The Usefulness of Aggregated Data in the Analysis of Farm Income Variability and Resource Allocation

ITH the steady increase in commercialization and size of the economical farm unit, the question of financing farm transfer also increases in importance. Further, with larger and larger amounts of operating capital required, shortand intermediate-term loans to farmers gain significance. While farm size and the importance of outside capital to the farm firm are growing, there is also a trend toward enterprise specialization. The latter may lead to year-to-year farm income fluctuations of considerably larger magnitude than is expected on highly diversified farms. As a consequence, it is possible that highly commercialized, large and specialized farms decrease their ability to meet loan commitments while their need for outside capital increases. Studies which address themselves to this question' are largely based on aggregated (county, economic area, state) data. It is the purpose of this paper to demonstrate that the use of aggregated data in such studies is likely to lead to wrong conclusions. This will be done 1) by first sketching the procedure for variance estimation and the statistical theory involved; 2) by presenting empirical estimates of enterprise variation; and 3) by comparing variance estimates derived from aggregated (county) data with estimates derived from individual farm data. These variance estimates will then be employed in computing variances of differing enterprise combinations in order to demonstrate the difference in results obtained from the aggregated versus the individual farm data.