Executive Constraints and Civil Conflict Onset

Sammanfattning: Do institutional constraints on a regime's executive decrease the likelihood of civil conflict onset? An unconstrained executive is free to loot state resources undermining the state's capacity to effectively deal with nascent rebellions, and may be more likely to engage in violent repression, especially in the form of indiscriminate violence. This can encourage political grievances, lead to a loss of legitimacy for the regime, and provide the opportunity for would be rebel groups to attract new members by offering protection against government violence. Furthermore, the lack of guaranteed checks to executive power may incentivise actors to avoid bargained solutions if they fear future extralegal retributions. Taken together, where there is a lack of effective constraints against the executive there may be an increased risk for the onset of civil conflict. To test this proposition I employ a bayesian latent variable model which estimates executive constraint as a latent factor derived from several manifest variables and relate it to a binary measure of civil conflict onset through a logistic regression. The primary finding is that there is a negative relationship such that the predicted probability for civil conflict is lower where there exists higher levels of executive constraints. However, this is conditional on the level of GDP per capita; for low-income country-years the relationship between executive constraints and civil conflict onset is indeterminate possibly because it is easier to recruit and mobilize fighters in such settings regardless of the overall level of executive constraints. The model, however, is a poor fit to the data meaning that the presented results should be considered tentative at best. Nonetheless, this study helps to further the work on examining specific political institutions as potential risk factors for the onset of civil conflict.