The rise of religion and the fall of the civilization paradigm as explanations for intra-state conflict

Abstract

The study presented here compares the impact of Samuel Huntington's concept of civilizations and that of religion on domestic conflict between 1960 and 2004 using the State Failure data set. The results show that examining the religious dimension of domestic conflict provides a better understanding of the dynamics of domestic conflict than does Huntington's concept of civilization. The results show that Huntington's predictions for conflict have not come to fruition as of 2004. Civilizational conflicts are a minority of all conflicts. Muslims, while engaging in a significant amount of intercivilizational or inter-religious conflict, primarily fight other Muslims. When one takes population size into account, Muslims participate in a disproportionate amount of conflict. In absolute terms, Muslims participated in the majority of all religious conflict for the entire period covered by this study and in 2003 and 2004. Finally, religion increasingly impacts on domestic conflict. Religious conflicts-including religious wars like those in Afghanistan and Algeria, which are not civilizational but clearly between factions within the same religion-are consistently more common than civilizational conflict and became a majority of all conflict starting in 2002. This rise in religious conflict as a proportion of all domestic conflict is not a post-9/11 phenomenon, but is, rather, the result of processes that date back at least to the late 1970s.

The Influence of Religion on Interstate Armed Conflict: Government Religious Preference and First Use of Force, 1946–2002

“…A religion that is disfavored by the state has less influence. Inspired by Fox's (2008Fox's ( , 2015 Religion and State dataset (RAS), GRP measures government favor/disfavor toward various religious denominations in five areas:…”

“…That paucity may also reflect that a lack of data on religion-state relationships had made empirical tests of hypotheses impractical. This is no longer true for the post-Cold War period (Fox 2008(Fox , 2015, but until this publication of this article it was still true for the period before 1990. The IR field's renewed attention to religion suggests that it is overcoming those prejudices and embracing the constructivist viewpoint of religions as intersubjective ideas-good or bad-that influence preferences alongside other ideas.…”

“…A Stata log of all tests is included as Appendix 2. Table 3 summarizes results of regressions on GRP scores and indices from the Religion and State dataset (Fox 2008(Fox , 2015. For equitable comparison, all three IVs are rescaled to 0-10 and regressions are run only for the time frame common to all three datasets (1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002).…”

“…The top row uses the GRP score for the preferred religion of the state, whichever it may be (i.e., the state's non-religion-specific GRP). The middle row uses the Government Involvement in Religion index (GIR) from the original RAS dataset (Fox 2008). The third row uses my construction of a composite score from variables in the expanded RAS2 dataset (Fox 2015).…”

“…The middle row uses the Government Involvement in Religion index (GIR) from the original RAS dataset (Fox 2008). The third row uses my construction of a composite score from variables in the expanded RAS2 dataset (Fox 2015). 13 All measurements except one indicate that the greater the state's association with religion broadly speaking, the greater its propensity to initiate an interstate armed conflict (the GIR coefficient in the A model barely missed significance at p < .068).…”