Timothy Peterson and Leah Graham recently published a study in the Journal of Conflict Resolution showing that, after you control for the democratic peace, similarities in human rights performance have an important effect on any two countries’ likelihood to go to war. The interesting caveat is that this finding holds true for states that abuse their citizens as well as those that don’t:

Although mutual norms of domestic non-violence are more pacifying than mutual disregard thereof, the authors argue that a wide disparity in norms is more aggravating than shared norms… that norm asymmetry is aggravating provides evidence for an “‘abusers’ peace…” Our results suggest the possibility for conflicts arising between newly democratic, human rights-supporting states and their more oppressive, authoritarian neighbors… It may be that installing an “outpost of democracy” within an authoritarian region and enforcing improved respsect for human rights on the domestic population will lead to increased regional violence.

Peterson and Graham are building on two earlier studies augmenting the democratic peace thesis by exploring the specific impact of human rights performance on war. IR scholars have long noted that democratic states almost never fight one another, but there is much debate over why. Although IR liberals have long treated “ideological commitment to human rights” as one of several “pillars” or indicators of the liberal peace, Mary Caprioli and Peter Trumdore showed that human rights performance alone is actually a good predictor of interstate violence even controlling for regime type. A separate study by David Sobek, M. Rodwan Abouharb and Christopher Ingram demonstrated that states with good human rights records, were more peaceful with one another regardless of democracy. Peterson and Graham’s study extends this finding in one more direction, arguing it is indeed dyadic norms that matter, but that there exists an “abusers’ peace” as well as a “human rights peace.”

The findings themselves should be critically analyzed and replicated further: among other problems they all rely on different and imperfect indicators of human rights performance (for a critique of quantitative data-sets on human rights see this article). However as a whole this line of research suggests two modest challenges to democratic peace theory.

First, it suggests that democratic institutions per se may be far less important in mitigating interstate war than a cluster of human rights norms that can include but are not limited to the “empowerment rights” associated with democracy. Second, it suggests that the causal mechanism translating adherence to these norms into pacific relations is not liberal but rather constructivist: a set of shared identities that can constitute shared interests among human rights abusers as well as champions, lessening the likelihood of violent conflicts. And at the level of policy, these studies do indeed encourage an emphasis on diffusing norms within neighborhoods rather than changing the regimes of specific states, if the goal is to achieve both rights and security.