Sorry, Europeans joining ISIS probably can't be explained in one chart

Why do Europeans join ISIS or other Sunni militant groups? Why would French and Belgian citizens commit terror attacks on behalf of ISIS, as they did last month in Paris? And why, for that matter, is Belgium producing more citizens per capita who are traveling to Iraq and Syria to join such groups than any other Western country?

One theory that keeps popping up over and over again is that it has something to do with economic trends. The argument goes like this: Muslim immigrants in Europe, many of North African descent, often face discrimination when trying to find jobs and are not well-integrated into society; this fuels resentment, anger, and a sense of injustice that pushes a small number of them to become violent.

These sorts of socioeconomic theories about radicalization are really attractive: They make sense intuitively and are something tangible we can understand (I'm poor, I'm angry), and they imply a relatively straightforward solution that can (at least theoretically) be addressed by policy.

The problem is that every time we look at the data, these theories just don't hold up.

Establishing a direct cause and effect with any degree of scientific rigor is incredibly challenging, both because of the many factors you have to take into account and because of the difficulty of getting adequate data. But maybe more importantly, as I've written previously, there is no one standard model of radicalization — every individual radicalizes for reasons that are unique and complex. So identifying a consistent trend is especially tough.

Still, when a bunch of terrorists all come from the same relatively small area around the same time, one can't help but wonder whether some root cause is at play. But to see how difficult it is to identify a single economic factor contributing to radicalization, it's worth looking at one recent study that attempts to do exactly that — and how the study ends up falling short.

What one study found on links between radicalization and economic exclusion

Chris Blattman, an associate professor of political science and international and public affairs at Columbia University whose research focuses on the causes and consequences of violence, recently wrote a blog post discussing research by Philip Verwimp, a Belgian economist who studies violence, that suggests there may be a link between ISIS recruitment in Belgium and the economic exclusion of immigrants.

Blattman explains that they're not saying that poverty is what's maybe making people want to join ISIS, but rather exclusion from the labor market. The difference is that because Belgium has a strong state welfare system, citizens — including immigrants — who are unemployed still enjoy full access to state welfare benefits that (presumably) keep them from poverty. But they are excluded from the labor market — ostensibly because of discrimination — and thus are not able to improve their lives.

Looking across several EU countries, Blattman finds that the number of fighters a country has in Iraq and Syria appears to be correlated with two things:

1. The percentage of immigrants ("non-EU nationals") who are employed in that country
2. The size of the gap in employment between immigrants and citizens who were born in that country ("own nationals")

(Graphs by )

In other words, Verwimp's findings suggested that 1) the more unemployed immigrants there were in a country, the more fighters that country had in Iraq and Syria; and 2) the bigger the disparity in employment between born citizens and naturalized citizens, the more fighters a country had in Iraq and Syria.

As Verwimp explained to Blattman:

One of my students from African origin, graduating from our MA program, told me (before the Paris attacks) "it is easier to get unemployment benefits in Belgium than to get a job." He decided to move to Canada. That summarises it. Migrants and their families have full access to the allocations of the welfare state, but face daunting challenges when they want to get ahead in life.

Verwimp doesn't explain exactly how this would make one want to join ISIS, but Blattman suggests that "the shame and injustice of exclusion, not poverty," could possibly be "what leads so many to rebel."

To their great credit, both Blattman and Verwimp are careful to emphasize that they are absolutely not proposing that this one factor alone explains ISIS recruitment in general, or even all ISIS recruitment in Belgium. Verwimp writes, "I refrain from attributing the high number of Syria fighters in Belgium and elsewhere solely to their situation in the labour market. I am not looking for one causal factor, that would be ridiculous for such a complex phenomenon."

The problems with the study

Even with that caveat, there are still some issues with Verwimp's research and the case that it shows a link between labor market exclusion and ISIS recruitment. This is not to say that additional research on this hypothesis isn't merited or that there is no link; just that the research they present doesn't make a strong case for it.

