Does poverty lead to violence? (The other view)

This is a guest post from Ethan Bueno de Mesquita, responding to my skeptical talk a few weeks back.

It has become almost conventional wisdom among academics that poverty and lack of education are not important causes and terrorism. The source of this view is work done by Alan Krueger, both joint with Jitka Maleckova and alone, showing that terrorist operatives’ economic and educational statuses tend to lie around, or even slightly exceed, the averages in their societies. Based on this evidence, in his book Krueger concludes, “there is not much question that poverty has little to do with terrorism.”

I disagree. In a 2005 article, I showed that one could generate the prediction that terrorists aren’t poor or ill educated in a model where there is nonetheless mobilization due to economic opportunity costs.
The key is that terrorist organizations screen potential recruits on an ability dimension that is positively correlated with both efficacy as a terrorist and success in the labor market.

To see the problem this creates for the argument that the relatively high socio-economic status of terrorist operatives implies that the economy is not an important determinant of terrorism, suppose that terrorist organizations accept recruits only over some competence threshold and that competence is positively correlated with income or education. Suppose, further, that economic downturns increase mobilization (perhaps by decreasing opportunity costs). In such a world, because of screening, the terrorist operatives actually observed are neither poor nor poorly educated, just as in Krueger’s data. Yet the conclusion is not
true: the supply of acceptable operatives and, therefore, the expected level of violence is affected by economic factors.

Of course, this theoretical argument does not establish that poverty causes terrorism. But it does suggest that data showing that terrorist operatives are neither poor nor poorly educated doesn’t settle the question. Doing so requires answers to two questions: First, is there a correlation between socio-economic status and effectiveness as a terrorist operative? Second, do economic downturns increase the supply of high quality operatives?

Two recent empirical papers show that, at least for the case of the Israeli-Palestinian conflict, the answers are yes.

In a 2007 paper, Benmelech and Berrebi study detailed biographies of all Palestinian suicide bombers between 2000 and 2005 and find that more highly educated operatives are more likely to (1) be assigned to high value targets and (2) successfully complete missions.

In a forthcoming paper, Benmelech, Berrebi and Klor add micro-level data on the Palestinian economy and (using a difference-in-differences approach) find that changes in local unemployment affect the quality of terrorist operatives coming from a particular area, as measured by education, age, and experience. Moreover, they show that this improvement in the quality of operatives translates into a more effective terrorist organization. During bad economic times, terrorist organizations recruit higher quality operatives, attack higher value targets, and do so with greater success.

My overall take on this (which may, of course, suffer from the self-serving bias) is twofold. First, the theory and empirical evidence point toward a relationship between the economy and the effectiveness of terrorist organizations, even though terrorist operatives are neither poor nor poorly educated. Second, I think the development of this literature is both a useful example of how the iterative dialogue between data and theory enriches and adds nuance to our understanding of complicated phenomena and a valuable cautionary tale about not jumping to strong policy conclusions too quickly based on limited evidence.

4 thoughts on “Does poverty lead to violence? (The other view)

  1. My first assumption about the studies by Krueger are that his inferences are based on the socio-economic and education levels of known terrorists, quite possibly from the ones who the U.S. has either killed, detained, or otherwise has information about. Otherwise this truly must have been a grueling study and if that is the case than I am mistaken. My contention with only using the terrorists that the US government (or anyone who has taken the time to do so) has clearly defined as such is that the odds are great that these are the most senior members of any terror cell- such as the suicide bombers mentioned in the post. As such, one would expect the education level as well as socio-economic status to be very high for these people. If they didn’t already have both to begin with I would think that a higher level of education than their peers would prompt them through the ranks more quickly which undoubtedly would lead to affluence, albeit undoubtedly via some illegitimate means. My concern about the study is that it doesn’t account for your “average” terrorist- if there is such a thing. It is all in the way that one defines terrorist: if you limit yourself to studying the top ranks then, just like any other business, you would find affluent and educated individuals, but if you also account for the terrorists who’s scope is to carry around a gun, then I would imagine you would not find affluence or high levels of education.
    Second, I will not question the research being done to find the increase or decrease in violence as economic conditions rise or fall, but I would be interested in just how much it would affect the expected level of violence or the quality of an operative: I just cannot imagine an economic downturn hurting those who are already affluent with such an impact that they decide to become a terrorist. Likewise, though possibly to a lesser extent, I do not see a middle-class Pakistani losing his job and deciding that it is best that he make the attempt to become a terrorist. It just seems farfetched.

  2. This may be all well and fine (although the last sentence also shows how these conclusions may just as likely unravel with more research). But doesn’t this miss the point a bit? “Terrorism” is a qualitatively different violence than “civil conflict and war,” Blattman’s main concern in his study (with perhaps slide 19 tossed in a bit haphazardly?). To explain the driver or trigger of poverty for the one does not seem to be the same as explaining it for the other. So, I’m not sure how these different explanations really clash?

  3. This is an excellent post. I had previously bought the argument that low education and poverty were not important determinants of terrorism.

  4. this is like arguing that poverty does or does not promote communism. Individual communists might not be from poor backgrounds any more than individual terrorists are, but what this would really be tracking would be relatively privileged elites’ response to poverty as a political problem, not proverty per se. You could easily argue that during bad economic times people would be more stirred to political action by the sight of the suffering they saw occasioned by the poverty.