Agriculture Reference
In-Depth Information
Table 1.1
Crop riskiness and share contracting
Yield coefficient of variation
Region
(% of share contracts)
Corn
Wheat
British Columbia
.27
.18
(20%)
(79%)
Louisiana
.29
.21
(62%)
(76%)
Nebraska
.12
.11
(69%)
(86%)
South Dakota
.14
.25
(64%)
(61%)
Sources:
Appendix A and Allen and Lueck (1995, 1999a).
the risk-sharing hypothesis in detail, but even a glance at the facts suggests that this is
not likely the case in modern farming. Table 1.1 shows the coefficient of variation for two
major crops (corn and wheat) in our four distinct regions as well as the prevalence of share
contracting for farming in those same regions. Contrary to the risk-sharing thesis, land
used to grow high variance crops is not cropshared more often. Table 1.1 actually suggests
the opposite: Crops with less yield variance are more often cropshared. As we show in
chapter 6, this finding is consistent with transaction cost models that focus on measurement
costs.
Our approach contrasts with the risk-sharing model.
20
In the classic risk-sharing model,
a typical model assumes that a principal maximizes some objective function subject to an
agent's incentive and individual rationality constraints. For sharecropping, most of these
models postulate a risk-neutral landowner (principal) leasing land to a risk-averse farmer
(agent). These models generate a trade-off between risk avoidance and imperfect incentives.
A principal who cash rents to an agent has no incentive problem, but the agent “bears all the
risk.” By sharing with an agent, the principal suffers from agent moral hazard, but the agent
no longer bears the full risk of the project and the payments can adjust accordingly. By using
the transaction cost approach, we avoid the empirical difficulties of risk-sharing models
while retaining other aspects (for example, uncertainty and moral hazard) of these models.
We abandon both the principal-agent distinction and the assumption of risk aversion, and
instead, we assume all parties are risk neutral.
21
By treating both parties as risk neutral, we avoid the problem of defining which party is
the principal and which is the agent, and also which party is more or less risk averse.
22
In
modern farming it is especially difficult to establish such a dichotomy because farmers and
