Democratic peace theory is true

The democratic peace has proved to be one of the most influential and durable findings in the history of international relations, if not political science as a whole. As with all influential findings, alternatives have been posed, for example, the “capitalist peace” (Gartzke 2007; Mousseau 2000, 2009, 2013), the “territorial peace” (Gibler 2012), or “political similarity” (Bennett 2006; Raknerud and Hegre 1997; Peceny, Beer, and Sanchez-Terry 2002; Werner 2000). Beyond such empirical challenges, the finding has also been criticized for not providing a fully convincing causal logic that explains the empirical finding (Rosato 2003). This puzzle of what makes democracies less likely to engage in conflict with other democracies still remains subject to a lively and inconclusive debate (Hegre 2014). Our purpose here is to use the new Varieties of Democracy (V-Dem) data, with its ability to more precisely disaggregate the components of democracy, to reconsider its causal logic in a more precise fashion and with a fresh perspective. We advance the debate by arguing that different forms of constraint on power holders help to explain why democratic dyads are less likely to engage in conflict. To date, the greatest attention has been focused on what we term electoral and horizontal accountability. Electoral accountability (sometimes referred to as vertical accountability) functions because elected politicians try to avoid alienating voters in anticipation of the next round of elections. Horizontal accountability operates through the ability of other branches of government and state institutions to check the power of the executive. We build on previous work that showed that greater electoral participation (Reiter and Tillman 2002; Clark and Nordstrom 2005; Bueno de Mesquita et al. 1999) and greater horizontal constraints on the executive (Reiter and Tillman 2002; Clark and Nordstrom 2005; Choi 2010) work to reduce conflict. An important part of our contribution is to add a new dimension—social accountability. It is provided by organized actors in civil society that have the ability to inflict audience costs on the executive in-between elections and to mobilize in support of opponents during elections. We also highlight a new measurement tool for capturing democracy as an aggregate and in its component parts in new and potentially fruitful ways. Whereas the vast majority of studies have relied on Polity to capture the degree of democracy among states, we use the V-Dem data, which have both better concept-measurement consistency and also allow us, because of the ability to disaggregate components of democracy, to directly test the different forms of constraint we discussed above. Our sample includes observations from 173 countries across the globe from 1900 to 2010. We test whether these three different forms of accountability—(1) horizontal, (2) electoral, and (3) social—restrain democracies from fighting with each other. When pitting the three mechanisms of accountability directly against each other, we find the weakest support for the form that has received the greatest attention in the literature—electoral. However, its effect becomes salient when it is combined with social accountability between elections. The democratic peace effect thus seems less a product of simple electoral constraint than on the ability of civil society, as well as other state actors, to constrain the incumbent. To our knowledge, this is the first time that anyone has tested the impact of civil society engagement on conflict behavior. In an Online Supplementary File (henceforth SF),1 we demonstrate that our main findings are robust in the face of contending theories and alternate specifications of our tests.