Examination of the effects of variation in the measurement of multidimensional poverty in the United States

Abstract

This chapter examines the extent to which methodological differences produce variations in multidimensional poverty index (MPI) values and corresponding measures of deprivation incidence and intensity in the United States. Application of the methodologies employed in 10 prior studies to a common data set and period (i.e., annual ACS-PUMS data for the years 2008 through 2019) reveals considerable heterogeneity in MPI values and the incidence of multidimensional poverty (i.e., the headcount ratio). Across methodologies, lesser variation exists in the average intensity of multidimensional deprivation. The primary factors driving variation in MPI values are differences in the numbers of indicators and dimensions employed, the choice of indicators, the grouping of indicators into dimensions, and the resulting differences in indicator and dimension weights. Given these findings, and as multidimensional poverty analysis has important implications for public policy, a set of best practices is recommended to produce more uniform measurement of multidimensional poverty in the United States.