Structure of Poverty

The entire population was characterized from poorest to richest. Applying the World Bank's US$1 and US$2 poverty headcounts, 33.8% (ADB 2007) and 77.7% (WDI 2007) of the Cambodian households are below the US$1/day and US$2/day poverty lines, respectively. As for Lao PDR, 28% of the population is below the US$1 poverty line (ADB 2003); while 74% is below the US$2 poverty line (WDI 2007). These poverty lines apply to all subsequent procedures in this paper.

The poverty income levels were taken back into the household strata to give an estimate of the headcount by stratum. Cross-tabulation of the variables strata and households obtained poverty rate per stratum, poverty headcount share in total poverty, and poverty headcount share in total population. Table 3 [ PDF 81.1KB | 1 page ] shows that in Cambodia, poverty is largely concentrated in the rural diversified followed by the agriculture sectors; in Lao PDR, the poorest are in agriculture, followed by the rural diversified sectors. The frequency tables for occupation and industry in the Cambodia and Lao PDR datasets (Table A1.5 [ PDF 113.4KB | 2 page ] and Table A2.3 [ PDF 104.1KB | 2 page ], respectively) substantiate these figures.

5.1 Earnings Share at the Poverty Line

Similarly, poverty income levels were taken back into the strata to estimate the average factor income shares in total household income, in the neighborhood of the poverty line. This was done by taking 10% of the population around the poverty line—5% below and 5% above the poverty line—for each stratum, and computing the average factor earnings shares in the total household income. To illustrate, 72% of the total income of the agriculture households around the US$1/day poverty line comes from unskilled agricultural labor, while for Lao PDR the number is closer to 58%. The trends in Table 4 [ PDF 189.1KB | 1 page ] show that households depend highly on (imputed) unskilled labor income.

5.2 Poverty Arc Elasticity

Poverty elasticity is usually computed by shocking income by one percent and calculating the change in poverty. However, if there is a gap between income levels at the poverty line and the households' income below such a level, then such an approach may understate actual poverty impacts. Taking the arc elasticity solves this problem by focusing on the changes in the neighborhood of the poverty line, and increasing the range over which impacts can be measured. Arc elasticity is obtained by computing the change in poverty headcount with respect to the change in the real income of households around the poverty line; this can be approximated by the following formula:

where the cumulative distribution function that computes poverty headcount ratio

when is the poverty income level in country c.

Lining up the stratum population from poorest to richest, poverty arc elasticities can be computed directly off the slope of the cumulative distribution function in the neighborhood of the poverty line (Hertel ,et.al, 2007a). For example, 48% of the population in the Agriculture stratum in Cambodia falls within 5% (above and below) of the US$1/day poverty level. A 32% change in income results in a 21% decrease in the poverty headcount, implying a poverty elasticity of -0.64 (Table 5 [ PDF 114.4KB | 1 page ]). The table also shows that, as expected, elasticities diminish from US$1/day to US$2/day as the density of poor households around the poverty line increases. This trend is more prevalent in Cambodia than in Lao PDR.

Download this Paper [ PDF 293.9KB| 26 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.