Recent cases of military intervention in Iraq, Afghanistan, Liberia, and Haiti have ignited a policy debate as to whether democratic government can be imposed externally. The results of panel (cross-temporal, cross-national) analyses indicate that military interventions have a positive effect on democratization in target states (Gleditsch, Christiansen & Hegre, 2004; Tures, 2003; Peceny, 1999a; 1999b; 1995; Herman & Kegley, 1998; Kegley & Herman, 1997; Meernik, 1996). But are these democracies stable? Drawing on our previous work on political stability and comparative analysis of the duration of different political systems (Gates, et al, 2004), we analyze the effect of military intervention and defeat in war. To estimate the survival time ratios of different political systems we utilize a log-logistic hazard function, which captures the non-monotonic nature of political stability. We find that intervention alone does not significantly affect the duration of polities. However, if the intervention is associated with military defeat, the post-conflict regime is markedly less likely to survive, ceteris paribus. We also look at the democratization effects of military intervention. For all regimes that shifted from autocracy or from semi-democracy to democracy, there is no statistically significant effect on political stability. The implication is that we can conclude that military intervention that entails defeating a country in war and imposing a regime change is politically destabilizing. Whether this new regime is a democracy or not does not make a statistically significant difference.
Also presented at the the WIDER Development Conference “Making Peace Work”, Helsinki 4-5 June 2004.