The Contagious Diffusion of Worldwide Terrorism: Is It Less Common Than We Might Think?

Abstract

Studies of the contagious spread of insurgency and conflict across national boundaries have generated a good deal of empirical research over time. While the contagious spread of terrorism has also been a policy concern, few empirical studies exist on the extent to which terrorism spreads contagiously. This paper uses methods developed by criminologists to study the spread of crime to examine the world-wide diffusion of terrorism from 1970 to 2013. We distinguish between contagious increases (based on shared borders) and non-contagious increases (where no borders are shared). We define the "domino effect" as a particular type of contagious diffusion where high levels of terrorism spread to an adjoining country but also remain high in the host country. Our analysis shows that both contagious and non-contagious diffusion has been rare over the past 43 years, non-contagious diffusion is more common than contagious, and when contagious diffusion occurs, it is very likely to occur according to the domino effect. THE CONTAGIOUS DIFFUSION OF WORLD-WIDE TERRORISM: IS IT LESS COMMON THAN WE MIGHT THINK?Researchers and policy makers have long been concerned with the possibility that terrorism and other forms of political conflict and violence may quickly spread from one country to another. Analogies are often drawn to infectious disease with the implication that once a major outbreak is under way in a particular country, if unchecked it may rapidly "infect" other regions and even countries, spiraling out of control as in an epidemic. While the domino theory was primarily about the contagious spread of communism rather than political violence, it assumed that this contagion would happen across countries that shared borders. Recently, the rapid rise of the Islamic State of Iraq and the Levant (ISIL) has raised similar concerns about the contagious spread of terrorism across the border of Syria and Iraq into Turkey, Jordan, and elsewhere. In this paper, we draw on methods developed for studying the spatial diffusion of crime 12 to examine the world-wide diffusion of terrorism over more than four decades . We distinguish between contagious diffusion where high levels of terrorism spread to an adjoining country from a host country but remain at high levels in the host country (referred to here as expanded contagion) from contagious diffusion where high levels of terrorism spread to an adjoining country but do not remain high in the host country (displaced contagion). Note that expanded diffusion references the specific outcome commonly known as "the domino effect."Both of these types of diffusion differ from non-contagious diffusion whereby the tactic may spread to countries that do not share borders.Our analysis of annual terrorist attacks for all non-island countries of the world shows that terrorist attacks since 2010 are more concentrated at the national level than at any other time since our data begin in 1970. We also find that contagious diffusion (both expanded and displaced) has been uncommon over this time pe...

Local alliances and rivalries shape near-repeat terror activity of al-Qaeda, ISIS, and insurgents

“…1 A and B clusters are found to mostly coincide with geo-political boundaries, due to increased border security, attacks occurring in civic centers in the interior, or terrorist familiarity with terrain/culture. In some cases we observe spillover or domino effects, due to weak borders or historical/political precedents (17). The Afghan–Pakistan cluster arises due to militant groups residing between the two countries; the Syria cluster is found to include Lebanon due to the Syrian civil war spilling over to its neighbor; the Nigeria cluster includes small portions of neighboring Chad, Cameroon, and Niger, as Boko Haram attempted to evade government scrutiny; since AQ-affiliated al-Shabaab often launched attacks from Somalia into neighboring Kenya, the two form a unique cluster.…”

Rethinking Democratic Diffusion: Bringing Regime Type Back In

“…This approach is particularly valuable because, as Zhukov and Stewart (2013, p. 273) note, misidentifying a network of spatial “neighbors” that partly but incompletely overlaps with the actual network risks failing to accurately identify the diffusionary process at work—what we believe has happened when networks based on regime type have been characterized as geographic ones. We therefore follow a growing body of work in comparative politics and international relations that explicitly tests the importance of geographic and nongeographic networks (Cao, 2010; Danneman & Ritter, 2014; Elkins, Guzman, & Simmons, 2006; Fortunato, Swift, & Williams, 2018; LaFree, Xie, & Matanock, 2018; Williams & Whitten, 2015).…”

A Bayesian Approach to Modelling Subnational Spatial Dynamics of Worldwide Non-State Terrorism, 2010–2016

“…A thorough review of 43 studies that was carried out at country level by Gassebner and Luechinger (), along with global studies that were carried out at local level (Python et al ., ; LaFree et al ., ; Nemeth, ), suggested a wide range of covariates that are relevant to the occurrence of terrorist events. Among those, we consider covariates that satisfy two essential characteristics: (a)a potential relationship with the lethality of terrorism, and/or severity and/or frequency of lethal terrorist attacks; (b)availability at high spatial resolution, to model subnational spatial dynamics of terrorism world wide. …”

“…A thorough review of 43 studies that was carried out at country level by Gassebner and Luechinger (2011), along with global studies that were carried out at local level (Python et al, 2017;LaFree et al, 2018;Nemeth, 2010), suggested a wide range of covariates that are relevant to the occurrence of terrorist events. Among those, we consider covariates that satisfy two essential characteristics:…”

“…Most empirical research in terrorism has focused on its temporal dimension (Hamilton and Hamilton, 1983;Porter and White, 2012;Brandt and Sandler, 2012;Barros, 2003;Suleman, 2012;Bilal et al, 2012;Sandler, 1993, 2005;Enders et al, 2011;Raghavan et al, 2013) or has considered purely spatial models only (Braithwaite and Li, 2007;Savitch and Ardashev, 2001;Brown et al, 2004). Moreover, studies that have explicitly combined both space and time dimensions have been carried out at country level, on larger geographical areas (LaFree et al, 2010(LaFree et al, , 2018Midlarsky et al, 1980;Neumayer and Plümper, 2010;Enders and Sandler, 2006;Gao et al, 2013) or at subnational level of analysis but within specific study areas Behlendorf et al, 2012;Nunn, 2007;Piegorsch et al, 2007;Öcal and Yildirim, 2010;Medina et al, 2011;Siebeneck et al, 2009;Mohler, 2013).…”