| Abstract: | The purpose of this study was to develop a method of forecasting the most likely consequences of public policy decisions on the economic well-being of domestic producers of food and natural fibers. To do this a new macroeconomic model, called GEM, was developed. While other models exist which attempt to evaluate the impacts of national policies on agricultural production, recent literature suggests that none of these models captures all the theoretically important linkages between agriculture and the rest of the economy. GEM, however, as derived from a general equilibrium framework, and thus incorporates the following linkages between agriculture and other economic sectors: (1) farm input markets, (2) markets for farm products, (3) impacts of agricultural and non-agricultural imports and exports on domestic markets, (4) effects of general economic conditions on off-farm employment opportunities for farm operator families, and (5) influences of general economic conditions on the cost of loan funds to agriculture. To answer positive economic questions, GEM was developed using econometric methods. A set of conceptual variables needed to estimate a theoretical model--developed without consideration of data availability--was identified and compared to the existing federal data base. Several "data gaps" were identified. While some of these "missing" variables were approximated by time series constructed from existing data, assumptions and/or the use of proxies, other data gaps forced modifications of the original theoretical model in order to derive the set of equations estimated in this study. The results from estimating the equations using the ordinary least squares (OLS) and the iterative instrumental variables (IIV) estimators were compared. No evidence was found that the IIV estimates were an improvement over those found using OLS. Given the higher costs of modifying the model associated with the use of the IIV estimator, the coefficients generated by the OLS estimator were chosen as "best" for use in this study. A simulator was developed using the Gauss-Siedel technique to solve GEM's 136 simultaneous equations. The General-Analytical-Simulation-Solution-Program, a package of computer programs providing many functions for running simulations, was used as the basis for constructing the GEM simulator... |
|---|