Sovereign debt default: Are countries trapped by their own default history?

Abstract

Why are sovereign debt defaults so persistent in some EMEs, even at relatively low levels of external debt? The empirical literature has argued that the country s record of defaults is the main determinant of the future default risk. However, there are two factors generating the e¤ect from history on the probability of default: state dependence and unobserved heterogeneity. Is a country more likely to default because it has experienced a default in the past (state dependence) or does the country have some previous speci c characteristics that make it more prone to default (unobserved heterogeneity)? Results indicate that state dependence e¤ects are large. Nevertheless, this paper presents evidence indicating that the omission of unobserved heterogeneity -which accounts for both unobserved and observed time invariant characteristics- has drastic consequences when assessing countries risk of default. When unobserved het- erogeneity is accounted for there are countries with high risk of default even if negligible levels of debt are assigned to them. Conversely, other countries show a low probability of default even with assigned levels of indebtedness higher than those observed in the sample. Finally, this paper presents evidence suggesting that unobserved heterogeneity could be associated to a set of di¤erent historical, political, and cultural factors that have deeply and persistently shaped institutions.