Estimating Parameters and Structural Change in CGE Models Using a Bayesian Cross-Entropy Estimation Approach

Go, Delfin S. ; Lofgren, Hans ; Mendez Ramos, Fabian ; Robinson, Sherman

World Bank Group, Washington, DC

关键词: ADVANCED COUNTRIES;    AGRICULTURE;    ARRANGEMENT;    BASE YEAR;    CAPITAL-LABOR RATIO;   

DOI  :  10.1596/1813-9450-7174
RP-ID  :  WPS7174

学科分类：社会科学、人文和艺术（综合）

来源: World Bank Open Knowledge Repository

【 摘 要 】

This paper uses a three-step Bayesiancross-entropy estimation approach in an environment of noisyand scarce data to estimate behavioral parameters for acomputable general equilibrium model. The estimation alsomeasures how labor-augmenting productivity and otherstructural parameters in the model may have shifted overtime to contribute to the generation of historicallyobserved changes in the economic arrangement. In thisapproach, the parameters in a computable general equilibriummodel are treated as fixed but unobserved, represented asprior mean values with prior error mass functions.Estimation of the parameters involves using aninformation-theoretic Bayesian approach to exploitadditional information in the form of new data from a seriesof social accounting matrices, which are assumed weremeasured with error. The estimation procedure is"efficient" in the sense that it uses allavailable information and makes no assumptions aboutunavailable information. As illustration, the methodology isapplied to estimate the parameters of a computable generalequilibrium model using alternative data sets for theRepublic of Korea and Sub-Saharan Africa.