On the Debt-Growth Relationship and Related GMM Issues

This article examines the effects of public debt on aggregate growth and some of its main channels utilising a larger dataset than those employed in most pre-COVID studies. We build a conditionalconvergence framework by performing a panel analysis that accommodates the autoregressive behaviour of the response variable. We discuss the difference-GMM and system-GMM techniques, focusing on critical issues inherent to such methods within the context of the debt-growth analyses, especially as concerns the internal instrument proliferation that typically arises when T is large. The estimates suggest that public debt adversely affects aggregate growth, productivity and capital accumulation, especially over the five-year horizon. However, the marked differences between advanced and emerging economies, along with crucial methodological issues typically associated with GMM approaches in macroeconomic frameworks, prompt crucial questions about the reliability of such techniques in analysing the debt-growth nexus and in generalising the resulting findings and policy recommendations.