This study has analysed two different Household Budget Survey (HBS) data to find out how the composition of poor households varies with the chosen method and index of distinguishing the households between poor and the non- poor. The objective of the study was to obtain evidence as to whether factors, other than large household size have negative or positive effect on household welfare. In other words whether poverty is explained largely by the small size of the household or whether demographic factors like age, household composition, employment status and other factors as well are significant. Also, to find out, as to whether the best method of identifying the poor households is to use total consumption or consumption per capita.
The researchers found out that small sized households make up a large segment of the poor when the method of identifying the poor uses total household consumption as a measure of welfare with no focus on the household composition. However, the poor are substantially made up of members from larger households when the measure of welfare is on the per capita basis and there is focus on the household composition. Household composition and use of adult-equivalence scales to get per capita measures as welfare measures have an important effect upon the welfare rankings of households. When welfare is measured at per capita basis and households are ranked accordingly, larger households are more prone to poverty than smaller households. It was also found out that with this measure of welfare, average household size of identified poor households is larger than the national average household size, while the average for non-poor ones is smaller than the national average.