The over-all purpose of this study is to investigate problems of aggregation error in supply estimates. Specific elements of the analysis include an exploration of the theoretical aspects of developing error-free or minimum-error aggregates, the development of empirical supply estimates for Iowa pork and beef (based on different stratifications of representative farms), determination of the relative magnitude of the aggregation error and possible factors contributing to it, and recommending practical research procedures for controlling aggregation error.

The grouping implied by Theorem I developed in the study extends beyond the practice of grouping farms by restrictive resources. Although grouping by restrictive resources is important, it does not reflect differences in the response patterns of individual farms. The requirements of Theorem I provide a basic principle to follow in controlling aggregation error.