The integration of alternative information systems: an application to the Hogs and Pigs report
用两种计量模型整合市场信息，改进美国农业部生猪与猪肉报告的初始估计，减少误差，为农业报告可靠性提供低成本方案。
Recent federal budget cuts have reduced the sample survey coverage of U.S. Department of Agriculture (USDA) crop and livestock reports. These cuts have renewed the concerns of the adequacy and reliability of the reports, in general, and of the Hogs and Pigs report in particular. The object of this study is to demonstrate that more consistent and reliable initial estimates of key hog supply and inventory categories can be generated by augmenting the survey-based Hogs and Pigs report estimates with market information. Two alternative pork sector econometric models with alternative expectation regimes are proposed as a relatively costless means to expand the information set on which the Hogs and Pigs report estimates are based;The first econometric model is dynamic and nonlinear and incorporates the rational expectation hypothesis with forward looking expectations. The second incorporates futures market price expectations. Breeding herd decisions of pork producers are based on distant closing futures prices of live hogs and corn. Both econometric models integrate restrictions based on the biological processes of pork production as prior information in their supply components;The market information from the two econometric models are synthesized in one-step ahead forecasts of key hog supply and inventory categories. These model-based forecasts are combined with the USDA initial estimates using alternative composite forecasting techniques. The results suggest that the inclusion of market information in the data evaluation and estimation procedures can reduce errors in the USDA initial estimates. Market information often compensates for errors in the USDA estimates. Thus, econometric models and composite forecasting techniques may provide a viable, cost-effective means to improve the consistency and reliability of the initial Hogs and Pigs report estimates.