Agriculture Reference
In-Depth Information
contracts over cash rent contracts, and as such, the variable YEARS DURATION could be
considered a proxy ratchet variable. The longer two parties contract with each other, the
more likely the contract involves commitment and lower powered incentives, despite the
learning done over time by the landowner. Again, the effect is small both in absolute size
and relative to other variables. The estimates in table 7.3 imply that an additional year to the
relationship, calculated around the mean of the dependent variable, leads to a 0.61 percent
increase in the probability of a share contract. Compared to the effect of having a row crop,
which increases the probability of a share contract by 56 percent at the mean, this effect is
economically unimportant.
In the context of tables 7.4 and 7.5, one might consider the YEARS DURATION coeffi-
cient along the same lines. In these cases, however, the estimated coefficients for YEARS
DURATION are not consistent with the ratchet effect. In table 7.5 the coefficient is posi-
tive and statistically significant, indicating that longer term relationships have greater share
terms. This is inconsistent with prediction 7.3 and suggests that landowners do not commit
to terms with ongoing farmers. Finally, in table 7.6, the estimated coefficient for YEARS
DURATION is not significantly different from zero. Taken together the effect of YEARS
DURATION on contract design offers little support for the existence of economically im-
portant ratchet effects in agricultural contracts.
7.4
Summary
In chapter 4 we saw that the choice between cropshare and cash rent contracts, at least for our
regions, is best explained by efforts to mitigate soil exploitation and crop underreporting.
Farmers have an incentive to overuse rented land under a cash rent contract, and a cropshare
contract curbs this incentive. At the same time, cropshare contracts create an incentive to
underreport part of the crop because the contract taxes reported output. Efforts to measure
underreporting are costly and prevent share contracts from dominating cash rent contracts
everywhere. Depending on which incentive problem is more serious, one contract or another
emerges as a second-best outcome.
Evidence to test a hypothesis like the ratchet effect is not easy to come by. Data on both
sides of the contract, as well as contract history, are required. Because agriculture and share
contracting are important theoretical applications of the principal-agent model, and because
our data meet the conditions laid out in the theoretical literature, this chapter provides an
important test. Our evidence suggests that the ratchet effect is generally unimportant in
modern agricultural land contracts. We did find that within cropshare contracts, limited
evidence exists that input and output shares move in ways consistent with ratchet effects,
although the effects are relatively small. We find virtually no evidence to support the ratchet
