Multidimensional poverty measurement without missing poor bias

Lifespan is strongly shaped by standard of living. When premature mortality disproportionately affects the poor, conventional poverty measurements are likely to be underestimated, since they only account for the living, among whom poor groups are systematically underrepresented. The missing poor refers to individuals who died prematurely and who, had they survived, would have been counted among the poor. This research revisits the missing poor framework within the context of multidimensional poverty measurement, a particularly important extension given that mortality inequality has gradients beyond income alone. The objective of this study is to identify and assess the missing poor bias and its relationship with poverty as measured multidimensionally. It is shown that the missing poor bias is driven by differences in the survival probabilities of population subgroups, as well as by the weights assigned to each dimension in the poverty index. Empirical evidence from Cambodia, El Salvador and Zambia illustrates these theoretical findings, demonstrating that mortality inequality plays a crucial role in multidimensional poverty measurement, and that the missing poor bias is substantially larger in countries where mortality rates and their inequality are higher.