Emergent Extremism in a Multi-Agent Model of Religious Clubs

Abstract

Over the past few decades, it has become increasingly common for terrorists to frame their objectives in religious terms and organize their activities within religious movements. This paper extends previous models of religious extremism to better account for observed patterns of extremism and test policies likely to reduce the appeal of extremism. By adapting existing models to a computational framework, it is possible to analyze the distributions of agents and clubs that make up a religious economy. Extremist groups emerge as just one tail of a distribution of religious clubs that ranges according to the magnitude of costly sacrifices demanded of members. By varying the simulation parameters that control the distribution of (secular) wages and the degree of substitutability between religious and secular commodities, we see where and when religious extremism is most common and which economic policies reduce its prevalence. In particular, the model demonstrates that direct transfers of wealth tend not to reduce the appeal of extreme groups, that extremism is more common when religious groups are able to produce close substitutes for standard secular goods and services, and that increasing access to public goods (and/or secular economic opportunities) can radically reduce the extremist share of the overall population. Quantile regression modeling of data from multi-nation surveys and institutional indices corresponds to the model's key results, as well as recent research connecting civil liberties and terrorist origination.

Keywords: Extremism, Religion, Sacrifice and Stigma, Terrorism, Multi-Agent Computational Model

JEL Classification: C63, Z12, H56, D71