Verwimp's research looks at data on European citizens who have "joined Sunni militant organizations in the Syria/Iraq conflict," but in Blattman's post both he and Verwimp discuss joining ISIS. While the majority of Sunni foreign fighters who have traveled to Iraq and Syria have joined ISIS, that is not the same as saying all of them have joined ISIS. Many have joined ISIS's rival, the al-Qaeda-affiliated jihadist group Jabhat al-Nusra (the Nusra Front), or one of the many other groups fighting in Syria. Not all of these groups share the same ideology nor are they all interested in training Western foreign fighters to carry out terrorist attacks back home.

This matters because without making this distinction, there is no way to evaluate how socioeconomic factors in Belgium may affect a person's decision to join ISIS instead of a more locally focused jihadist group that is fighting primarily to topple the Assad regime and make a better Syria, rather than to create a new Islamic caliphate and take over the world.

The data on foreign fighters who have traveled to Iraq and Syria also does not address those who radicalize and decide to carry out terrorist attacks on their home countries without ever having been to Iraq or Syria. The motivations for wanting to fight in Syria against the Assad regime and the motivations for wanting to blow oneself up inside a Paris theater may be quite different — or they may be the same. The data doesn't illuminate that distinction, so there's no way to tell.

Another issue is that Verwimp defines "migrants" as "anyone with a recent migration experience in the family." That definition is incredibly vague and broad. According to Verwimp, the data he uses in his study is "Eurostat data for the labour market from the year 2010." Because he doesn't provide any additional information, I'm not certain which specific data set(s) he used, but he appears to draw the data from this report.

That report contains labor market information for "migrants and their immediate descendants" — already a more precise definition than the one Verwimp provides — that is broken down by a number of clearly defined variables, including:

• Citizen born in the country, citizen not born in the country, or noncitizen
• Country of birth of father and mother
• Number of years spent in residence in the country in question
• Legal work status (are there legal restrictions that prevent the immigrant from working, or is it just discrimination?)

So the report has a lot of factors that one might look at to evaluate the degree to which certain groups of people are excluded from the labor market. But Verwimp merely compares "number of fighters in Iraq and Syria per million inhabitants" with "% of non-EU nationals in employment" and then with "gap in employment between own nationals and non-EU nationals."

In other words, his measure of radicalization (how many people per capita from a country go to fight in Iraq and Syria) doesn't quite measure radicalization, and his measure of labor market integration is oddly imprecise.

There's also an issue of timing. The data Verwimp uses is a snapshot of the labor market in 2010, which Verwimp says he used because "given the age of the Syria fighters at the time of departure (mid-twenties), data for the year 2010 give an adequate picture of the state of migrants on the labour market, to wit at the time in their lives that the Syria fighters were confronted with the labour market."

But the problem is that in 2010, the Syrian civil war hadn't even begun, so many of the people who would go on to fight in Syria and Iraq once the civil war was raging were probably only in their late teens or early 20s in 2010. Of the 440 Belgian foreign fighters in Syria and Iraq as of January 2015, 146 had arrived in just the preceding three months. And Pieter Van Ostaeyen found that among a group of 202 Belgian fighters, the average age at death was 25.7 (meaning they were even younger when they'd actually left Belgium).

This means that many of the Belgian citizens who traveled to Iraq and Syria — Verwimp's metric for radicalization — would have been somewhere around age 20, give or take a couple of years, as of 2010, when Verwimp's data was collected. As such, their lack of participation in the labor market could potentially be explained as an issue of youth unemployment, not exclusion due to racial or religious discrimination.

This matters because if we're evaluating Blattman's hypothesis that it's "the shame and injustice of exclusion, not poverty" that "leads so many to rebel," the fact that younger people have a harder time getting hired than older adults worldwide wouldn't seem to bear out the "shame and injustice" idea in the same way systematic racial or religious discrimination would.

This is not to pile on Blattman or Verwimp, but rather to illustrate how hard it is to establish a clear link between macro-level economic factors such as labor market participation with something as individualized and complex as radicalization. That's also not an argument against trying to study and understand which factors, if any, may contribute to radicalization — indeed, it's imperative that we do. But as badly as we might want to find a clear, straightforward, relatable chart or graph that will answer the question of why a group of young men in France and Belgium would decide to murder so many innocent people on a November day in Paris, it just may be the case that no such chart exists.