Category : Articles

Sub Category : N/A

who becomes a terrorist? 205
7 For a theory of participation in these kinds of insurgencies and civil wars, see Humphreys and 
Weinstein 2008. 8 On eta, see Clark 1983; on the Red Brigades, see Weinberg and Eubank 1987; and Ferracutti and 
Bruno 1981; and on a mixed sample of international terrorists, see Russell and Miller 1983. 9 However, they did not present any data on the nonterrorist sample to support their comparison. 10 Sageman 2004; Federal Research Division 1999; Pedahzur, Perliger, and Weinberg 2003; Hassan 
2001.
the testable implications of a variety of alternative theories. Section 
IV discusses the Bengali case and the structure of the data. Section V 
presents the results of the analysis, and Section VI examines potential 
confounding factors and robustness checks. Sections VII and VIII dis-
cuss the generalizability and broader implications of these findings.
I. Definitions and Previous Literature
I define terrorism as the use of violence to attain political ends, when 
this violence is primarily intended to cause fear among the civilian 
population and when the psychological value of acts of violence is pri-
oritized over their immediate military value. These last criteria differ-
entiate it from insurgencies and coups, where force, rather than the 
fear of force, is the primary means used. Such a definition of “terrorist 
group” would thus exclude insurgent groups like the ltte of Sri Lanka, 
whose use of terror, while large in an absolute sense, was subsidiary 
to the broader military and territorial strategy of the organization. It 
seems probable that such groups also exhibit recruitment patterns that 
differ from other terrorist groups, as their control of territory enables 
them to extend better material inducements to members or use coer-
cion in recruitment.7
Previous empirical studies of participation in terrorism have long 
stressed the possibility of a socioeconomic root cause. In the 1970s a 
literature grew up on the social backgrounds of those involved in the 
emerging “third wave” of terrorism.8 All of these studies found that 
the perpetrators were in general wealthier and better educated than 
the populations from which they came, although only Ferracutti and 
Bruno attempted to make an explicit comparison to an outside refer-
ence group.9
 The increase in Islamic terrorism in the 1990s, particu-
larly suicide terrorism, generated more work on the social backgrounds 
of terrorists. Despite the radically different cultural and ideological 
backgrounds of these movements, the results have been broadly simi-
lar: Islamic terrorists seem socially privileged,10 although this result is 
less strong for suicide bombers than for other terrorists.
Like the older literature, many of these accounts are handicapped by