The purpose of this paper is to examine the dynamic and long-run relationships among public debt, FDI and output growth in five individual Caribbean countries over the period 1975–2015.
Zivot and Andrews (1992) unit root test with structural break is used to examine the stationarity of the variables and then the autoregressive distributed lag bounds testing procedure is used to ascertain existence of cointegration among them. Finally, order-invariant generalized forecast error variance decomposition (GFEVD) is used to establish the strength of the causal relationship between the examined variables.
The results confirm that the examined variables are cointegrated. FDI, domestic investment, trade openness, human capital (HC) and institutional quality were found to have significantly positive effects on economic growth, while higher public debt and inflation rates hampered growth. GFEVD revealed unidirectional Granger causality running from FDI to economic growth in two countries; unidirectional causality from growth to FDI in two other countries; and bidirectional causality between growth and FDI in one other country. The results also indicate one-way causality from output growth to public debt in three countries and bidirectional causality between these two variables in two other countries.
The implication is that the Caribbean Governments may need to adopt effective debt management as a major policy and intensify efforts at utilizing loans obtained judiciously for human and capital projects that have direct positive net present value but, to secure strong and inclusive growth, these strategies must be linked to policies that enhance macroeconomic stability and the quality of their institutions, encourage capital inflows and domestic investments vis-à-vis domestic savings, and increase HC and trade earnings.
In contrast to extant studies of the public debt–FDI–output growth nexus, this study controls for the possibility of structural breaks in unit root tests along with performing bounds test for cointegration, variance decomposition analysis, Granger causality tests, and CUSUM and CUSUMSQ tests for the stability of the dynamic output growth model. This is a unique contribution to the existing literature, and highlights the originality value of this paper.
