The Political Economy of FDI 609 of democracy to employ this empirical technique.36 The standard OLS estimates are presented in column 1 and the selection-corrected estimates are presented in column 2.The standard OLS regression yields very similar results as in Table 6: democratic regimes attract roughly 0.37 percent more FDI as a percentage of GDP (DICTATORSHIP)than their authoritarian counterparts.With the Heckman selection model,one finds that a significant selection bias exists in the OLS results.57 In this case,one has vastly underestimated the effects of democratic governance on FDI inflows.When these selection effects are taken into account,democratic govern- ments attract almost a full I percent more FDI as a percentage of GDP per year! These selection effects are explained by the size of FDI inflows to developing countries.Although the majority of raw FDI is between developed countries,when measured as a percentage of GDP,developing countries attract the highest amount of FDI in the sample.These developing countries also tend to be authoritarian,or at least more authoritarian than the developed countries in this sample.This can lead to a spurious correlation between authoritarian regimes and high levels of FDI inflows,when in actuality it is lower levels of economic development that explain the high levels of FDI as a percentage of GDP. Essentially,the standard OLS regressions have understated the effects of democ- racy on FDI.The OLS regressions ignore the fact that poor countries attract higher levels of FDI as a percentage of GDP,and that poor countries also tend to be authoritarian.When these selection effects are controlled for using a standard Heck- man selection model,one finds that the unbiased positive effects of democratic institutions on FDI inflows are even more massive than reported in the OLS estimates. Democracy and Sovereign Debt Risk Although this article argues that the informational and representational character- istics of democratic systems have positive effects on FDI inflows,the greatest ben- efit is how democratic systems increase the credibility of political leaders in the eyes of international financial markets.In the previous sections I examined the effects of democratic institutions on levels of FDI inflows.According to my theory, democratic institutions should decrease the potential risks of government leaders choosing policies that negatively affect multinational operations,leading to higher levels of FDI. To empirically examine the causal mechanism leading democratic institutions to higher levels of government credibility,I examined how democratic institutions affect the sovereign debt ratings of governments.Granted,this is not a direct test 66.Alvarez et al.1996. 67.Both Lambda and the likelihood ratio test confirm the significance of the selection model.See Table 7.of democracy to employ this empirical technique+ 66 The standard OLS estimates are presented in column 1 and the selection-corrected estimates are presented in column 2+ The standard OLS regression yields very similar results as in Table 6: democratic regimes attract roughly 0+37 percent more FDI as a percentage of GDP ~dictatorship! than their authoritarian counterparts+ With the Heckman selection model, one finds that a significant selection bias exists in the OLS results+ 67 In this case, one has vastly underestimated the effects of democratic governance on FDI inflows+ When these selection effects are taken into account, democratic govern￾ments attract almost a full 1 percent more FDI as a percentage of GDP per year! These selection effects are explained by the size of FDI inflows to developing countries+ Although the majority of raw FDI is between developed countries, when measured as a percentage of GDP, developing countries attract the highest amount of FDI in the sample+ These developing countries also tend to be authoritarian, or at least more authoritarian than the developed countries in this sample+ This can lead to a spurious correlation between authoritarian regimes and high levels of FDI inflows, when in actuality it is lower levels of economic development that explain the high levels of FDI as a percentage of GDP+ Essentially, the standard OLS regressions have understated the effects of democ￾racy on FDI+ The OLS regressions ignore the fact that poor countries attract higher levels of FDI as a percentage of GDP, and that poor countries also tend to be authoritarian+ When these selection effects are controlled for using a standard Heck￾man selection model, one finds that the unbiased positive effects of democratic institutions on FDI inflows are even more massive than reported in the OLS estimates+ Democracy and Sovereign Debt Risk Although this article argues that the informational and representational character￾istics of democratic systems have positive effects on FDI inflows, the greatest ben￾efit is how democratic systems increase the credibility of political leaders in the eyes of international financial markets+ In the previous sections I examined the effects of democratic institutions on levels of FDI inflows+ According to my theory, democratic institutions should decrease the potential risks of government leaders choosing policies that negatively affect multinational operations, leading to higher levels of FDI+ To empirically examine the causal mechanism leading democratic institutions to higher levels of government credibility, I examined how democratic institutions affect the sovereign debt ratings of governments+ Granted, this is not a direct test 66+ Alvarez et al+ 1996+ 67+ Both Lambda and the likelihood ratio test confirm the significance of the selection model+ See Table 7+ The Political Economy of FDI 609