Quasi-maximum likelihood estimation of a censored equation system with a copula approach: meat consumption by U.S. individuals

A copula approach to censored system estimation is proposed. The quasi-maximum likelihood estimator departs from the multivariate normal error distribution predominantly used in existing estimators and resolves the computational difficulty with multiple probability integrals in high-dimensional censored systems. An application to individual meat consumption demonstrates that the procedure produces very different empirical estimates from existing Gaussian full-information and quasi-maximum likelihood estimates. Copyright (c)2008 International Association of Agricultural Economists.