Technology, labour regulation, and nonparametric panel data modelling

We study the relationship between technology and labour regulation using rich annual country-level large panel data. We depart from previous studies by questioning the assumptions of linearity and cross-sectional independence and exploit recent advancements in semi-nonparametric panel data econometric methods. Alternative proxies for technology as well as various measures of labour regulation are considered. This work refines previous results by showing threshold effects, nonlinearities and complex interaction effects that are obscured in parametric specifications and that have relevant implications for policy. Our findings also highlight that standard approaches that adopt parametric formulations or do not consider cross-sectional dependence are seriously biased.