CeBER Working Paper No. 2017-7

The study of the relationship between public debt and economic growth came again to the spotlight with the financial crisis (2007-2008) and with the sovereign debt crisis that followed in Europe. This literature aims to shed light about the sign, magnitude, mechanisms and threshold regimes relating debt to growth and to make policy recommendations with important consequences in terms of government’s policies.

We empirically investigate this relationship for a group of 60 countries for a long-time period (the shorter one from 1970 to 2012) using the historical public debt database (HPDD) built by the International Monetary Fund (IMF) and we defend that the empirical strategy underlying most of the studies on this topic should be revised. We claim that: a) the study of the long-memory property of the public debt GDP ratio and stationarity (using the last generation tests) has to be performed as a first step of the empirical analysis, what has been done using 87 countries; b) In the presence of a non-stationary public debt GDP ratio cointegration analysis was used to estimate the relationship between the public debt GDP ratio and output; c) under the no rejection of the null of no cointegration, the above mentioned relationship was studied between the public debt GDP ratio first difference and GDP growth rate using threshold models and searching for thresholds using a wide variety of variables.

The main conclusions of this study are that the debt series have a long memory and should not be analyzed in a short-term framework; additionally, the non-stationarity of the debt series does not allow researchers to apply stationary econometrics methods to model its behavior. This finding implies that the relationship between economic growth and national debt that has been characterizing the literature on the subject, has disputable econometric foundations. We thus recommend our empirical strategy to overcome the above-mentioned drawbacks of the existent empirical literature. Finally, it should be mentioned that the relationship between the public debt GDP ratio first difference and GDP growth rate is always negative despite the different threshold regimes identified.