Agriculture Reference
In-Depth Information
a
Table A1.1.
Ordered probit estimates of individual participation in non-farm employment
(standard error in parentheses)
Marginal effects
b
Coefficient
No non-farm
employment
Some
non-farm
employment
Only
non-farm
employment
(1)
(2)
(3)
(4)
Zone
c
0.369
*
-0.128
***
0.045
***
0.083
***
Locally linked
(0.191)
(0.017)
(0.015)
(0.031)
0.812
***
-0.292
***
0.086
***
0.206
***
Leakage to urban areas
(0.176)
(0.012)
(0.022)
(0.027)
0.720
***
-0.259
***
0.079
***
0.180
***
New urban links
(0.179)
(0.013)
(0.020)
(0.028)
Household characteristics
0.027
*
-0.009
*
0.003
*
Number of members
0.005
(0.016)
(0.005)
(0.002)
(0.004)
-0.053
*
0.017
*
-0.007
*
-0.010
*
Farm size (ha)
(0.028
(0.009)
(0.004)
(0.006)
Ratio land/members
d
0.130
-0.042
0.017
0.026
(0.142)
(0.046)
(0.018)
(0.029)
Individual characteristics
-0.006
*
0.002
*
-0.001
*
-0.001
*
Age
(0.004)
(0.001)
(0.0004)
(0.001)
0.446
***
-0.144
***
0.056
***
0.088
***
Gender (F = 0; M =1)
(0.125)
(0.013)
(0.016)
(0.025)
-0.345
*
0.104
***
-0.044
***
-0.060
*
Child (under 15)
(0.181)
(0.029)
(0.003)
(0.036)
Education: primary
e
-0.040
*
0.120
0.015
0.024
(0.079)
(0.020)
(0.011)
(0.032)
Education: secondary
e
0.459
***
-0.160
***
0.055
***
0.106
***
(0.110)
(0.016)
(0.016)
(0.030)
Education: post-secondary
e
0.626
***
-0.232
***
0.065
***
0.166
***
(0.208)
(0.017)
(0.019)
(0.032)
-1.166
***
Constant
n.a.
n.a.
n.a.
(0.218)
*
Significant at 10% level;
**
Significant at 5% level;
***
Significant at 1% level.
a
An ordered probit model was fitted using maximum likelihood estimation. Number of
observations = 1797. Log likelihood function = -1281. Covariance matrix was adjusted for
clustering according to 30 villages.
b
Marginal effects are evaluated at means of regressors.
c
The 'traditional' zone is used as the benchmark.
d
Ratio of land to household members is calculated as total hectares divided by total number
of household members using an adjustment factor for children.
e
For education level, 'no (formal) education' is used as the benchmark.
Search WWH ::
Custom Search