(1) | (2) | (3) | |
---|---|---|---|
Farming ability (estimated from model results in Table 3) | 0.248** (2.43) | 0.180* (1.70) | |
Log of household income (lagged) | 0.090* (1.65) | ||
Log of adult equivalents | 0.378*** (4.33) | 0.379*** (4.37) | 0.345*** (3.84) |
Log of acres owned | -0.450*** (7.13) | - 0.373*** (5.25) | - 0.585*** (6.24) |
(Log of acres owned)2 | 0.064* (1.90) | ||
Log of rental rate | 0.087 (0.67) | 0.099 (0.77) | 0.083 (0.64) |
Log of value of agricultural assets | 0.217*** (4.80) | 0.204*** (4.50) | 0.185*** (3.91) |
Log of household head’s age | 12.814*** (4.01) | 12.855*** (4.04) | 12.646*** (3.94) |
Log of household head’s age squared | -1.631*** (4.04) | - 1.637*** (4.07) | - 1.608*** (3.96) |
Household head completed primary school ( = 1) | -0.135 (1.28) | -0.144 (1.38) | -0.139 (1.31) |
Female-headed household ( = 1) | - 0.280*** (2.68) | -0.267** (2.51) | - 0.252** (2.45) |
Distance to extension services (km) | 0.008 (0.71) | 0.006 (0.49) | 0.009 (0.77) |
Death of household head since prior survey ( = 1) | 0.108 (0.49) | 0.103 (0.47) | 0.097 (0.44) |
Distance to main road (km) | 0.022 (0.54) | 0.022 (0.56) | 0.025 (0.62) |
Population density (persons per km2) | 0.306 (0.99) | 0.249 (0.84) | 0.391* (1.94) |
Log of rainfall during main growing season (mm) | -3.647* (1.69) | -3.224 (1.53) | -3.671* (1.72) |
Log of rainfall during last main growing season (mm) | 0.309 (0.33) | 0.205 (0.23) | 0.225 (0.23) |
Distance to nearest piped water source | 0.000 (0.03) | - 0.000 (0.04) | 0.000 (0.04) |
Distance to the nearest electricity supply | 0.006 (0.31) | 0.002 (0.09) | 0.008 (0.40) |
Credit constrained ( = 1) | -0.130 (0.49) | -0.109 (0.42) | -0.188 (0.69) |
Observations | 1,155 | 1,156 | 1,154 |
Note: Robust z-statistics in parentheses. All the standard errors are adjusted for clustering effect at the village level. Agroecological zone and village dummies are included in all the regressions. Other controls include village soil types and topology features. Inverse probability weighting is used to account for potential attrition bias (see methods section for description).
* Significant at 10%;
** Significant at 5%;
*** Significant at 1%.