the six-year lag structure is insignificant; the feedback effect, however, is negative and significant; and in Uruguay the relationship is insignificant.
Finally, our empirical investigations for the presence of unidirectional Granger-causation from real per capita gross unilateral transfers to real per capita gross fixed capital formation, as related in Table 15.6, are negative in Bolivia, Ecuador, and Uruguay. Similarly, the sign of Granger-causal ordering from gross fixed capital formation to unilateral transfers is negative in Chile, Colombia, and Paraguay; in Argentina the relationship is insignificant. For Peru and Venezuela the direction of Granger-causality from capital formation to unilateral transfers is also negative and significant, but only at three and four annual lag lengths, respectively. In addition, results indicate the presence of significant negative feedback effects in each country at the appropriate longest lag lengths. Moreover, in Brazil the initial effect from capital formation to unilateral transfers is significant and negative, but the feedback effect is significant and negative only for a five- year lag.
This chapter presents a dynamic bivariate distributed lag causal test of the aid-cum-development hypothesis. The main conclusion of our empirical investigations is that there appears to be a noncontemporaneous and statistically significant Granger-causal relationship between unilateral transfers and the recent economic development process in several South American countries. However, the direction of causal ordering is not the favorable one predicted by the aid-cum-development hypothesis. It may be inappropriate, therefore, to consider such external flows as a modus operandi for economic development in these countries.