34
The land area is an important factor in determining household welfare. We include two
categories of land: farmland and land for other agriculture.
The physical capacity is difficult to measure. Following the literature, the physical capital
includes: cattle, buffaloes, and pigs.
Human capital is one of the most crucial factors to determining welfare. We used the
education level of the household’s adult members as a proxy for human capital.
Technology is an important factor in determining household welfare. We included two
variables as proxies for technology: chemical fertilizers and whether the household runs a
business.
Household Characteristics
Household Characteristics are also important factors to determining household welfare. For
household characteristics, we include ages, dependency ratio
25
, and gender of household
head.
Environmental factors
Environmental factors refer to the impact of infrastructure on household income earning
capacity. We include six variables to capture the impact of infrastructure: village access to
electricity, roads, health services
26
, safe water, markets
27
.
The results of the Logit model are shown in Table 2
28
. Out of 23 explanatory variables, 13
were found to be statistically significant in determining absolute gains or losses of household
welfare (Gi), which includes the factors of production, household characteristics, and
environmental factors.
Factors of production such as inputs of land are important factors to determining household
welfare. If the extent of a household’s ‘other agriculture land area’ has a positive and
significant impact on welfare, it means that farmer will gain benefits from AFTA, particularly
in lowland area. Should irrigation access have a positive, statistically significant, this implies
that access to irrigation could result in more productivity and thus contribute to household
welfare. Following the approaches described in other literature, physical capital includes
cattle, buffalo, and pigs, thus measuring the contribution of ownership of farm animals to
household welfare. Human capital is measured by education level of the household’s adult
members. All variables for human capital are strongly significant, and it appears that the
literacy of members in household and education level of household heads are positively
related to household welfare. This is consistent with government policies to contribute to
investments in human capital for building labor capacity of households across the country,
especially for the poor to escape from poverty. The variables related to technology- chemical
25
Dependency ratio was calculated by dividing the number of dependents by the total number of household
members.
26
Two variables were used to capture village access to health services: whether village has community health
work and the distance of the village to the hospital.We used two variables to capture village access to health
services: whether village has community health work and the distance of village to hospital.
27
To calculate the access to market, we used whether the village has a market located in the village.Access to
market, we used variable whether village has market at village.
28
See Summary Statistics of Variables in appendix 3 and correlation between variables in appendix 4
35
fertilizers and whether the household runs a business- appear to have different effects on
welfare. The use chemical fertilizers has a positive and significant effect but running a
business has negative and significant effects, meaning that households which run businesses
would lose welfare from AFTA. It is possible that they might lose welfare for only the short
term, but in the long term they would gain more benefits from AFTA.
Table 2: Results of factors influence on absolute gains or losses of household welfare (Gi)
Dependent variable : Gi
Coefficient
z-value
P>│z│
Explanatory variables
Factors of production
Irrigation Access to village
0.1448 *
1.80
0.0720
Number of Cattle per household
0.0226 ***
4.18
0.0000
Number of Buffalo per household
0.0619 ***
6.44
0.0000
Number of Pigs per household
0.0278 ***
4.00
0.0000
Total number of Literate per household
0.0476 **
2.21
0.0270
Literate Females
0.0451
1.53
0.1270
Education level of household head in
years
0.0626 ***
9.10
0.0000
Adults
0.0332
1.54
0.1230
Rice farm land Area per household
-0.0121
-0.71
0.4750
Other Cultivation Land Area per
household
0.0824 ***
4.24
0.0000
Village
located at Lowland Area
0.3063 ***
5.61
0.0000
Chemical fertilizers
0.2088 **
2.63
0.0090
Business
-0.1798 ***
-3.08
0.0020
Household characteristics
Dependency Ratio
-0.1270
-1.24
0.2140
Male Head of Household
0.0722
0.69
0.4880
Age of Household Head
-0.0053
-0.44
0.6610
Age Squares of Household Head
0.0001
0.58
0.5650
Environmental factors
Access to All Season Road
0.0543
0.87
0.3870
Electricity Access
-0.0639
-1.05
0.2920
Access to Safe Water
-0.0048
-0.09
0.9240
Community Health
0.2097 ***
4.38
0.0000
Distance from Village to Hospital
0.0005
0.36
0.7200
Access to Market
0.1833 **
2.07
0.0380
Constant
-1.4223 ***
-4.63
0.0000
Number of observations
8293
LR chi2(8)
433
Prob > chi2
0.0000
Pseudo R2
0.0383
Note: the superscripts *, ** and *** denote rejection at 10, 5 and 1 per cent critical values.
Source: Authors' calculations based on LECS 4