Measuring Household Welfare

Consumption expenditure is probably the most common and preferred welfare indicator; however, its measurement is a challenging and time-consuming task. Although short consumption modules have potentially enormous advantage in terms of time and money savings, a recent and comprehensive literature on available experiments comparing short versus long modules is still lacking. The present paper aims at filling this gap trying to draw conclusions in terms of the accuracy of consumption and related poverty and inequality (distributional) estimates based on short modules. First, the paper briefly reviews the literature on how to accurately measure consumption and how survey design can influence consumption estimates; then, the empirical literature is discussed. The literature review mainly focuses on studies from the 1990s on developing countries. Available evidence seems to indicate that short modules underestimate consumption with respect to longer ones resulting in lower levels of recorded consumption and therefore less accurate estimates and higher poverty rates. However, one of the most complete, recent and authoritative studies in the field (Beegle et al., 2010) finds that short modules may actually result in a smaller downward bias compared to the benchmark than other longer consumption modules. In terms of relative ranking of households, the literature is scant; however, results from rigorous studies indicate that household consumption rankings obtained through short consumption modules are largely consistent with rankings derived from long modules. A critical review of the available evidence points to a number of factors that hinder the ability to draw firm conclusions; it indicates that there is still room for further investigation and provides some guidance for future field experiments in order for them to reach conclusiveness.