Burden Sharing, Defense Networks, and Security Communities

• In Event: Security Organizations in Evolving Global and Regional Orders

Abstract

How should the burdens of collective defense be distributed? This question represents a longstanding puzzle. The key dilemma is to ensure that member states adequately contribute to collective defense arrangements despite the incentive to free ride. Much attention, both from scholars and policymakers, focuses on formal military alliances like the North Atlantic Treaty Organization (NATO). We introduce two innovations. First, we shift focus from formal military alliances to bilateral defense cooperation agreements (DCAs). Alliances represent only a small fraction of substantive defense cooperation. Further, new alliances are rarely signed, and the overall structure of the global alliance system has remained static for decades. By contrast, DCAs are now ubiquitous, numbering well into the thousands. These agreements implement frameworks that attempt to coordinate a wide range of defense activities, from joint military exercises to defense research to arms trade, and they thus provide a novel opportunity to explore burden-sharing problems.

Second, we conceptualize defense cooperation as a global network of bilateral ties. Governments’ defense expenditures depend not only on their bilateral defense commitments, but also their position within the global network. Dense clusters of DCA partnerships lead to the emergence of distinctive security communities -- i.e., groups of states that have more extensive ties to one another than to any other states. In principle, security communities enhance security. But communities also exacerbate the free rider problem by allowing countries to displace defense burdens to their various partners. States thus face conflicting influences. On the one hand, their direct DCA partnerships put upward pressure on defense expenditures. On the other hand, membership in dense communities reduces the urgency of allocating resources to defense.

Using new data on defense cooperation, we estimate a series of simultaneous network coevolution models. These models account for complex influences, including selection effects, mutual causality between DCAs and defense expenditures, and, most importantly, network effects. The analysis leads to a paradoxical conclusion. Ceteris paribus, DCA partnerships increase defense expenditures. But when those partnerships embed countries in dense communities of collaborators, expenditures decline. The analysis reveals the importance of considering both defense cooperation beyond alliances and the broader structural context in which defense cooperation occurs.

Author

• Brandon J. Kinne, University of California, Davis