3
will extend the empirical evidence already developed in the literature pointing to the
emergence of the democratic peace at the turn of the twentieth century (Gowa 1999;
McLaughlin Mitchell, Gates, and Hegre 1999; Cederman 2001; Cederman and Rao 2001;
Russett and Oneal 2001). While this body of work has confined itself to the systemic and
dyadic levels of analysis, I will first present similar results confirming this conclusion and
then extend this finding with a monadic research design. These tests confirm two key
findings: democratic states were more aggressive than their autocratic counterparts in the
nineteenth century; and a crucial change in the relationship between regime type and
foreign policy occurred in the aftermath of World War I. But perhaps more importantly,
this evidence creates not only a new empirical question to be answered, namely why were
democratic regimes aggressive in the nineteenth century, but it also offers a means to
evaluate the relative strength of existing explanations for the democratic peace. A
discussion section examines how most theories of democratic peace cannot yet account
for this important historical break in the behavior of democracies.
The presentation found in this paper proceeds in the opposite fashion of most
research papers. Instead of beginning with a theoretical review, it leads with a series of
statistical tests, across multiple research designs, demonstrating that a radical shift in the
foreign policy behavior of democratic states occurred in the aftermath of World War I. A
robust democratic peace did not emerge until after 1919. It then evaluates some of the
most prominent explanations for a democratic peace in light of this important empirical
shift. Third, I offer some suggestions and clues to this puzzle while acknowledging that
much work still remains to be done.
Convention