Vox has an interesting article today about whether the increased numbers of defections from ISIS that are recently being recorded reflect growing problems inside the movement. The post is an interview with Peter Neumann from the International Centre for the Study of Radicalisation, which has recently released a report claiming a rapid increase in defections. The article states:

ICSR’s researchers verified 58 publicly-reported cases between January and August of this year alone. The true number of defectors is likely higher — and the pace of defections from ISIS, according to ICSR, is increasing.

This information, and reports from the defectors themselves, is taken as a sign that the pace of defections has increased, and the organization is facing increasing problems retaining recruits. I’m not convinced that this is the case based on the evidence presented.

We know that last year there was a significant increase in the pace of recruitment of foreign fighters. If foreign fighters leave at a constant rate, with the median time period to defection of about, say, one year, then we would expect that the number of defections would lag the surge in recruitment by about a year. In general, it’s very easy to confuse a constant rate for an increased rate if the background population is not well understood and is increasing – as appears to have been the case in the past year for ISIS, if we are to believe the reports from Western “intelligence” agencies.

This problem is especially likely if defectors being detected by western agencies is a rare event, because specific numbers of rare events can fluctuate by large multiples of their average within a fixed period of time, due to the discrete nature of the events and the vagaries of probability. So while it’s possible that the rate of defections has increased, it’s also possible that this simply reflects a delayed effect of increased recruitment.

While it’s possible that the narratives reflect an increasing level of disaffection amongst foreign recruits, this could also simply be responder bias – people who want to bail are likely to speak with others who seem in the same position, and more likely to remember negative events than positive ones. So I think it’s dangerous to draw too many conclusions about the behavior of a large number of potential psychos on the basis of the reports of a small number of disaffected psychos.

This is a good example of the difficulties researchers face in understanding the epidemiology of disease (infectious or not) when the size of the affected population is not understood, and all the conclusions have to be inferred from observed cases. Unless one has a very good idea of the dynamics of the disease generation process, it’s very dangerous to draw conclusions from observations of a single point in the process, as is happening here. Of course, no one is going to be able to find out the dynamics of this process and we need information on which to base policy responses (a few of which are given in the article), but a lot of caveats are necessary when we want to draw conclusions on the basis of such limited data.

The deradicalization process is an interesting challenge. Assuming that the number of foreign fighters grew last year as suggested, and that at some point ISIS collapses, what is going to be done with these people? The Vox article suggests that these people are stone-cold extremists, whose disillusionment is not driven by horror at ISIS itself. But as the number of fighters increases the chance that any one of them has committed any serious war crimes will decrease (I assume) so a growing number of them will presumably be only guilty of having fought overseas, which is probably not even a crime in some countries (it is in Australia). Some kind of effort is going to be needed to deradicalize these people when (when!!!) ISIS collapses, and my guess is that process will be made more difficult by a punitive response to the returnees. It seems like jailing them is not going to help, and unless we want to take a step towards barbarity ourselves they are going to need to be subject to legal proceedings, and the full protections of the law, in determining their guilt for any crimes committed – a process that is unlikely to be successful given the difficulty collecting witnesses from a force that has dispersed to the four corners of the globe. But not deradicalizing them is surely not an option, and I suspect that leaving them overseas won’t be either – stripping people of citizenship is a dangerous precedent and in any case whatever society forms in the aftermath of ISIS is going to want to deport these people quick fast. But at the same time, I don’t know if anyone knows good ways to deradicalize terrorists – how hard is it? Is it comparable to deradicalizing guerilla fighters in places like Aceh, Northern Ireland or Angola? Or is it a completely different process?

I wonder if any government has any plans in place to deal with this? I suspect not, and I guess that any policy other than abandoning them overseas is going to get short shrift domestically. Even though a short-term punitive approach is going to make long term problems, my guess is that this is all the western powers are going to be willing to try…