Abstract
Two political risk variables (level of demaocracy and political instability) are created and included in a probit model of rescheduling. Both the coefficient estimates and the forecasting performance strongly support the importance of including quantified proxies of political events in the assessment of overall country exposure faced by international lenders. An inverse relationship between rescheduling probabilities for a given country and its level of democracy is found, while a direct relationship is established between the rescheduling probabilities and the level of political instability.