Aggregate versus disaggregate analysis: corn and soybean acreage response in Illinois
1994
Park, W.I. | García, P.
Tradeoffs between aggregation and specification errors and their implications for forecasting and policy analysis make it difficult to determine the most appropriate level of aggregation for econometric analysis. Aggregation introduces bias in parameter estimates, but may reduce the effect of specification errors in the micro-level relationships. Here, tests and procedures for assessing the tradeoffs are applied to corn and soybean acreage response equations in Illinois. The results indicate that modeling acreage response in Illinois at the state level, rather than at the crop reporting district level, leads to little loss in predictive accuracy and provides parameter estimates more consistent with expectations.
显示更多 [+] 显示较少 [-]