Incorporating uncertainty in poverty dynamics: how can we assess the economic impact of AIDS mortality in the presence of measurement error and missing data?
Alessandra Garbero, London School of Hygiene and Tropical Medicine (LSHTM)
Disentangling the relationships between HIV and AIDS and poverty has proven methodologically difficult despite many studies attempting to quantify their contributions to households’ vulnerability in sub-Saharan Africa. Is AIDS exacerbating poverty or is household-level poverty fuelling the HIV epidemic? The existing economic literature is concerned largely with two issues inherent in estimating the impact of AIDS on household welfare: the potential endogeneity of adult deaths and unobserved heterogeneity of households. Most of the existing demographic literature measures welfare uses asset indices. No evidence to date exists concerning the implications of using poor welfare proxies to estimate the impact of AIDS deaths on households' consumption or, more broadly, welfare. This paper addresses the problems of measurement error and missing expenditure data in the consumption modules of the Africa Centre Demographic Surveillance System from KwaZulu-Natal, South Africa and compares expenditure-based estimates of poverty dynamics associated with mortality with those from asset indices.
Presented in Session 155: Measuring poverty