Evaluating Prior Beliefs in a Demand System: The Case of Meat Demand in Canada

Abstract An almost ideal demand system for meats is estimated using Canadian data. A Bayesian approach is used to impose inequality restrictions on substitution elasticities, via Monte Carlo integration and importance sampling, in order to conform with prior beliefs about curvature and monotonicity restrictions and substitution relationships. Results are more consistent with the concavity and monotonicity restrictions from demand theory than with the added restriction that all meats are substitutes.