Working Paper Series: Special Edition of 2016 to 2018 Interns

additional data and information that are necessary for better inferences. The forward orthogonal method subtracts the estimates from their actual mean values while keeping data points consistent, rather than losing data through differencing and finding lag differences. 5.3 Empirical Results The system GMM estimated regressions passed all the specification tests, except for the one highlighted in the column for Model 2-Latin America in table 7 . The null hypothesis for no first order serial autocorrelation was not rejected at the five percent level of significance. Likewise, regressions are not impinged by simultaneity bias, as orthogonality conditions rendered results failed in rejecting the null hypothesis; that the error term is uncorrelated with the instruments used. In light of these result of both the Hansen and Arellano-Bond tests, the parameter’s estimates resulting from models are consistent and empirically non-biased to be analysed and discussed. 5.4 Discussion and Analyses of Results Much can be discerned from the results presented in tables 6 and 7 . Column 1 of tables highlight the short name of variables used, column 2 report results for the whole Latin America and the Caribbean region (without considering the exchange rate regime), columns 3 and 4 report results for the fixed and floating exchange rate economies respectively, while columns 5 and 6 show estimation results for Latin American and Caribbean countries receptively. One important thing to note from results garnered from the SGMM is that estimates are mostly significant in both per capita GDP models for the region as a whole, and the same is true for economies operating under a fixed exchange regime. Hindrances to Regional Growth : Estimates in all models show that both remittances and population growth have a significant negative impact on the economies in the region. This means that remittances inflow to the region fit the pure altruism theoretical framework that coincides with its counter-cyclical property highlighted by Ukeje et al., Feeny et al., Chami et al. and others. Likewise, results show that countries with higher population growth rates record slower growth patterns than the smaller populated countries. These results coincide with the Solow (1956) convergence theory,

136