The time for causal designs: Review and evaluation of empirical support for mechanisms of political radicalisation

Highlights

• Current empirical evidence for mechanisms of political radicalization is reviewed.

• Evidence cumulates on a limited number of factors and interventions.

• The way forward for research on radicalization lies in causal designs.

Abstract

This paper evaluates the most influential current approaches to the mechanisms of radicalisation on the basis of their empirical evidence and calls for a focus on research designs capable of arbitrating on matters of causality, not just correlation. It shows how the existing evidence converges on a handful of factors involved in radicalisation, including negative life experiences leading to fundamental uncertainty or loss of significance, which spur on the search for and identity shift towards groups with strong norms and ideals, including sacred values that enable extreme ingroup defences (e.g. acts of terrorism). The cumulative empirical data indicates support for some, but not all, kinds of interventions. Finally, because both theoretical approaches and current interventions propose cause-and-effect relationships, the paper argues that it is imperative that the field shifts its focus to experimental research designs capable of making causal inferences.

Introduction

How individuals, often young people, come to accept the use of violence as a legitimate means to achieve political change has emerged as one of the most pertinent questions for policymakers and social scientists in the last decade. The consequences of this radicalisation are often devastating, both in terms of specific acts of violence and the fallout from these events. Our explicit and implicit theories of the mechanisms of this process matter, because interventions - also those targeting the community or societal level - work (or do not) through individual psychological mechanisms (Webber & Kruglanski, 2018). Interventions that explicitly target radicalisation to political violence include programmes as diverse as mentoring and coaching, dialogical workshops, exit programmes, community outreach and collaboration and punitive measures (Romaniuk, 2015). However, the proposed mechanisms through which these programmes are supposed to work are often vague or rest on untested assumptions (Horgan, 2016; Thomas, McGarty, & Louis, 2014) despite the fact that “getting it wrong” can have dramatic iatrogenic effects and possibly contribute to further radicalisation (Lindekilde, 2012). As models of radicalisation and the interventions based on these models make claims of relationships in the “real world”, we must evaluate these models based on their empirical support. In this paper, I review and evaluate the most prominent psychological theories of political radicalisation on their empirical merits. I argue that current evidence indicates support for a handful of central factors and mechanisms that should not be neglected when designing interventions targeting groups and individuals at risk of political radicalisation.

The following section presents and evaluates the empirical evidence for the most impactful psychological approaches of the last decade. As the goal of reviews should not merely be to create a list of relevant factors, but actively synthesize our current knowledge (Borum, 2015), the paper goes on to discuss similarities and disparities in the different theories with a synthesis and ranking of the supported mechanisms. These mechanisms are then used to evaluate central kinds of interventions that exist in current policies across the world, illustrating the usefulness of this kind of research. The paper concludes with a call for a new research focus, one that employs research designs that can arbitrate on matters of causality, not merely correlation. Pre-empting the following review, the central seven approaches presented below are summarised in Table 1, which provides an overview of the central tenets, conceptualisation of radicalisation, proposed explanatory factors, as well as a comparative evaluation of the empirical evidence in support of the internal validity (causal connection), external validity (generalisability) and measurement validity of the approaches.