5 Company–Community Partnership and Climate Change Adaptation Practices …
89
Table 5.2 (continued)
Variables
Partnership
Non-partnership
p
Mean
SD
Mean
SD
Log of total farm income
16.466
1.137
16.017
1.144
0.000
Adopting grafting (=1 if
yes)
0.945
0.228
0.872
0.334
0.001
Adopting conservation
techniques (=1 if yes)
0.873
0.334
0.825
0.380
0.079
Planting spice crops (=1 if
yes)
0.708
0.455
0.611
0.488
0.006
Planting hardwood (=1 if
yes)
0.190
0.393
0.128
0.334
0.022
Observations
363
360
Estimation Strategy
We conduct estimation in two phases. The first stage of regression estimates the
determinants of joining the company partnership, and the second estimates whether
company partnership substantially drives the adoption of sustainable practices, i.e.,
the soil and water conservation and grafting methods, and plant diversification, and
improves farm income.
For the first step of regression, we evaluate the marginal probability effects (MPE)
of explanatory variables to be in a partnership with companies based on probit and
logit models. Through the result of this analysis, we intend to define what factors
determine the partnership with companies. We estimate
∂ P
Dummy Partnership i = 1
∂ x it, j
= ∅
x
it β
β j MPE in Probit
(5.1)
∂ P
Dummy Partnership i = 1
∂ x it, j
=
x
it β
1 −
x
it β
β j MPE in Logit
where t implies the year of 2012, 2013, and 2014; Dummy Partnership = 1 denotes
membership in company partnership and Dummy Partnership = 0 denotes the nonmembership; x it, j represents household characteristics relevant to the membership
in the partnership in each year of 2012, 2013, and 2014; β is the vector of coefficients
to be estimated; and ∅
x
it β
is the value of the standard normal probability density
function at x
it β.
In the second step, we examine the effects of having these partnerships with MNCs
on farmers’ income and climate change adaptation status. However, individuals may
self-select into treatment based on their anticipated monetary and nonmonetary benefits, resulting in possible reverse causality (endogeneity) between the partnership
89
Table 5.2 (continued)
Variables
Partnership
Non-partnership
p
Mean
SD
Mean
SD
Log of total farm income
16.466
1.137
16.017
1.144
0.000
Adopting grafting (=1 if
yes)
0.945
0.228
0.872
0.334
0.001
Adopting conservation
techniques (=1 if yes)
0.873
0.334
0.825
0.380
0.079
Planting spice crops (=1 if
yes)
0.708
0.455
0.611
0.488
0.006
Planting hardwood (=1 if
yes)
0.190
0.393
0.128
0.334
0.022
Observations
363
360
Estimation Strategy
We conduct estimation in two phases. The first stage of regression estimates the
determinants of joining the company partnership, and the second estimates whether
company partnership substantially drives the adoption of sustainable practices, i.e.,
the soil and water conservation and grafting methods, and plant diversification, and
improves farm income.
For the first step of regression, we evaluate the marginal probability effects (MPE)
of explanatory variables to be in a partnership with companies based on probit and
logit models. Through the result of this analysis, we intend to define what factors
determine the partnership with companies. We estimate
∂ P
Dummy Partnership i = 1
∂ x it, j
= ∅
x
it β
β j MPE in Probit
(5.1)
∂ P
Dummy Partnership i = 1
∂ x it, j
=
x
it β
1 −
x
it β
β j MPE in Logit
where t implies the year of 2012, 2013, and 2014; Dummy Partnership = 1 denotes
membership in company partnership and Dummy Partnership = 0 denotes the nonmembership; x it, j represents household characteristics relevant to the membership
in the partnership in each year of 2012, 2013, and 2014; β is the vector of coefficients
to be estimated; and ∅
x
it β
is the value of the standard normal probability density
function at x
it β.
In the second step, we examine the effects of having these partnerships with MNCs
on farmers’ income and climate change adaptation status. However, individuals may
self-select into treatment based on their anticipated monetary and nonmonetary benefits, resulting in possible reverse causality (endogeneity) between the partnership