We report the regression results of the models presented above in Table 2. Before discussing the results, there are a couple of comments worth making about computational issues. In addition to the visual evidence of probable heteroskedasticity, we formally tested whether heteroskedasticity is present in our data sets. Breusch-Pagan test rejects the null hypothesis of homoskedastic errors. In Table 2, for each model, we present estimates of heteroskedasticity- robust standard errors. We checked the variance inflation factors (VIFs) for each of the explanatory variables. VIFs suggest that multicollinearity is not present in our data sets.