Figure 6.2 Poverty headcount rate, by season
6.4 Equality in consumption expenditure
Overall inequality in Afghanistan, represented by the national Gini index see Text box of 29, is low and also lower than
neighbouring countries in South and West Asia Figure 6.3.
There is, however, differentiation across regions in the sense that in several south-western regions inequality is very low. It
can be noted that the regions Southwest and West-Central link high equality with particularly widespread poverty compare
Figure 6.1.
Figure 6.3 Gini index, by residence and region, and for selected countries
a
Poverty incidence and poverty profiling
57
Gini Index
The Gini Index measures the extent to which the distribution of consumption among individuals or households within a
country deviates from a perfectly equal distribution. A value of 0 represents absolute equality with everybody consuming the
same amount, a value of 100 represents absolute inequality, where all consumption is concentrated in one person.
a Source for other countries: World Bank 2009.
Poverty incidence and poverty profiling
58
The distribution of per-capita real consumption expenditure shows that the richest quintile has a share in total consumption expenditure of 39 percent, more than four times higher than that of the poorest quintile
Table 6.2, panel a. However, the shares of the two quintiles below the richest do not much deviate from the 20 percent that would be implied
by equality. This indicates that inequality importantly originates in a relative wealthy top quintile and that the other successive quintiles only moderately differ from one to another. In fact the per capita expenditure of the middle quintile
is close to the poverty line, suggesting that a significant proportion of the non-poor can be considered vulnerable to falling under the poverty line.
Table 6.2 Mean per-capita monthly real consumption expenditure and share in consumption, by a population quintile, b poverty status
Poverty variable Population
share Consumption
Per-capita expenditure Share
Cumulative share
a. Poverty quintile
Poorest 20
758 9
9 Second poorest
20 1,100
13 22
Middle 20
1,416 17
39 Second richest
20 1,858
22 61
Richest 20
3,231 39
100
b. Poverty status
Poor 36
950 20
20 Non Poor
64 2,079
80 100
6.5 Profiling the poor
Poverty is closely linked to a wide variety of household and individual characteristics, such as household size, education, disability and employment. In addition, contextual characteristics – for instance, the labour market, accessibility and
geographic features – may bear an impact on the poverty status. However, the causal relation between poverty and such individual, household and context variables may not always be directly evident. Moreover, they are likely to interact
in intricate ways and more advanced multivariate analysis is required to disentangle this interplay and to find the true determinants. This is, however, beyond the scope of the present report.
6.5.1 Household size and composition
There is an apparent correlation between poverty and household size. Overall, the average size of poor households is 8.0 persons, compared to 6.9 for non-poor households.
Figure 6.4 shows that the proportion of poor households is progressively higher for households with more members up to household size nine.
5
A likely cause for this initial correlation is the increase in number of dependents, particularly children who do not, or only to a limited extent, contribute
to household income. In this perspective, the levelling off of the effect for households with more than nine members could be related to the possibility that these large households increasingly involve extended or composite households
with additional adult members engaged in income generating activities. This is also in line with the difference in the dependency ratio between the poor and non-poor; whereas the ratio of dependents children and elderly to the working-
age population for the latter is 106, for the poor population it is a very high 128. Accordingly, the proportion of poor households is positively and strongly correlated to the number of children, but negatively with the number of persons in
the working age 15-59 for households with three or more persons in this age group see
Figure 6.5.
_________________________________________________________
5
Caution in the interpretation of this correlation is warranted since the definition of household wellbeing, per capita consumption, is intrinsically linked to household size.
Figure 6.4 Percentage of poor households, by household size
Figure 6.5 Percentage of poor households, by number of children under 15 and adults aged 15-59
Even though reverse causality is well possible, it is likely that large household size due to high fertility and a large number of children negatively affects the per-capita consumption level in the household. Therefore, the higher fertility among
Kuchis resulting in a larger number of children per household see chapter 3 can be a contributing factor to their higher poverty incidence and poverty depth.
6.5.2 Characteristics of the head of household
The head of household is in most instances the single most determining person for the conditions of living in the household. This is especially the case in Afghanistan where households are almost exclusively male-headed 98 percent;
see Section 3.3.1 on household structure and men are the main or sole decision makers in almost all life domains
Poverty incidence and poverty profiling
59