In a game-theoretic framework, we study how scalar multidimensional poverty measures affect
the strategic interactions of ministers responsible for reducing deprivations in the measure’s dimensions.
Ministers share a common interest in reducing measured multidimensional poverty, but
also have preferences over alternative uses of their allocated budgets; we think of improvements
in the scalar multidimensional poverty measure as a public good for ministers, who therefore
can free ride on each other’s antipoverty spending. The allocation of resources across ministers
and the measure’s parameters (the weights assigned to each dimension and the extent of deprivation
depth aversion) affect equilibrium size and composition of antipoverty spending. For
common parameterizations, a reallocation of budgets that improves measured poverty reduction
always decreases total antipoverty spending in equilibrium. Similarly, choosing weights such
that the poverty measure is effectively unidimensional favors measured poverty reduction, while
antipoverty spending is larger with a multidimensional measure. Increasing deprivation depth
aversion may increase or decrease the resources actually spent on the poor, depending on whether
disparities across dimensions are due mostly to the number of deprived households, or to their
average deprivation. We illustrate using data from Mexico, the first country to adopt an official
multidimensional poverty measure.