Terrorism, credit, and investment: evidence from a natural experiment in pakistan

Global terrorism driven by radical Islam proliferated in the past two decades because of advances in the funding and recruitment of extremist movements (Feldstein (2008)). The nature of these events stimulated academics and policymakers to study the organisational economics of terrorist groups (Berman (2003), Berman (2011), Shapiro (2013)). In so doing, a consensus has emerged on the importance of financial counter-terrorism (Bueno de Mesquita (2005a,b, 2007, 2013)).Despite innovative theoretical contributions on this topic, there is lack of quantitative evidence on their predictions, which leads to a debate on the costs and existence of financial counter-terrorism.

In this research, I offer an empirical contribution to understand the role of terrorism financing and recruitment in promoting terrorist attacks. I study two specific questions: 1) Do the timing and location of terrorism financing correspond to terrorist attacks? 2) What is the relation between financing and recruitment in generating attacks? To answer the first question, I follow 1,545 cities over 96 quarters between 1992 and 2015, containing the universe of terrorist attacks (e.g. 12,000 events). I also build a panel of 20 terrorist groups operating in 485 cities over the same period. To address the second question, I combine data from Jihadist fora operating in the dark web, with the work of two judges and a machine learning algorithm, leveraging novel techniques from the computer science literature.

The main findings of the study are as follows. First, Sunni-majority cities exposed to higher financing experience more attacks, with terrorist organizations reacting to temporary funding shocks. This terror escalation takes place only for highly capital-intensive activities (e.g. bombs, chemical, biological and radiological weapons). Second, supply exclusively explains the increase in terror attacks. I dissect the demand and supply of terrorist strikes by exploiting city-organization-time variation. Third, the elasticity of terrorist attacks to terrorism financing is 0.25, meaning that doubling the funding to terrorist groups leads to a 25% increase in the attacks. Finally, terrorism recruitment exhibits complementarities with financing. I show the latter by constructing a measure of terrorist recruitment using dark web data and machine-learning algorithms.

This work joins the literature on the organizational economics of terrorist and violent groups, and it contributes to that literature by providing empirical evidence that terrorist groups face non-negligible costs in storing and moving their funding. It is also connected to studies on crime and conflict in developing countries, and it offers insights through the introduction of a novel identification strategy. Finally, it contributes to the broad literature on the determinants of terrorism. In terms of the potential policy impact, this research is the first to provide quantitative evidence on the relationship between terrorism financing, recruitment and attacks. It emphasises the importance of financial counter-terrorism and provides key insights to design future policies