Search results

How do nation-states' political institutions affect the relations between states? This article addresses that question by testing the predictions of different theories linking political institutions to war outcomes. Specifically, rent-seeking and regime legitimacy theories predict that all democratic belligerents are more likely to win wars because they fight more effectively. Alternatively, other theories focusing on the domestic political vulnerability of leaders and the marketplace of ideas predict that democracies are likely to be more careful about choosing when to start war. This would mean that only democratic initiators are more likely to win. Analyzing all interstate wars from 1816 to 1982 with a multivariate probit model, we find that democratic initiators are significantly more likely to win wars; democratic targets are also more likely to win, though the relationship is not as strong. We also find empirical support for several control variables, including strategy, terrain, and capability. (American Political Science Review / FUB)

The literature on the democratic peace has emerged from two empirical claims: (1) Democracies are unlikely to conflict with one another, and (2) democracies are as prone to conflict with nondemocracies as nondemocracies are with one another. Together these assertions imply that the democratic peace is a dyadic phenomenon. There is strong support for the first observation, but much recent scholarship contravenes the second. This paper assesses whether the democratic peace is a purely dyadic, a monadic, or perhaps a mixed dyadic and monadic effect. Our analysis offers two important advances. First, our model directly compares the dyadic and monadic explanations by using the state as the unit of analysis rather than the potentially problematic dyad. Second, our model controls for an important but overlooked confounding variable: satisfaction with the status quo. Our results indicate that the initiation of violence within crises is predominantly a dyadic phenomenon, but we also find evidence suggesting a strong monadic effect regarding the emergence of crises. (American Political Science Review / FUB)