The idea that public debt has a negative impact on economic growth has been very popular in literature, and it has been the guide towards the adoption of policies of debt reduction. Through empirical and theoretical models, many studies have attempted to provide support for this claim and for the existence of general debt thresholds above which debt would negatively affect growth. However, because of heterogeneous conditions, it is unlikely that such thresholds could be generalized to any country and any period. Going beyond the estimation of debt-thresholds, our work aims to deepen the relationship between public debt and economic growth by analysing a slightly unbalanced panel data set with 25 Western and Eastern European countries, with quarterly data from 1999Q1 to 2015Q4. After testing for unit roots, we perform a cointegration analysis both at country and panel level to understand for which countries a cointegration relationship between GDP and debt exists, and thus for which countries of our data set a long-run relationship between these variables can be described. Finally, we specify a panel model to compare our groups of countries, and we decompose GDP into its main subcomponents to find out whether or not cointegration arises at the aggregate level only. Our findings show that a) a long-run relationship between public debt and GDP exists for some countries but it cannot be generalized, b) the debt-to-GDP ratio does not always correspond to this long-run relationship, and c) the short-run linkage between public debt and GDP is negative, but it is mainly determined by the events that followed the financial crisis.

Key words: Economic growth, Public debt, Cointegration analysis