Size Distribution and Growth in a Sample of Illinois Cash Grain Farms

Abstract

Falling agricultural incomes, declining exports, and emerging biotechnologies have rekindled concern over farm structure and the viability of the medium-sized (often identified as the family) farm. (See Stanton 1984 for a comprehensive discussion of farm structure issues.) Recent studies by Edwards, Smith, and Patterson; Zulauf; and the Office of Technology Assessment (OTA) examine the changing nature of the size distribution of U.S. farms. Gardner and Pope and also Leiby and Sumner investigate forces affecting farm growth. These studies are difficult to compare because of differences in models, time periods, and size measures used in empirical analyses. For example, a comparison of the OTA and the Edwards, Smith, and Patterson studies leads to very different views of farm sector structure in 2000. The OTA results, based on analysis of deflated gross value of farm income from 1969 to 1982, suggest a movement towards a bimodal distribution of very small and very large farms. In contrast, the USDA study, using acreage to measure farm size over a shorter period, 1974 to 1978, predicts a more stable distribution similar to today's. Interpretation of these results is complicated by the difficulty of defining farm size for a heterogenous set of multiple enterprise farms, the limited range of categories used to describe farm growth, and the lack of microlevel evidence to corroborate findings based on aggregate agricultural census data. The purpose of this analysis is to assess evidence of change in the size distribution of a sample of 161 Illinois cash grain farms over the past decade. Three issues are addressed in the empirical investigation. Markov transition probability matrices are estimated first, to investigate changes in distribution of the sample over the period 1976 to 1985. Second, we consider whether the results of this analysis and our predictions of future farm structure are invariant with respect to the measure of size used. Third, discriminant analysis is applied to identify factors associated with change in farm size.