Agriculture Reference
In-Depth Information
Thus, cropshare contracts are more likely to be chosen for row crops where tillage is
important.
27
To summarize, we expect negative coefficients for DENSITY and IRRIGATED
and a positive coefficient for ROW CROP.
Cash Rent versus Cropshare Estimates
We use our contract data to estimate the determinants of cropshare contracts (versus cash
rent contracts) and test the predictions from our model. We use the same general empirical
specification from chapter 3, so that for any farmland contract
i
the complete model is
C
i
=
X
i
β
i
+
i
i
=
1,
...
,
n
;
and
(4.5)
1,
C
i
>
if
0
C
i
=
(4.6)
C
i
≤
0,
if
0,
C
i
where
is an unobserved farmland contract response variable;
C
i
is the observed dichoto-
i
mous choice of land contract for plot
, which is equal to 1 for cropshare contracts and equal
X
i
to 0 for cash rent contracts;
is a row vector of exogenous variables including the constant;
β
i
is a plot-specific error term. We use
a logit model to generate maximum likelihood estimates of the model given by equations
(4.5) and (4.6) for various contract samples.
Table 4.3 shows the logit coefficients estimates of the influence of selected variables on
the choice of contract for the Nebraska-South Dakota sample. The first equation includes
all contracts (3,432). The coefficient estimates for the output division variables—HAY and
INSTITUTION—are both negative and statistically significant as predicted. The coefficient
estimates for the soil exploitation variables—DENSITY, IRRIGATED, and ROW CROP—
also have the expected signs and are statistically significant. This implies that when soil
exploitation is of less concern, a cropshare contract is less likely.
28
In the second and third
equations in table 4.3, all of the coefficient estimates for these variables still fulfill the
predictions, although in a few cases the t-statistics fall. These estimates offer support for
our theory of contract choice and are also consistent with the observation of Gray et al.
(1924, 589) that “especially in the Corn Belt it is frequently customary to cash rent the hay
land while sharing the grain crop.” Dubois (2002) finds that in the Philippines, corn—a crop
that is hard on the land—is more likely to be cropshared than are the main alternatives of
rice and sugar.
While we do not have data on soil quality directly, two recent empirical studies support
our model. Using data from a special U.S. Census survey, Canjels (1996) finds that erodible
land is more often cropshared. Similarly, Sotomayer, Ellinger, and Barry (2000) use a set
is a column vector of unknown coefficients; and
i
