Religious Institutionalism: A Domestic Explanation for External Support of Rebel Groups

ABSTRACT

What motivates state support for rebel groups? The literature on state support for rebel groups has made critical developments. In particular, scholarship has shed light on the impact of religious identity links. Less work, however, examines the level of religious institutionalism in external states. I argue that the impact of religious links is conditional on the extent to which religion is institutionalized in the external state. Religiously institutionalized states allow domestic religious forces more space in the political arena, which increases pressure on state leaders to support co-religious rebel groups. Using statistical analysis, I find that the interaction of religious institutionalism and religious links affects the likelihood of support. When an external state and rebel group have religious links, and the external state has a high level of religious institutionalism, the likelihood of support is high. In contrast, without a high level of religious institutionalism in the external state, religious links do not increase the likelihood of support. The relationship is driven by cases where the external and target states do not share a religion. This article contributes to existing literature by moving beyond transnational religious links and focuses on how religious institutionalism increases domestic pressure to support co-religious rebel groups.

KEY WORDS:

• Civil war
• conflict
• external support
• religion

Supplementary material

Supplementary data for this article can be accessed on the publisher’s website

Notes

¹ I distinguish between external and target states. Target states refer to the states where and against whom the rebel group is fighting. Other studies have referred to target states as “host states.” “External states” refer to states in the geographic proximity of the target state.

² They find that the presence of moderately strong rebels, groups with transnational linkages, and interstate rivalries are all important predictors of the likelihood of a rebel group receiving support.

³ Scholarship has examined religiously institutionalized states but in the context of interstate war (for example, Henne 2012). This paper adopts the concept of religious institutionalism to examine another aspect of foreign policy: support for rebel groups.

⁴ Maoz et al. (2012) also employ a triadic approach.

⁵ The biggest contemporary religion (Christianity), for instance, is bigger than the biggest contemporary ethnic group (Han Chinese).

⁶ The lack of contemporary empirical examples of state support for rebel groups because of nonideological ties (see Byman et al. 2001) discourages further its focus.

⁷ See (Carlson and Owens 2003; Casanova 2011; Dark 2000; Fox and Sandler 2004; Rudolph 1997; Thomas 2005).

⁸ This phenomenon is more relevant for international religions with transnational networks. The Catholic Church, with its wide reach, transcends national boundaries, and provides an example. Other examples include Aga Khan IV, the current Imam of Nizari Islam, who provides guidance to followers across countries. Shaykh ‘Abd al-Halim Mahmud (Shaykh of al-Azhar University between 1973 and 1978) sent letters to Arab leaders for money—receiving money from Saudi Arabia and Kuwait (Zeghal 1999), illustrating religion’s capacity to reach across national boundaries. Ayatollah Khomenei, who spent significant time studying and working in Shiite religious centers in Iraq, worked to portray himself as the religious leader for followers of Twelver Shiite Islam both inside and outside of Iran. Beyond Ayatollah Khomenei, a long history of shared and continued scholarship, networks, and religious leaders exist between Shiite organizations, seminaries, and mosques in Lebanon, Iran, and Iraq as well as between more distant locations, such as Zanzibar, Tanzania, and Michigan (Nasr 2004).

⁹ Numerous theories on the “democratic peace” (for instance, Oneal and Russett 2001) attest to the attention on domestic institutions in explaining foreign policy.

¹⁰ This definition follows Henne (2013). See also Fox (2008); Gill (2008); Kuru (2009); Pew Forum (2012); Philpott (2007).

¹¹ I describe a state’s religion based on the majority of its population. Thus, the US is a “Christian” state, not due to its ideology but because of its religious demographic make-up. Pakistan provides an example of a state with a high level of religion–state connections. Pakistan provides provisions for Islam as the official religion for legal code, enforces a certain form of Islam and restricts other types of religious activity, allows the participation of Islamic parties in governing coalitions, provides state support for Islamic activities, and uses religious organizations to administer the state (see Henne 2013:22).

¹² The ulema is a term describing a body of Islamic scholars that are recognized as having specialized knowledge of Islamic law and theology.

¹³ It is worth noting, however, that the Shaykh was appointed. It is likely that in some ways incorporating religious organizations into the political arena and process moderates their demands and changes their preferences. While this may be true, it does not mean that religious organizations have to change their stances completely. Furthermore, religious organizations often maintain different parts. Involvement of the organization’s core in the political process may limit its demands. The organization’s periphery, who operate more independently, are still able to use the organization’s institutions and networks to promote its interests. Egypt during the early 1980s provides an excellent example, as periphery members of al-Azhar maintained higher degrees of leeway to express their preferred demands (see Zeghal 1999:386).

¹⁴ Newly formed committees in the Academy of Islamic Research tried to codify Sharia law and use Parliament as the main agent for change. They tried to do so by encouraging members to discuss their preferred topics (see Zeghal 1999:383).

¹⁵ See (Appiah 2009; Carlson et al. 2003; Casanova 2011; Horowitz 2009; Thomas 2005).

¹⁶ For example, Egypt’s 1961 reform encouraged “modernization” of al-Azhar, giving them the job of legitimizing political decisions (Zeghal 1999:373).

¹⁷ Since rebel groups are nonstate actors, rebel group religious institutionalism may look different from state religious institutionalism. Future research should theorize more on the ways in which rebel groups and other nonstate actors can institutionalize religion. Firmer theory on nonstate actor institutionalization of religion will provide a foundation for future theoretical arguments and hypothesis testing of its impact on support.

¹⁸ Proximity impacts states capacity to engage in both cooperative and conflictual behavior. As such, this article is limited to dyads that are likely have some sort of opportunity to engage in conflict. Gleditsch and Ward’s (2001) relevant states include states whose minimum distance is 950 kilometers or less. Other scholarship has defined states as relevant if the minimum distance between capitals is 950 kilometers or less, yet this tends to overstate distances between states and creates a bias for geographically larger political entities. In addition, great powers are more likely to be capable of providing support across large geographic distances. I also include as relevant external states the five countries in the UNSC.

¹⁹ To enter the data set, the rebel group must meet the UCDP criteria: 25 battle deaths in that year.

²⁰ See Fox’s Religion and State Codebook (Fox 2012).

²¹ Robustness checks include a dichotomous indicator of democracy—regimes with scores of 6 and higher are coded as democracies (1), while all others are coded as non-democracies (0). Results are consistent with the primary analysis and are found in the Appendix.

²² Rebel groups in countries rich with natural resources may be more attractive organizations to support. To address this potentially confounding variable, I include three yearly measures for resources. The measures come from Buhaug, Gates, and Lujala (2009). The first measure is a dichotomous indicator of whether or not significant gemstone production occurs in the target country. The second measure indicates if the target country has oil and gas in a conflict zone. The third measure is an indicator if the target country has drug cultivation in a conflict zone. The data is available until 2003 and are run as robustness checks. Results are consistent with the primary analysis and can be found in Appendix.

²³ Variables come from the Expanded Armed Conflict v2.3 (EACD) data set (Cunningham et al. 2009).

²⁴ Note that Model 1 includes external state–target state religious links as a control variable.

²⁵ When the external state, target state, and rebel group share a religion, an increasing level of religious institutionalism does not correlate with a higher likelihood of support.

²⁶ When the external state and rebel group do not share religious links, a higher level of religious institutionalism does not increase the likelihood of support.

²⁷ The difference is not statistically significant at a conventional level.

²⁸ The measure is positive when the interaction term is not included and is consistent with findings from previous literature (San-Akca 2016).

²⁹ This is likely due to the numerous subsampling of the data and lowered number of observations in each subsample.