Deradicalization programs that might help prevent events like the Manchester concert bombing (above) have been largely untested.

Rex Features/Associated Press

Can terrorists be deradicalized?

Investigators are still piecing together exactly what drove Salman Abedi, the suspected assailant in the recent concert bombing in Manchester, U.K., to kill 22 people and wound dozens more, but early indications suggest he had become a radicalized jihadist. How formerly harmless members of society go on to embrace violent extremist ideologies is a looming question in the world of counterterrorism, yet increasingly so is the problem of “deradicalization,” or convincing people to abandon an extremist mindset.

Worldwide, hundreds of deradicalization programs have sprung up. They typically consist of trained counselors either convincing the extremists their religious views aren’t founded in proper theology, treating the subject’s extremism as a mental health issue, or trying to nudge the extremist’s value system away from violence.

Despite their ubiquity, there’s been precious little effort spent evaluating whether these programs actually work, writes Daniel Koehler, director of the German Institute on Radicalization and De-radicalization Studies based in Stuttgart, in a commentary published today in Nature Human Behaviour. He discussed his work with Science, as well as the dangers of failing to establish deradicalization program standards. The interview has been edited for clarity and brevity.

Q: Why are there so many deradicalization programs in the United States and abroad, despite little evidence they work?

A: Citizens, policymakers, and security officials have come to recognize that we cannot arrest or kill our way out of this problem. We can physically defeat groups like ISIS or other violent extremist and terrorist groups, but this leaves their appeal untouched or even strengthened.

Q: Is it dangerous to implement deradicalization techniques before they've been proven?

A: Using the wrong methods could cause a lot of damage. It could mean not detecting a high-risk person, or even shielding him from the authorities. The fact that a person is in a deradicalization program could make security officials and people in his community less vigilant, blinding them to his continued danger.

Q: What are the hurdles to assessing the efficacy of deradicalization programs?

A: There are many. First, there are no generally used definitions and concepts in the field, making it difficult to even find out if everyone is talking about the same thing. “Deradicalization” versus “disengagement,” for example, are concepts that sometimes are used to mean the same thing and sometimes aren’t. Also, we don’t know when a deradicalization case is considered to be finished, or how to measure recidivism. So we need standardization in the basic definitions.

When it comes to specific metrics, to use recidivism as an example, we do not have any base rates for terrorists who do not go through a program. Does recidivism only count as going back to the same terrorist group? What about going back to the same ideology? What about a different violent extremist ideology? Or does it simply mean any form of crime?

Q: How should a good program evaluation be structured?

A: In each program, there are multiple factors like staff training, risk assessment, case ranking, individually adapted treatment methods, double-blind peer review, etc. Each can be graded by an evaluator and an overall score can provide a measure we call structural integrity. This does not necessarily measure a program’s impact in any particular case, but rather whether it’s capable of having a positive impact. Any individual outcome depends on so many additional uncontrollable variables, but a program’s structural integrity tells us whether it can work at all.

Q: Can deradicalization programs help with “lone wolf” attackers who aren’t formally part of any terrorist group?

A: We know that in the vast majority of “lone wolf” cases, family and friends knew about the radicalization process or attack plan. Just look at the recent Manchester concert bombing case, where the attacker’s friends and community members made multiple attempts to alert the authorities. What deradicalization provides is a program in the middle acting as some sort of communicative bridge between security officials, family members, and communities. But the basic mechanism is the same across all groups: You need to have access to the radicalized person, identify the factors driving their ideology, design an intervention plan, and then track its impact.

Q: Your commentary in the journal accompanies a research paper that found that Colombian paramilitary terrorists tend to believe that ends justify means to an extreme degree. How likely is it that these and other findings in the academic literature would generalize to other terrorist organizations’ ideologies?

A: I strongly believe that the main methods and dynamics behind deradicalization research can be applied across different terrorist organizations and cultures. Research has shown a great deal of overlap regarding people’s individual motives for joining and leaving terrorist groups. Of course, the program type and structure must be adapted to each circumstance, but in the end, it’s about discovering the nature of a person’s bond to the terrorist group and ideology, and to design a tailored exit strategy for that person.