term in the absence of any explicit work on religious and non-religious ideological beliefs is difficult to say at this point. More broadly for the task of evaluation, the results show that to understand ‘what works’ we need to assess the contextual conditions that influence client progress. Not all support provided to program beneficiaries will be the same and progress is idiosyncratic, therefore aggregate and individual outcomes must be assessed.

Acknowledgements

This research was supported by the Australian Research Council Future Fellowship FT170100061. The author would like to thank the relevant agencies for their support and Emma Belton for her research assistance.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Australian Research Council Future Fellowship: [Grant Number FT170100061].

Notes

1 The study presented in this paper is part of a larger research project examining three state-based case-managed programs in Australia.

2 For results from a quantitative evaluation of Intervention 01 and another state scheme led by police, see Cherney & Belton (2021).

3 The amount each attribute was mentioned or detected did not hold weight in how it was captured in the final dataset.

4 See Cherney & Belton (2021) for the quantitative evaluation of intervention 01 and the application of a five-point metric to access client change.

5 At the time of writing there currently does not exist any published studies on the RADAR tool. It is informed by Barrelle’s (2015) pro-integration model.

6 This was pointed out by one of the reviewers.

7 This was raised as a concern by one of the reviewers.

8 This was also raised as a concern by one of the reviewers. For the relevant government website see https://www.livingsafetogether.gov.au/identify-and-act/the-signs (accessed 24-09-2021).

9 The risk of disguised compliance was very much highlighted by the Usman Khan case (London Bridge attacker) in the U.K., where Khan participated in various disengagement and rehabilitation programs, but where it appeared he deceived authorities about his deradicalisation. Similar cases of client deception have arisen with programs in Germany (see Reinhard, 2020) and Austria (see Oltermann et al., 2020).