data have no information on the use of any farm inputs. A study by Shaban (1987), however,
offers support for our implications for input choices. Using data from Indian villages,
Shaban found that input use was significantly lower (from 19 to 55%) on shared land
compared to owned land. He concludes that the data refute the idea that land owners are
able to stipulate and perfectly monitor input uses in a share contract. More recently, Canjels
(1996) finds heavier input use (for example, fertilizer, pesticides) in cash rent contracts
compared to cropshare contracts using data from the United States in the late 1980s.
Empirical Analysis: The Choice of Contract
To test the above predictions we use data from the same leasing surveys used in chapter 3.
Again, each observation is a single contract between a farmer and a landowner, and the
variables used in this chapter are listed in table 4.2. Though the appendix describes the data
in more detail, it is again useful to examine some features of the data used in this chapter.
CONTRACT is a dummy variable that identifies cropshare contracts, and ADJUSTMENT
is a dummy variable that identifies the presence of an adjustment clause in a cash contract.
These are used as dependent variables in the analysis that follows. For the Nebraska-South
Dakota data we have over 3,400 contracts, and for the British Columbia-Louisiana data
we have over 1,000 contracts. Information on ADJUSTMENT is only available for the
Great Plains data. For some estimates we divide the Nebraska-South Dakota data into
two subsamples: one for which farmers provided data (1,261 contracts) and one for which
landowners provided data (2,171 contracts). Some information is only available for farmers
or for landowners. 21 These anomalies are discussed as they arise.
Independent variables can be grouped into those that measure output division costs and
soil exploitation. HAY and INSTITUTION are dummy variables that indicate the presence
of hay and institutional landowners, respectively, are used as output division variables.
IRRIGATED, RICE, ROW CROP, and TREES are dummy variables that indicate the
presence of irrigation systems, rice, row crops, and tree crops, respectively, and are also
used as soil exploitation variables. DENSITY is the population density for the county in
which the plot lies and is also a soil exploitation variable. 22 A number of variables are used
as controls, and depending on the data set these include the age of the farmer (AGE), the size
of the leased plot (ACRES), a dummy for leases between family members (FAMILY), the
total farm sales (FARM SALES), and the fraction of income derived from the farm (FARM
INCOME). ROW*HAY is an interaction variable also used as a control. 23 ABSENT is a
dummy identifying absentee landowners and ACRES OWNED measures percent of farm
acres owned by the farmer. These last two variables are used to test hypotheses from other
sources than our model.