The Determinants of Government Debt: An empirical analysis, 1995-2017

This data-rich study analyses the determinants of government debt based on a balanced panel dataset of 140 countries from 1995 to 2017. The dependent variable stems from the IMF Global Debt Database. The empirical approach to the analysis of this work is pursued with a linear Fixed-effects model and a linear dynamic panel-data estimator (System-GMM) model to encounter endogeneity issues. The dataset is divided into five broad subgroups, namely income groups, debt categories, regime type, economic country groupings and lending programs. The dataset is further split into pre- and post-crisis periods to account for evident changes in macroeconomic stability in the aftermath of the financial crisis. Key aspects tested in this work are the effect of annual GDP per capita growth, trade openness, domestic investment and state fragility on government debt. The main findings are that higher GDP per capita growth is an effective measure of reducing the government debt ratio in any setting. Trade openness increases government debt in high-income economies and decreases debt for autocratic governments, highly indebted economies and countries involved in an official lending program. Domestic investment programs are highly effective in reducing public debt burden in high income economies, among highly indebted governments and in stable economic environments also for countries involved in a lending program. Higher FDI inflow has a decreasing effect on government debt for economies receiving international development assistance as well as for least developed countries.