Executive Summary
The purpose of this report is to provide guidance to policy-makers and implementers of countering violent extremism (CVE) and risk reduction (RR, also referred to by others as deradicalisation) programmes. While the examples provided are mostly from Africa, Asia and the Middle East, reflecting the authors’ professional experiences of programming in these regions, many of the tools and techniques presented will also be relevant to those operating in ‘the West’. CVE and RR provide two increasingly prominent frameworks for countering the influence of individuals and entities involved in violent extremism (VE). Widely understood to describe a range of preventative and non-coercive measures, CVE may involve, for instance, community debates on sensitive topics, media messaging, interfaith dialogues, training of state governance and security actors, and a variety of initiatives with individuals deemed to be ‘at risk’ of being drawn to violence, such as vocational training and mentorship programmes. While there are substantial overlaps between CVE and RR in terms of activities, and many authorities group them under the same umbrella, RR can be considered distinct because its activities more narrowly target individuals who were previously directly or indirectly involved in the production of violence, such as defectors from VE entities, or those serving sentences for terrorism-related charges.
This paper aims to assist policy-makers and implementers by examining approaches through which to understand the drivers of VE and the wider context in which this violence occurs. It also looks at the design of CVE and RR programmes, and outlines key issues relating to programme and evaluation. We recommend the following:
• Adopt robust classification systems for VE drivers: The commonly used categorisation of VE drivers into push and pull factors is overly simplistic and somewhat ambiguous. We suggest that the CVE and RR communities apply an alternative system, based on existing theory, that distinguishes between structural motivators, individual incentives, and enabling factors.
• Apply the ‘results frameworks’ and ‘theories of change’ approaches: These approaches should be applied in order to maximise the extent to which CVE and RR programmes contribute to their intended impacts. Through encouraging practitioners to articulate the programme logic, these methods assist with the identification of questionable assumptions and other potentially problematic aspects of interventions, and thus promote critical thinking about superior routes to the desired end.
• Recognise that CVE is not rebranded development programming: There is a substantial risk that undue emphasis is placed on unemployment, poverty and other such ‘structural’ drivers, particularly as many practitioners in this burgeoning field arrive from a background in development programming. While such issues may be of considerable relevance in particular locations, they are certainly not by themselves sufficient conditions for VE, and in specific contexts where VE occurs they may not even be necessary. Alongside such ‘root causes’ it is also important to consider the relevance of social networks, ‘radical’ mentors, revenge-seeking, the pursuit of status, and a host of other motivating and enabling factors. Put simply, the CVE community should ensure that the framework develops into a holistic preventative measure.
• Target ‘at risk’ individuals: CVE efforts to influence the broad community of actual or potential supporters of VE may be of substantial relevance in particular where support levels are elevated or such assistance is critical to perpetrators of this violence. However, CVE efforts should also specifically target individuals identified as ‘at risk’ of being drawn to VE, through mentorship, vocational training, and so on, to the extent possible in each location. Such targeting does not consistently occur at present – for example, it is not a requirement of USAID CVE efforts. Failure to concentrate efforts in this manner will invariably result in programmes that underachieve in their contribution to the reduction of VE (the issues associated with identifying such individuals notwithstanding).
• Mitigate risk without being excessively risk-averse: Implementers of CVE and RR programmes should mitigate their many possible negative effects, such as stigmatising specific communities, exposing implementing partners to an excessive risk of being targeted by VE entities, enabling VE entities to rally support through highlighting ‘Western meddling’, and so on. However, implementers should also seek to avoid the converse temptation to become overly risk-averse as this will impinge on their ability to achieve their intended impacts.
• Explore possibilities for experimental and quasi-experimental designs: Randomised controlled trials and quasi-experimental methods should be explored as a means to evaluate the performance of CVE and RR programmes. In the case of CVE this may involve (a) identifying a number of ‘at risk’ individuals, (b) applying ‘treatments’ (vocational training, mentorship programmes, etc.) to approximately half of these, and (c) assessing changes in their attitudes or behaviours as compared with the control group. While such methods should not be treated as a panacea, and they certainly cannot be utilised to answer the full spectrum of evaluation questions, the authors of this paper believe that they have the potential to provide evidence of programme performance in advance of that which can be supplied through non-experimental approaches.