A theory-based living standards index for measuring poverty in developing countries

Living standards indices based on asset ownership are increasingly being used to describe poverty in developing countries. Yet, less work has been done on constructing meaningful weights using economic theory. In this paper, I develop a novel, cardinal index by incorporating asset prices, which can be feasibly collected at low cost. I weight items based on marginal utilities, estimated using a structural model of household preferences. Using data from the Living Standards Measurement Surveys from Malawi (2004, 2010, 2013), Nicaragua (1998, 2001, 2005), and Bulgaria (1995, 1997, 2001), I find that the marginal utility weights are positively but not linearly increasing in price for most items. I compare weighting methods and find that weights, household rankings, and poverty profiles are sensitive to construction method, suggesting that a theory-based method offers some advantages over commonly used statistical methods. Finally, I discuss methodological issues of spatial aggregation, temporal variation, and quality differentiation.