Using the specification in Column 2 as my basic model, I test for robustness by including economic variables (government consumption, the inflation rate, financial depth and growth rates) and a measure of political risk. The results reported in Column 3 shows that the basic model is robust to changes in specifications. Furthermore, the economic variables and the measure of political risk are not significant. 15 The insignificance of the estimated coefficient of the political risk variable agrees with the findings of Edwards (1990), Jaspersen et al. (2000) and Hausmann and Fernandez-Arias (2000). Indeed, this result is not surprising. For example, in 1998 and 1999, Angola, a highly unstable country ranked first in FDI flows among SSA countries (UNCTAD, 2000). A plausible explanation is that FDI to Angola (which is mostly in petroleum) is so profitable that the return after adjusting for risk is quite substantial. It is interesting to note that the Africa dummy remains significant after controlling for a wide range of factors. 16 This indicates that there is an unaccounted for “Africa effect'' -suggesting that the inability of countries in sub- Saharan Africa to attract FDI may be partly blamed on the fact that these countries are located in a continent that happens to have a bad reputation.