Using Baltagi and Li (2002)’s semi-parametric fixed effects regression estimator, we investigate the existence of an aggregate ‘Kuznets curve’ in a sample of 113 countries over the 1960–2000 period. Our results show that misspecification of the functional form in a panel model with...

This paper shows that improvements in life expectancy (LE) had a non-linear effect on income per capita over the 1940-1980 period as this effect was conditional on each country's initial level of LE. Whereas higher LE had an initial statistically significant negative impact on income per capita...

Alfaro, Kalemli-Ozcan, and Volosovych (2008) argue that accounting for differences in institutional quality makes the Lucas Paradox disappear. We show that their key finding is driven by the presence of outliers. Once we control for them, we find that the Lucas Paradox remains.