Poverty indices are usually calculated on the basis of (cross section) income data from the past. As past income is fixed and future income is uncertain such measurement does not reflect the ex ante risk of falling into poverty. This paper presents refinements of commonly used poverty measures to account for income risk. It is shown that the standard headcount ratio underestimates poverty in societies with moderate poverty if income is risky. The Foster, Greer and Thorbecke measure always underestimates poverty if income risk is neglected.
|Journal||Jahrbücher für Nationalökonomie und Statistik|
|Publication status||Published - 2004|