Regime-switching effects of debt on real GDP per capita the case of Latin American and Caribbean countries

In this paper, we try to investigate how the debt and real GDP per capita relationship varies with indebtedness levels and other country characteristics in a balanced panel of 21 developing Latin American and Caribbean countries over the period 1992–2006. The empirical results indicate that there exist two threshold values of 32.88% and 55.89%. The latter is lower than the Maastricht criterion and Stability and Growth Pact of a total external Debt per GDP ratio at 60% in the OECD countries. Both thresholds divide our panel into three regimes. In the middle (stimulus) regime, the Debt per GDP ratio has a positive impact on real GDP per capita, which is consistent with the stimulus view (Eisner, 1984). However, the impact becomes negative and consistent with the crowding-out view (Friedman, 1977, 1985) in the left and right (crowding-out) regimes. Based on our findings, we find no supportive evidence for Ricardian view (Barro, 1989). Therefore, our empirical results have important implications for fiscal policymakers in these Latin American and Caribbean countries.