Household size and composition

Poverty incidence and poverty profiling 60 see Section 10.2.1 on household decision making. Therefore, it can be expected that the attributes of the head of household have considerable predictive power for the poverty status of the household. From Figure 6.6 it can be deduced that in the rare cases that households are female-headed, on average poverty is more widespread than in male-headed households. Underlying factors could be that women have less access to employment see Section 4.2 on labour force and employment or that widows lack the support of a partner for income generating. Figure 6.6 Percentage of poor households, by selected attributes of household head Literacy and educational attainment apparently are key factors in the explanation of poverty, and therefore should be incorporated in poverty reduction strategies. The ability to read and write and knowledge learned in the education system are strong facilitators for adequate performance on the labour market and in social life. Households of illiterate heads are 31 percent more likely to be poor than those of literate heads, and the household poverty rate decreases steadily with higher levels of education: the likelihood of being poor for households with uneducated heads is 1.5 to 5 times higher than those with heads that have any educational attainment above middle school. However, there is not very much differentiation between households of heads with middle school and primary school. The age of the household head does not show any clear effect on poverty data not shown. Poverty incidence and poverty profiling 61 Households headed by a disabled person are also somewhat more likely to be poor, but the effect is small. The correlation is likely to work again through labour force participation and employment, as the likelihood that disabled persons participate on the labour market is only half of that of non-disabled persons see Section 8.5.3. With regard to labour characteristics of the head of household, Table 6.3 indicates that those who are unemployed are significantly 14 percent more likely to be poor than households with employed heads. Somewhat surprisingly, households with inactive heads report the lowest incidence of poverty. A possible explanation could be that these concern elderly heads who reside with younger generations who can provide for sufficient household income. The table also clearly shows large variation across industries in which the household head is working. If households have heads working in sectors that require extended levels of education – such as communication, public administration, health and education – the likelihood of securing adequate provisions is significantly higher than when heads work in the large agricultural sector and in construction, manufacturing, and mining and quarrying. Table 6.3 Percentage poor households, by a activity status of household head and b Industry of working household head a. Activity status Share poor b. Industry Share poor Employed 32.9 Transportcommunication 22.3 Unemployed 37.5 Public administration 24.0 Inactive 29.5 Health 25.5 Total 32.8 Trade 28.5 Education 29.9 Other services 33.0 Agriculture and livestock 35.5 Construction 35.9 Manufacturing 38.4 Mining and quarrying 49.1

6.5.3 Characteristics of household members

Apart from the attributes of the head of household, specific characteristics of other household members may also be linked to poverty. There are some 266 thousand households in Afghanistan in which at least 1.2 million children perform child labour. 6 Usually, child labour is performed in view of pressing needs to supplement household income. It is, therefore, not surprising to find that households with at least one child engaged in child labour are significantly more often poor than those without child labour Figure 6.7. The effect of having one or more disabled household members is visible in the graph, but very small. The relation between poverty and migration is a complex one, and one of which the direction of causality is not immediately evident. No effect is found for households with or without members who left to live somewhere else, mostly for work- related reasons. However, a substantial difference in the proportion poor is shown between households that have one or more seasonal migrants 38 percent and households with no seasonal migrants 32 percent. This effect can readily be interpreted in the sense that seasonal migration provides a coping strategy to many vulnerable and poor households. Additional analysis is required to further explore the interaction between poverty and migration. _________________________________________________________ 6 For a definition of and information on child labour, see Section 4.4 of this report.