In this study the causality relationship among variables in chicken and beef markets were investigated based on annual data from 1974 to 2004 in the I.S. of Iran. For this purpose, causality algorithms emerging from directed acyclic graphs were used in two cases, one based on co- integration analysis and innovation correlation matrix of Vector Error Correction Model (VECM) and the other using data directly. In the investigation of causality process, PC algorithm based on partial correlations and GES (Greedy Equivalent Search) based on Bayesian network model were used. The results revealed that there is no specific causality relation between chicken and beef consumption and their price indices. Thus it seems other variables and government interventions are effective factors in these markets. Therefore, this model is an instrument for forecasting changes of these variables. In these markets there are causality relations among price indices, quantities of consumption and other variables. Beef price index as endogenous variable, is under the effect of chicken price index, non meat food price index, non food price index, per capita expenditure and quantity of chicken consumption. The quantity of beef consumption is predetermined and isn’t under the effect of other variables. Chicken price index is endogenous and under the effect of beef price index, per capita expenditure and non food price index. Also chicken quantity consumption is endogenous and under the effect of beef quantity consumption and non meat food price index. With respect to the findings, for effectiveness, policy initiatives aimed at improving in meat industry should be different for different meat markets and the method of directed graphs can be used a s a guideline.

Keywords: Directed Graphs (DG), Co-integration analysis, Endogeneity testing, Meat demand.

Full-Text [PDF 259 kb]   (3190 Downloads)