Volume 21, Number 1 Article by Sambit Kumar Mishra March, 2009
The Effect of Price Index on Food Poverty: A study of Urban Poverty in Chhattishgarh :
Mistakes in poverty estimation are often triggered when researchers follow less careful procedures in measuring prices and do not make adjustment for spatial price variation. In the poverty literature, little attention has been devoted to the effects of the choice of price indices on poverty measurement. This paper examines the sensitivity of food poverty measures to spatial price deflators in the seven main urban centres in Chhattishgarh, using the Chhattishgarh Urban Household Survey and adopting the food energy intake (FEI) approach to compute the poverty lines. It compares regional price indices calculated from official average prices for the main urban areas with indices derived from household data based on unit values derived from the household survey itself, which correct for quality effects, units of measurement and selectivity bias.
As measured by the headcount index, household food poverty rates in Chattishgarh vary from 23% (Ambikapur) to 64% (Jagadalpur) across the main urban areas according to the author's preferred spatial price deflators while it varies from 16% (Ambikapur) to 53% (Jagadalpur) according to the deflator obtained using external price information. Thus, the results caution against the use less careful procedures in estimating prices and price indices. They show that using official average prices rather than household survey-based unit values substantially biases poverty downwards. The differences between the poverty rates indicated by the external price measure and preferred measure are in the order of 7% (for the least food poor city) and 11% (for the most food poor city). The paper also examines the determinants of household welfare and food poverty using regression techniques. The models identify variables such as household composition, location, labour market status, asset ownership and level of schooling as important determinants.
Reprint No 09103
