Shocks and Turbulence: Globalization and the Occurrence of Civil War

Abstract

Several scholars argue that systemic global trends are pulling individuals not only upward toward the global level, but also downward to the local level; the result is a potential loss of authority for the state (Ferguson and Mansbach 2004; Rosenau 1990). Their theory of “fragmegration” can provide a causal mechanism for why longstanding grievances may erupt into civil war at a particular time. While increased global exposure does provide both states and individual citizens with tremendous benefits, sudden “shocks” of globalization can overwhelm a state's capacity to offset the negative impacts of globalization, thus weakening a state's capacity to deal with rival polities for the allegiance of its citizens. The present study conducts a cross-sectional logistic regression with discrete duration analysis to test the impact of globalization shocks on the onset of civil wars between the years 1970–1999. The results demonstrate that increasingly dramatic changes in the level of global integration are associated with an increased risk of civil war onset.

Keywords:

• civil war
• coping mechanism
• globalism
• globalization
• state capacity

Acknowledgments

An earlier version of the paper was presented at the 67th Annual Midwest Political Science Association Conference held in Chicago. I thank conference participants Jeffrey Michael Cavanaugh, Johann Park, and Horace Bartilow for their helpful comments. I give special thanks to Sara Mitchell for the great deal of valuable advice, comments, editing, and insights that she provided throughout the process of writing this manuscript. Finally, I would like to thank Cameron Thies, Thania Sanchez, Robert Urbatsch, Fred Boehmke, Amanda Frost Keller, one anonymous reviewer from the International Systems and Global Governance seminar at the University of Iowa, International Interactions editor Paul Diehl, and three anonymous reviewers for their thoughtful comments. All remaining errors are my own. Replication data are available on the International Interactions dataverse page at http://dvn.iq.harvard.edu/dvn/dv/internationalinteractions.

Notes

¹Note that even the least engaged states, such as Bangladesh, Benin, Burundi, or Rwanda, have become more globally integrated, particularly relative to their starting points.

²This is similar in logic to the inverted “U” found by Hegre et al (2002) in regards to democratization and civil war propensity.

³For examples of globalization being used to describe an increase in trade, see Dollar and Kraay (2002) or Milner and Kubota (2005). Recent work by Elbadawi and Hegre (2008) measures globalization and economic shocks in much the same way.

⁴Economic globalization is more than just trade. It involves an integration of financial and investment sectors that contribute to creating a more fully integrated world market between individuals and firms. This requires us to look at more than just how governments engage each other and explore contacts between individuals and firms. As the current credit crisis illustrates, even if states intervene in the global economy, their role is limited and systemic changes impact states far more than they impact the system.

⁵This also fits well with the established literature of “international” and “world society” developed under the rubric of the English School and more generally within sociology. For an in-depth discussion from the English School perspective on this topic, see Brown (2001), Buzan (1993), and Buzan and Little (1994). For a sociological perspective, see Meyer et al (1997). Finally, to examine the development of state and nonstate actors coexisting within a “world society,” see Rosenau (1988).

⁶This term was first coined by Rosenau (1990:3) and refers to the work done by several contemporary scholars, such as the previously noted Rosenau (1990, 1997, 2003) and Ferguson and Mansbach (2003, 2004), as well as Ruggie (1998). Postinternational scholars claim lineage from classic theorists such as Thomas Paine and Immanuel Kant, as well as more recent figures such as Karl Deutsch. The interdependence and regime theories focus on nonstate actors, much like Keohane ([1984] 2005) and Keohane and Nye (1971, [1977] 1989), but postinternationalists tend to emphasize their role to a greater extent.

⁷Proponents of the state weakening argument are not without their critics. For a countering view, see Gilpin (2001), Hirst and Thompson (1999), Walt (2002), or Waltz (2000).

⁸Legitimacy is the accepted right to rule or exercise authority over a population.

⁹The first outcome is similar to the greed/grievance explanation for civil wars that is proposed by Collier and Hoeffler (2002, 2004) while the second resembles the state weakness theory posited by Fearon and Laitin (2003).

¹⁰An exception to this is the work of Thyne (2006, 2010). Using a bargaining model he finds that cheap signals by foreign states create discord between their expected behavior, resulting in miscalculations of either their strength or that of the opposition among government and insurgent leaders.

¹¹The term “moral community” should not be considered to hold any actual normatively superior connotation; rather, it simply means that a group is united by a commonly accepted value or belief and they believe it to be normatively superior. See Buzan (1993:339) for a similar sentiment. Haas' (1992) idea of “epistemic communities” can be considered one type of moral community based on technical expertise.

¹²These rewards can include eligibilityfor International Monetary Fund and World Bank loans, as well as formal aid from other states.

¹³Some argue that these resources are used to buy off opposition and actually serve to prevent civil war. See Fjelde (2009).

¹⁴It is important to keep in mind that these measures are proxy variables for the rather opaque concept of globalization. They are in no way intended to be exhaustive, but they are able to capture the general concept. In particular, while the measurements based on McDonalds, Ikeas, and the trade of books are not ideal, they are intended to capture how culturally linked a country is with the rest of the world and, for better or worse, the two companies are large global companies and books have long spread ideas across borders.

¹⁵Various other lag periods were also tested. Models with the globalization and globalization shock variables that were lagged more than four years found no statistical significance. This is not unsurprising given the expectation that the impact of shocks of globalization is temporary.

¹⁶GDP per capita income is sometimes criticized for failing to account for different intricacies of governments (Thies 2010). However, for the purposes of this study, it serves as an acceptable proxy variable.

¹⁷Croatia, in the midst of civil war when it emerged as a new state in the data set in 1992, was the only occurrence of a new state to experience civil war onset in the time period relevant to this study. However, both Croatia 1992 and Jordan 1970 are dropped due to lack of data.

¹⁸This estimate was calculated using Clarify Software (King et al. 2000; Tomz, Wittenberg, and King 2003). Prior war, noncontiguous state, oil producer and stability are set to their modal values, while GDP/capita, population, mountains, democracy, and ethnic and religious fractionalization are set to the means.

¹⁹Marginal effects were found using the Grinter program in Stata 10 (Boehmke 2008).