Previous work has documented a negative correlation between internal conflict and state capacity. We attempt to shed light on mechanisms that underlie this relationship, using data for Colombian municipalities. We rely on identifying heterogeneous effects of different types of violent events on state capacity, taking advantage of variability across municipalities in the prevalence of specific manifestations of conflict and their intensity. Our findings suggest that events making civilians feel targeted affect the state’s capacity to collect taxes, while those reflecting a stronger military capacity of illegal armies, in particular their large-scale attacks, affect the state’s capacity to provide public goods.

We re-examine the view that wars make strong states, taking advantage of panel data to address two of the most obvious endogeneity concerns that arise in this context: initial conditions and persistence of state capacity. Our main message is that, in modern times, there is no evidence that wars lead to strong states. In contrast to findings for earlier periods, our results show that external conflicts have displayed a negative correlation with traditional measures of state capacity in recent decades, which becomes insignificant after controlling for initial conditions and the persistence of state capacity. As in previous work, we find a negative capacity- internal conflict correlation, robust to controlling jointly for initial conditions and persistent effects.