The Millennium Development Goals call for a two-thirds reduction in under-five mortality rate. Can democratic reforms contribute to this goal? The focus of previous studies has mainly been on verifying the existence of a relationship between democracy and child mortality and not on the dynamics of the relationship. This paper addresses this question by empirically testing the dynamic effects of important changes in the level of democracy on the percentage change in child mortality using a distributed lag model. The findings are that during the 5 to 25 years following a democratic transition child mortality decreases significantly. Following this decrease, child mortality stabilizes at a new, lower level. Disaggregating democratic transitions into different subcomponents the finding is that the single most important aspect for child mortality is the competitiveness of executive recruitment. The results on the effects of an autocratic experience mirror those of a democratic transition, child mortality increases for a number of years but the results are not as stable as for positive democratic changes.
