Counter-extremism

Bridging the Evidence Gap: Proving What Works in Preventing Extremism

Commentary
Posted on: 31st January 2020
Cleo Blackman
Director of Programmes, Tony Blair Institute for Global Change

The year 2019 marked the first time evaluators became winners of the Nobel Peace Prize. Economists Abjihit Banerjee, Esther Duflo and Michael Kremer are pioneering experimental, on-the-ground research to test effective ways to tackle seemingly intractable poverty in the developing world. Their research is hailed as dramatically improving our ability to practically fight poverty. It demonstrates the need to adopt new approaches based on field trials and scientific evidence, unpacking the complicated roots of a problem and breaking it down into more precise questions.

The rise in Randomised Control Trials (RCTs) is considered essential to addressing persistent problems in empirically assessing development outcomes. There have now been thousands of trials aiming to isolate what works from what does not. Evidence from these trials is steering funding allocations for scaled up work all over the world. Valid criticisms of RCTs have been raised–concerns about ethics, reliability, and lack of insight on causality and cost. However, the intent behind their innovation, to turn a problem on its head, is worth thinking about more deeply when it comes to preventing or countering extremism.

Development has become smarter about unpacking what works and investing in understanding cost-effectiveness, for example which interventions have the best outcomes for financial investment or which offer the greatest value for money at scale. Those working to prevent and counter violent extremism (P/CVE) must also catch up with demand and get better at evaluating what works in practice and at scale to stop the over-reliance on academic research that is not grounded in any practical application.

This means researching how to take one-to-one mentoring in deradicalisation to scale in countries facing large numbers of returning fighters into communities affected by the violence they perpetrated (for example, Horn of Africa, Colombia, Syria, Jordan, Turkey and Lebanon).

It means understanding which kind of educational activities–dialogue, teacher training or resources–can best build resilience in young people through traditional and non-traditional religious education systems. This is particularly important in communities that suffer from a chronic lack of infrastructure and resources and in contexts where young people are more vulnerable to extremism.

It also means investing in research and evaluation that helps governments spot the signs of emerging violent extremism (what do the early warning signs look like?) so that we can build a case for why early, upstream intervention is more effective in the long term rather than trying to stick a plaster on the problem once it has got out of hand.  

The Government of Kenya provide a good example of forward thinking on this issue. They have developed county action plans both in communities targeted by extremists (with high volumes of incidents) and in those that are not currently experiencing the problem as much as their neighbours. While currently in planning stages, this recognises that a comprehensive approach to countering violent extremism must address both the ideology and the structural grievances which extremists use as a prism for recruitment to their causes. It recognises that social services and economic opportunities need to be equitably provided to everyone to build collective resilience to extremism. If not, there is a risk that targeted communities get opportunities that others do not. This would simply fuel more grievance, playing into the hands of extremists who argue that their approach to governance and state-building would be fairer and more beneficial for those feeling marginalised.

So, what role does the monitoring and evaluation (M&E) community have in advancing this thinking? Providing the evidence–and providing evidence that is hard to come by. Evidence on this issue requires proving, to some degree, the positive impact of interventions that build equitable inclusion. Effectively, that positive peace mechanisms can be shown to reduce propensity towards violence.

This could be done by comparing countries with different levels of radicalisation and extremism problems and different approaches to inclusion, diversity and equality. This would highlight some of the reasons why some contexts may appear to be more immune to violent extremism taking root than others. This is obviously a long-term endeavour, but we need to start leveraging indicative evidence now in order to further invest in approaches that show promise, demonstrate how milestones in the journey are being achieved, and what we are learning as we go.

Efforts to assess effectiveness when it comes to P/CVE have improved and are continuing to improve all the time. However, in the very few cases where the results are made publicly available, it reveals substantial constraints in applying standard M&E tools to these kinds of programmes. These include, the complexity of the intervention and change process, the consistent evolution of our understanding of the problem and terminology, the mismatch between project lengths, funding periods and time it takes to effect meaningful change, and the hyper-localisation of causes and effects that make standardisation impossible.

The P/CVE field has advanced a lot since Prevent was introduced in 2006, but its approach to M&E has been slower in pace than in the broader development sector. The catch-up needs to be accelerated because the risk of making uninformed policy decisions on issues related to national security is severe.

The recent London Bridge attack by Usman Khan demonstrates this risk keenly. Waiting to publish lengthy academic studies, when politicians are under short-term pressure with limited windows to enact meaningful change, is not going to work. It is not working. And we cannot afford to wait another 20 years to prove how to prevent extremist violence from emerging, or how to stop it when it does.

One thing that needs urgently to change is the conflation of preventing extremism with countering it. These are two distinct outcomes, with different change pathways, and they need different ways of thinking about measurement. One is about creating the positive conditions that can prevent extremist ideas, both violent and non-violent, taking root. The other is about changing attitudes and behaviours once those ideas have taken hold of someone.

There is simply not enough effort to separate these in the same way that has been done in the peacebuilding sector. Countering terrorism is focused on an individual and their social network. Preventing terrorism needs to be much more focused on communities and structural factors and, of course, has a much bigger range of potential consequences and outcomes linked to pathways that might never lead to extremism.

If a young person receives a quality education, has economic opportunities, a chance for social mobility and a life filled with purpose and meaning, and they never even contemplate engaging with extreme content or views, that is as successful an outcome as you could hope for. But from an evidential perspective, where is the proof that it was a specific factor, or a combination of factors, that occurred to make vulnerable people more likely to resist extremism? There are too many variables and external influences. In this sense, policymakers are chasing the pot of gold at the end of the rainbow–a simple action that will stop radicalisation in its tracks.

It is not just the prison system that could fail the victims or perpetrators of violent of extremism. It could be the education system. It could be the economic system that fails to give people opportunities that channel energies into positive outlets. It could be the friends and family who perhaps failed to spot or act on the signs that people are being pulled away. It could be the kids at school who bullied others for being different in some way. It could be the sophistication of the marketing strategies of extremists that thrive in an uncontested space. The point is that the culpability is a collective one of an entire ecosystem.

The adage “prevention is better than a cure” is often cited at this point. It stands to reason this is cheaper and easier. But this is challenging even for simpler issues. Let us take, for example, preventable diseases where a combination of awareness raising (to change behaviours) and vaccinations are tangible and proven methods to get results. Even these programmes struggle for 100 per cent success. There are immunities, reactions to drugs, external factors that impact on efficacy, and there are people who do not believe in vaccines or people who believe more in traditional remedies or the power of prayer. When you are dealing with people, nothing is simple. Everything is a complex change. And that means there is no simple answer or magic bullet for defining “what works” to prevent extremism.

So, what is the answer?

In short, M&E needs to be more central to P/CVE approaches–with the right tools, more data can be collected to give greater insight into why certain interventions are more likely to change behaviours. This requires being honest about what we can measure and what we cannot.  

We also need to share more data on localised approaches and outcomes that demonstrate positive trajectories. There needs to be more openness and willingness to share successes and failures across similar sectors (we know the issues are multidimensional, so are the solutions).

Lastly, we need to innovate more smartly and take risks where possible, for example around RCTs that help map impact on attitudes and behaviours but also tell us about how interventions work, so we can build positive conditions and replicate those. Ultimately, there is also a need to accept that we might be unlikely to be able to prove unequivocally the role of specific interventions alone in causing a positive impact. However, this elusiveness should not serve as a barrier to progress.