I have a new paper, co-authored with Andrea Dean, that empirically investigates the democratic "domino theory." Here's the abstract:
The “domino theory” has a long and impressive history guiding global foreign policy. According to this theory, increases or decreases in democracy in one country spread and “infect” neighboring countries, increasing or decreasing their democracy in turn. Using panel data that covers over 130 countries between 1900 and 2000, this paper empirically investigates the domino theory. We find that democratic dominoes do in fact fall, as the theory contends. However, these dominoes fall significantly “lighter” than the historical importance of this model suggests. Countries consistently “catch” only about 11 percent of the increases or decreases in their average geographic neighbors’ increases or decreases in democracy. We also explore whether
U.S. military intervention is capable of spreading democracy, as policymakers reasoning according to the domino theory have argued. Our results are mixed. Post-WWII occupation of Austria successfully spread democracy to its neighbors. More recent occupations in Haiti and the Dominican Republic failed to do so.
You can download the full paper here.