Posterior Analysis of Environmental Damage Evaluation in Europe

In this paper we consider environmental damage evaluation using two types of econometric models to analyse tree damage due to acid deposits, using cross-country data for several European countries. First, we use a set of univariate Poisson models and second, a multinomial probit model with the difference that we do not observe class specific data. The damage function is parameterized in terms of a number of economic and environmental variables. Statistical inference is conducted using Bayesian methods relying on Markov Chain Monte Carlo simulation. Natural by-products of our method include an implied ranking of countries according to environmental damage, and (exact, finite sample) posterior distributions of average country-specific environmental damage.