Robust Poverty and Inequality Measurement in Egypt: Correcting for Spatial-price Variation and Sample Design Effects

The paper estimates inequality and absolute poverty in Egypt for 1997 with measures that are robust to sample design effects and corrected for spatial variation in price levels. Standard errors for inequality indices are calculated using a bootstrap approach which replicates the sample design. Standard errors for poverty indices are corrected for the design effects resulting from sample stratification and clustering. The authors use data from the Egypt Integrated Household Survey and follow the cost‐of‐basic‐needs methodology to construct region‐specific poverty lines. It is found that 15.7 million people were poor in Egypt in 1997, or 26.5% of the population. The estimates indicate a sharp sectoral difference with rural areas being significantly poorer, but significant differences in poverty were not found between Upper and Lower Egypt. This finding differs substantially from the conventional wisdom that Upper Egypt is poorer than Lower Egypt and results from the correction for spatial‐price variation.