Power-law distribution in the external debt-to-fiscal revenue ratio: empirical evidence and a theoretical model
Gilles Dufrenot  1@  
1 : Aix-Marseille University (Aix-Marseille School of Economics), CNRS and EHESS; CEPII; Banque de France  (AMU (GREQAM), CEPII, Banque de France)
Aix-Marseille Université - AMU, CEPII, Banque de France

This paper provides evidence that the external debt-to-fiscal revenue ratio in the emerging countries follows a power-law distribution. Such a distribution reflects the fact that external debt distress or debt crises correspond to extreme events that have been found to happen fairly often. We formally test the hypothesis of a power-law, going further than the usual visual inspection of the distribution of the variable of interest on a doubly logarithmic scale. We further show that such a distribution can be derived from a theoretical model in which uncertainty comes from tax evasion and corruption. Using the framework of an optimal stochastic growth model, we model the external debt-to-fiscal revenue ratio as a diffusion process for which the stochastic steady state distribution is derived using the properties of Itoˆ diffusion processes.


Online user: 1