Agriculture Reference
In-Depth Information
attributes (such as soil fertility). In other problems the various inputs may be seed, fertilizer,
pesticide, or other nonland nonlabor inputs; labor time as opposed to effort; the number of
tasks; and so on. The complexity of the models progresses throughout the topic. The chapters
on contract choice are simpler because they suppress issues of timing and specialization
and focus on moral hazard and enforcement-monitoring costs. Later, these other aspects
are introduced to discuss ownership types and vertical integration.
As these complications are added, the specific form of production function in equation
(1.1) alters, but regardless of the specific form, our production function contains in it the
basic structure necessary for the existence of transaction costs. Notably, output is determined
by human action
is
observable, it is a complex asset and cannot be perfectly measured. As a result, effort can
be altered to suit the private interests of one party at the expense of the other. For example,
low levels of output that result from low labor inputs could be blamed on poor weather
conditions. Suboptimal level of inputs might also include applications of the inputs at an
incorrect time. The simple additive uncertainty component also simplifies the analysis in
models that maximize expected values.
In addition to this basic structure, we also make several assumptions regarding the func-
tion
h(
·
)
and nature
θ
. Inputs are not observable, and although the output
Q
always has positive but diminishing marginal products.
Second, we assume that all inputs are independent of one another. Both of these assumptions
are intuitively appealing and create models that generate clear predictions. The assumption
of independent inputs simplifies the model and increases the number of testable implica-
tions. Not only do we have no a priori theoretical grounds to assume which inputs are
substitutes or complements, but there is empirical justification for their independence. First,
were they not independent, contracts could adjust some input prices upward, others down-
ward, to influence farmer behavior. This, however, is not observed for the cases we study.
15
Second, in chapter 5 we show that input cost sharing in cropshare contracts exhibits an all-
or-nothing dichotomy; that is, input costs are either shared in the same proportion as output
or are not shared at all. This result is consistent with independent inputs.
Price taking is another common feature of our models. For example, we always assume
that the opportunity cost of the farmer's input is the competitive wage rate
h
. First we assume that
h
w
per unit of
farmer's effort, and the opportunity cost of the unpriced land inputs is
per unit. We also
assume that farmers sell their output on world markets and that they cannot influence this
output price. These assumptions seem reasonable in the context of modern agriculture with
world trade, where individual farmers are small relative to both the input and output markets.
The logic of our models is straightforward. Once the precise production structure is
constructed, we begin by deriving the first-best, zero transaction cost outcome for a specific
problem. By “first-best" we mean not only that inputs are used in the optimal amounts, but
also that inputs are fully specialized and applied at the appropriate time. We use this outcome
r
