Civilian Defense Forces, Tribal Groups, and Counterinsurgency Outcomes Open Access

Abstract

How do states maintain their authority given that even weak non-state groups are now capable of challenging states by resorting to irregular warfare? In Part I of the dissertation, I propose that one of the ways in which states pursue internal security is by recruiting civilians into lightly armed self-defense units referred to as civilian defense forces (CDFs). In their role as government auxiliaries, these units assist the counterinsurgency (COIN) effort by counteracting insurgent guerrilla tactics. CDFs' relative affordability, local intelligence, and the propaganda value of coopting civilians, help offset insurgents' superior mobility, concealment, and covert civilian support. Empirical tests of the argument on two different crossnational data sets of all insurgencies completed between 1945 and 2006 reveal that incumbents who deploy CDFs are 53 percent more likely to be victorious. Given that CDF deployment is a more easily manipulable variable than many state attributes such as economic development or military power, CDFs appear to be an effective COIN policy deserving of further academic and policy attention.

The finding that this particular COIN policy is associated with incumbent victory raises an important theoretical puzzle. Namely, if CDFs really are as beneficial to COIN, why do all incumbents not deploy them? In Part II of the dissertation, I lay out a formal principal-agent model which posits incumbent fear of civilian defection to rebels as the main impediment to CDF formation. However, the model also reveals that this impediment should be more easily overcome in areas predominantly populated by tribal groups, where pre-existing intertribal blood feuds incentivize them to comply with the principal's wishes in order to maintain CDF membership and therefore an advantage relative to local rivals. Consistent with theoretical expectations, empirical analyses of a novel province-level data set (N=100) encompassing Turkey and the Philippines suggest that areas with substantial tribal populations tend to station twice as many CDF personnel as those with no tribal populations.

Table of Contents

1 Introduction 1

1.1 Civilian Self-Defense and Internal Order . . . . . . . . . . . . . . . . 4 1.2 Civilian Rivalries and CDF Deployment . . . . . . . . . . . . . . . . 7 1.3 Layout of the Dissertation . . . . . . . . . . . . . . . . . . . . . . . . 14 I CDFs and Counterinsurgency Outcomes 15 2 Insurgency, Counterinsurgency, and Civilian Auxiliaries 16 2.1 Insurgency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.2 Counterinsurgency . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2.2.1 COIN Policies. . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.3 Civilian Defense Forces . . . . . . . . . . . . . . . . . . . . . . . . . . 35 2.3.1 CDFs and Rebel Mobility . . . . . . . . . . . . . . . . . . . . 36 2.3.2 CDFs and Rebel Concealment . . . . . . . . . . . . . . . . . . 38 2.3.3 CDFs and Popular Support for Rebels . . . . . . . . . . . . . 41 2.3.4 Hypothesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 2.3.5 Caveats and Concluding Remarks . . . . . . . . . . . . . . . . 44 3 CDFs and COIN Victory: An Empirical Test 48 3.1 Research Design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 3.1.1 Dependent Variable. . . . . . . . . . . . . . . . . . . . . . . . 52 3.1.2 Explanatory Variables . . . . . . . . . . . . . . . . . . . . . . 54 3.2 Statistical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 3.3 Alternative Model Specications. . . . . . . . . . . . . . . . . . . . . 63 3.4 Endogeneity and Selection Eects . . . . . . . . . . . . . . . . . . . . 64 3.5 Maximizing CDF Eectiveness. . . . . . . . . . . . . . . . . . . . . . 73 3.6 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 II Tribal Groups and Civilian Defense Forces 81 4 Intertribal Rivalry and CDF Formation 82 4.1 CDF Formation as a Delegation Problem. . . . . . . . . . . . . . . . 83 4.2 Formal Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 4.2.1 Players and Moves . . . . . . . . . . . . . . . . . . . . . . . . 86 4.2.2 Payos. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 4.3 Formal Model Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 90 4.3.1 Non-Rivalrous Village . . . . . . . . . . . . . . . . . . . . . . 91 4.3.2 Rivalrous Village . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.3.3 Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 4.3.4 Equilibria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 4.3.5 Comparative Statics . . . . . . . . . . . . . . . . . . . . . . . 94 4.4 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 4.5 Tribal Society: A Proxy Measure of Rivalries. . . . . . . . . . . . . . 96 4.5.1 Denition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 4.5.2 Lex Talionis: The Imperative of Blood Vengeance . . . . . . . 100 4.5.3 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . 105 5 CDFs in Turkey and the Philippines: An Empirical Test 107 5.1 Research Design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 5.1.1 Dependent Variable. . . . . . . . . . . . . . . . . . . . . . . . 109 5.1.2 Explanatory Variable . . . . . . . . . . . . . . . . . . . . . . . 111 5.1.3 Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 5.1.4 Estimation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 5.2 Empirical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 5.3 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 5.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 5.4.1 Illustration: Tribal Rivalries and CDF Service . . . . . . . . . 125 5.5 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 6 Conclusion 130 6.1 Unanswered Questions For Future Research . . . . . . . . . . . . . . 139 Appendix A Documentation for COIN Outcome Codings 143 Appendix B Revisions to the Lyall/Wilson Data 178 Appendix C Documenting Civilian Defense Forces 190 Appendix D The Data At A Glance 213

About this Dissertation

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Laney Graduate School
Department	Political Science
Degree	PhD
Submission	Dissertation
Language	English
Research Field	Political Science, International Law and Relations Political Science, General
Keyword	CDF Self-defense Militia Insurgency Counterinsurgency
Committee Chair / Thesis Advisor	Reiter, Dan, Emory University
Committee Members	Beardsley, Kyle, Emory University Carrubba, Clifford, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Civilian Defense Forces, Tribal Groups, and Counterinsurgency Outcomes ()	2018-08-28 14:55:56 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions