As discussed earlier, the measures of openness and infrastructure development are on the average lower for countries in SSA (Table 4). It is therefore possible that the differential impact of these variables can be explained by a ''threshold effect'' and not a ''regional effect.'' 18 I tested this hypothesis by including quadratic terms for INFRAC and OPEN in the regressions. The estimated coefficients for the quadratic terms were not significant, suggesting that there is no second-order effect (the results are not reported). Finally, I checked the robustness of the results by examining whether the results hold when panel data is used. The results reported in Column 5 indicate that the results are robust.