Clint Watts of PJ Sage has released part two of his study of the foreign fighter data from Sinjar, Iraq. The CTC at West Point was the initial conduit for the data and they wrote a useful accompanying report. Clint has gone further by recoding the data (all of which he makes freely available on his site). His new look at the numbers led him to some important findings, including:
- Al-Qaeda does little of its own top-down recruitment in Middle Eastern and North African countries.
- The Internet plays a limited role in radicalizing, recruiting, and coordinating young men in many Middle Eastern and North African countries.
- Returning veteran fighters play a crucial role in radicalizing, recruiting, and coordinating young men to fight in foreign countries.
- A handful of cities (what he calls “flashpoint cities”) produce a disproportionate number of foreign fighters.
Based on Clint’s findings in part two of his study, he counsels the following:
- Focus less attention on the Internet in the Middle East and North Africa and more on local social networks, especially in the flashpoint cities.
- Put greater effort into tracking foreign fighters leaving Iraq and Afghanistan.
This “flashpoint city” thing, isn’t that just (or mostly) the result of what data you happen to lay your hands on? Isn’t it perfectly reasonable to assume that al-Qa’ida or cooperating groups are more well-developed in some cities than others, and that recruiters tend to channel their angry young men through specific contacts in Syria or wherever, whereby they end up in the same places & under the same commanders in Iraq…? Especially given how compartmentalized a lot of this is. Then when the U.S. raids that particular place, they get data that will reflect only a small and very particular subset of the whole foreign insurgent population, sifted through the same recruiting networks and thus with some gross regional overrepresentations. Isn’t it Statistics 101 to know that you can’t draw conclusions about a larger population by looking only on a small exceptional subset that you dug out yourself — in this case, names of known Jihadists in one particular network?
It seems every month there is a new feature story on Terrorist City in some Maghreb country, and questions on why oh why so disproportionately many of its lost sons are going to Iraq, and I’m always struck by how the reporters manage to miss these obvious questions.
I saw your note and appreciate the feedback.
I actually agree to a small extent with what you say and actually address that in Appendix A at , http://www.pjsage.com/products.htm. There I discuss the weaknesses of the Sinjar data and other parts of the study.
All samples are the result of what one has but the question is whether the sample is representative of the population. I believe the sample is fairly representative of the foreign fighter population but likely skewed to represent more North Africans and Syrians due to the location of Sinjar, Iraq; where the records were recovered. That being said, the Sinjar records are the only verifiable data which illustrates where foreign fighters are from by city, atleast that I have seen. The vast majority of reports to date have been gross estimates and conjecture. These records provide tangible data with which quantitative approaches can begin. While not the solution, it at least provides definitive information on foreign fighters and a necessary counterbalance to anecdotal internet accounts and ‘expert’ opinions of where fighters are coming from.
What is important in my opinion is that recruitment is highly concentrated based on these records. It is not national or regional, but instead local, thus demanding an alternative approach to the nation-state system we currently utilize to counter radicalization and recruitment. Also, many of these locations are not remarkable and I must not have been sufficiently clear about that. It is not surprising that a half dozen people came from a city of millions such as Casablanca. But, the recruitment of more than 50 fighters from Darnah, Libya, and more than a dozen from Sanaa, Yemen in this sample alone suggests we need to explore what is happening in these particular locations to create such intense recruitment. As to whether this is indicative of only one network or not remains to be seen. However, other analysts have noted that many of these city locations were major foreign fighter producers during the Afghan jihad.
As for your Statistics 101 issue, statistics is by definition the inference of population characteristics through analysis of a sample. The power of one’s inference and analysis usually improves with an increase in the sample size. The minimum cited in social sciences often times is 30 for doing analysis, more than 1200 is usually ideal, and this sample (563) falls about halfway in between. The sample is also supposed to be normal but at this point, I have no way of determining if it is or isn’t. Unfortunately, I haven’t stumbled upon a uniformly administered survey of foreign fighters yet, but I hope to continue to build the database as I go and improve its accuracy. Getting a sample of those foreign fighters currently held in detention would help this greatly but I have only speculative numbers thus far and while some have recommended it, no one has published the numbers.
I guess what I am saying is some data is better than no data, and I would rather analyze now and try to develop solutions to end this than wait to do a historical study later of limited utility and with little improvement in my sample. As I get better/more data, I will reanalyze and re-direct trying to improve my findings.
For me, an 80% solution now is more useful than a 90 % solution in a decade.