The Estimation of a Polluting By-Production Technology Using Statistical Copulas

We estimate the parameters of a two-equation by-production technology defined as an intersection of two sub-technologies producing one good, and two polluting outputs using a sample of 130 countries for the period between 2000 and 2011. We use Frank, Farlie-Gumbel-Morgenstern, and Ali-Mikhail-Haq copulas in order to account for stochastic dependence between technical inefficiency components of the composite error terms characterizing the two sub-technologies. Our empirical results suggest a positive stochastic association between the technical inefficiency of producing a good output, or GDP, and that of producing carbon dioxide and methane. Introducing copulas results in significant changes in the values of the estimated parameters of the by-production technology.