Meta-Regression Estimates for CGE Models: A Case Study for Input Substitution Elasticities in Production Agriculture

Abstract

The selection of appropriate parameters for computable general equilibrium (CGE) models critically affects the results of applied economic modeling exercises. Valid and reliable parameter selection models are needed, and typically comprise direct estimation, expert opinion, or copycatting of results from seminal studies. The purpose of this study is to use meta-analysis to summarize and more accurately estimate elasticities of input substitution, specifically between labor and other inputs in agricultural production. We construct a comprehensive database of elasticity estimates through an extensive literature review, and perform a meta-regression analysis to identify structural sources of variation in elasticity estimates sampled from primary studies. The use of meta-analysis contributes to improved baseline analysis in CGE simulations because it allows for the computation of input parameters tailored to a specific CGE model setup. We correct for variations in research design, which are typically constant within studies, and account for bias associated with undue selection effects associated with editorial publication decision processes. Improved accuracy and knowledge of the distribution of imputed input parameters derived from a meta-analysis contributes to improved performance of CGE sensitivity analyses.