Re-examining the saving-investment nexus: threshold cointegration and causality evidence from the ECOWAS

The aim of this paper is to re-examine the relationship between saving and investment in the case of ECOWAS countries. We then employ the Gregory and Hansen (J Econometr 70:99?C126, 1996a; Oxford Bull Econ Stat 58:555?C560, 1996b) threshold cointegration tests and the Toda and Yamamoto (J Econometr 66:225?C250, 1995) causality procedure. Threshold cointegration approach allows estimating changes in the saving-retention coefficients over time. Using yearly data for the period from 1960 to 2007, we show out two groups of countries regarding the cointegration test results. In the first group (The Gambia and Togo), capital is perfectly mobile since there is no long-run relationship between savings and investment. The second group comprised of the ten other ECOWAS countries studied, is characterized by an incomplete capital mobility. Moreover, in the second group the saving-retention coefficient is time-varying in most of the ECOWAS countries (except for Benin and Niger) and is different from a country to another one. The changes in the saving-retention coefficients indicate how international capital mobility has changed. Moreover, the inclusion of trade openness and country size in the saving-investment equation reduces the saving retention coefficients implying improvement of capital mobility over time. Furthermore, investment rates cause saving rates in Benin, Guinea-Bissau and Niger while saving rates cause investment rates in Cote d??Ivoire, Nigeria and Sierra Leone.