ABSTRACT Using data on 167 countries over 1970–2019, this study gathers new evidence about the government debt-growth relationship. We jointly address several features of the panel – foremost asymmetry, cointegration, endogeneity, country heterogeneity and cross-sectional dependence. We find that increases in per-worker debt are detrimental to the long-run dynamics of output, but our nonlinear estimates suggest that changes in government debt propagate their effects through different channels. Applying alternatively gross and net public debt as a measure of indebtedness yields different results too, as happens with the adoption of different estimators. The magnitude of the effects depends on the model specification, indicating that cross-country estimates of the debt-growth nexus require parsimonious interpretations.
Read full abstract