Counting the Poor: The Liquidity-Adjusted Supplemental Expenditure Poverty Measure

Most countries use income as a measure of available resources when calculating household-level poverty status. We introduce a new measure of poverty status to the literature which is based on household expenditure instead. We show that an expenditure-based measure is the theoretically correct measure of resources actually transferred into a given period in the life cycle because it includes spending financed by borrowing and asset drawdowns, which income does not. We also show that the illiquidity of both service flows from durables and the expenditure needed to maintain them, as well as the value of in-kind transfers received, requires an adjustment to both the poverty threshold and available resources because those flows and transfers cannot be used to satisfy any minimum consumption need other than that for a single specific good. Our expenditure-based poverty measure constitutes an internally consistent comparison of liquidity-adjusted needs to the liquid resources available to meet those needs when calculating who is poor and who is not. When applying our new poverty measure to the Consumer Expenditure Survey data from 2009 to 2022, we find expenditure poverty to be below that using income but that the relationship between the two varies strongly with where the poverty threshold is drawn. We show that the impact of tax and transfer programs on poverty using our measure differs from that using the standard approach and could yield a slight reduction in the anti-poverty impact of those programs, that the impact of adding a transportation need to the minimum bundle required to not be poor could raise the poverty rate by about two percentage points, and that our expenditure resource-based measure is lower than that yielded by a consumption measure when using a poverty threshold similar to that used by the U.S. government.