Edward Mansfield and Jack Snyder's article “Democratization and War” (Foreign Affairs 74 [May/June 1995]: 79–97) arrived in 1995 like a thunderclap, riveting me to questions about domestic political changes and their implications for peace and conflict between states. The article appeared when the intellectual wars over the democratic peace were hot, and when research critical of democratic performance in general was receiving heightened scrutiny. In general, the early response to the authors' claim that democratization significantly increases the war-proneness of states focused narrowly on their research design. However, my review of Mansfield and Snyder's book-length treatment of the subject focuses on the broader weaknesses in the book, of which there are few, and the book's strengths, of which there are several.
In Electing to Fight, the central puzzle is as follows: “[W]hy do citizens in well-established democratic systems vote for governments that rarely wage war upon each other, while electorates in transitional democracies so often support aggressively nationalist policies, even against democracies?” (p. 23). The answer, the authors argue, is that while democratic voters in newly democratizing states do not start out preferring aggressive foreign policies, the incapacity of nascent democratic institutions to channel popular demands effectively makes democratic voters receptive to political distortions and, perhaps, calls to engage in war abroad.
For Mansfield and Snyder, the central link between democratization and causing war is the process of incomplete democratization, in which a volatile mixture of weak political institutions, protective elites, and popular involvement in the political process increase the attractiveness of aggressive foreign policies. In short, elites in the former authoritarian regime seek to slow the democratization process, a process that would ultimately weaken elite prerogatives. Appeals to nationalism centered on notions of national security enable elites to accomplish the twin goals of finding common cause with “the people” and overempowering the executive branch in an effort to restrain democratic reform.
The authors test their claim that incompletely democratizing states are more war-prone with monadic- and dyadic-level data for a global sample of states covering the modern interstate system (1816–1992), and in doing so they report substantial evidence that incompletely democratizing states are significantly more likely to engage in wars abroad. In turn, the authors “process trace” cases of war initiation by democratizing states, a method enabling them to gauge the relative importance of the causal mechanisms that their theory anticipated to be important precursors to war, as well as to explore false-positive predictions produced in their quantitative analysis. Additional qualitative analysis facilitates their exploration of six international wars that occurred between 1992 and 2000, the period absent from their large-n quantitative analysis. In general, the qualitative analysis supports the authors' theoretical claim regarding incomplete democratization.
There are few outright weaknesses in the volume, but three spring to mind. First, little attention is devoted to the origins of domestic regime change. It seems plausible that whether a political change is indigenous or imposed from abroad might have important implications for the democratization and conflict link, as well as for policies developed to minimize the strength of this link. Second, a survey of the appendix, which reports cases of complete and incomplete democratization and war (pp. 285–87), suggests some striking regional patterns. For example, not a single case of democratization, either complete or incomplete, occurs in Latin America or Africa when the authors rely on the Polity III combined regime index, a frequency that increases only slightly when the additional measures of regime change are calculated. A lengthier exploration of the potential theoretical implications of these patterns seems warranted. Finally, although the process-tracing methodology is instructive, it is not abundantly clear why one benefits from tracing only cases of democratization wherein war initiation occurred, rather than also examining a representative subsample of cases in which democratization occurred but war initiation is absent.
These issues aside, the volume exhibits several strengths. First, it remains provocative, even 10 years after the initial article in Foreign Affairs. To my mind, this provocation stimulates critical thinking about the research question, as well as many other research-related issues. Second, it is firmly grounded in the dynamics of political institutions, rather than the static approach that characterizes many studies of conflict. Emphasis on the dynamic qualities of domestic political institutions reinforces an important element of the scientific study of peace and war: Virtually every cause of war is cast in probabilistic language and estimated empirically as such. Thus, powerful states deter weaker challenges from initiating wars, but in some instances, weak states do attack stronger targets. The democratic peace is presented in similarly probabilistic terms. During regime transitions, the very institutional and normative factors that enable democratic pairs of states to avoid open warfare in resolving disputes might fail to prevent the occurrence of interdemocratic hostilities. Mansfield and Snyder's case study of the Kargil War between Pakistan and India certainly suggests as much.
Third, the volume provides an excellent illustration of the way in which scholarship can inform policy in a meaningful way and in a timely fashion. Specifically, if weak democratic institutions do increase the proclivity of new democracies to resort to war, then the current democratic community should seek to fortify fledgling democracies and prevent the resort to arms. Fourth, Mansfield and Snyder demonstrate that the emphasis on regime change in the comparative literature and on war and peace in the international relations literature are complementary, thereby underscoring the need to draw knowledge from across subfield boundaries to answer research questions that are of contemporary importance. Finally, their volume demonstrates the benefits of relying on research designs that combine qualitative and quantitative methods of theory exploration and hypothesis testing. Even if one does not subscribe to the process-tracing methodology employed here, the juxtaposition of large-n and small-n analysis in the same study is likely to be beneficial.
In sum, Electing to Fight raises a number of important issues both explicitly and implicitly that warrant further consideration. In particular, if we wish to encourage the proliferation of democracy, we should keep in mind that, much like economic development, the road to the ideal type is likely to be bumpy, even jarring, at times. Mansfield and Snyder encourage one to think about the dynamic systems that underlie this journey, and in turn, how scholars and policymakers alike might smooth this path together.
