The Terms of Agricultural Contracts: Theory and Evidence

H. Peyton Young and Mary A. Burke


Standard contract theory predicts that contractual terms between tenants and landlords should reflect the quantity and quality of factor imputs (labor and land), the degree of risk aversion of the parties, and various structural features of the labor market such as alternative wages, turnover rates, and search costs. We examine these theories in the light of data on the terms of contracts, as well as the resulting real returns to labor and land, in contemporary Illinois agriculture. Contrary to standard theory, contractual terms exhibit a very high degree of uniformity in spite of demonstrable variations in underlying fundamentals. In particular, contract coalesce around certain regional norms that are insensitive to local variations in land quality. A particular implication is that labor of higher quality land captures a sizable portion of land rent: about one-third of the incremental surplus from farming more productive soils goes to the tenant, even after adjusting for differences in labor quality, riskiness in returns, and security of tenure.



Cropshare tenancy has been a source of controversy in the economics literature since classical times. Aside from debates concerning the relative merits of cropshare versus fixed-rent tenancy, there is the vexing matter of the shares themselves: almost all accounts of cropsharing practices, in both developed and developing countries, find that tenants and landlords divide the crops according to simple fractions, such as one-half, three-fifths, or two-thirds. Futhermore, these shares seem to respond only weakly to observable differences in economic fundamentals, such as differences in land quality, reservation wages, riskiness in returns, and the like. In sum, contracts appear to be too uniform given the degree of heterogeneity in economic fundamentals (Bardhan, 1976; Bardhan and Rudra, 1980, 1981, 1986a, 1986b; Murrell, 1983, Stiglitz, 1989). Is this conformity effect an illusion that would go away if we could examine real as opposed to nominal returns? Or is it a genuine effect, and if so, what are its implications for factor returns?

Previous studies of this question have focused for the most part on village-level data in developing countries (see, in addition to the above references, Bliss and Stern, 1982; Bardhan, 1984; Robertson, 1987; Lanjouw and Stern, 1998). Typically these studies find that the nominal shares cluster around a few focal values, and that within each village the shares exhibit a high degree of conformity, whereas among villages the terms may differ substantially. But none of these studies has addressed the crucial issue of whether these peculiarities in the nominal terms flow through to the bottom line and distort the factor returns to labor and capital. In this paper we examine this issue using survey data from Illinois that contains information on net returns to labor and land in addition to nominal contract data.