Agriculture Reference
In-Depth Information
programs would imply, since the payments do not depend on the allocation of input costs.
Furthermore, in the context of our model, it is not at all clear how government programs
would influence measurement costs or soil exploitability. These programs reduce income
variability, so—assuming risk aversion—one might argue that nonprogram crops be treated
differently from program crops. But, for the Nebraska-South Dakota data, virtually all crops
are program crops, so there is no way to test for effects even if implications were available.
Interestingly, the only nonprogram crop is hay, and it is virtually never cropshared, contrary
to the risk-sharing hypothesis.
Our data show that cropshare contracts are more likely when crop division costs are
low and where the ability of farmers to adversely affect the soil is high, and that cash
rent contracts often contain clauses that discourage exploitation of the soil. Our coefficient
estimates support our general theory that the variation in contracts is largely determined
by the costs of enforcing the contracts in various situations. Not only are the signs of our
estimated coefficients consistent with our predictions, but the magnitude of the coefficients
dwarf the coefficients for both the control variables and the variables testing other theories.
The next chapter extends our model to the case of input sharing and provides more evidence
in favor of the transaction cost framework.
