Rasch Model and Multidimensional Poverty Measurement

The topic of the multidimensionality of poverty is currently at the heart of many theoretical, empirical and institutional debates in the European Union. Despite this increasing interest, there seems to be no consensus on how to define and measure multidimensional poverty. Two aspects may be considered in measuring poverty: the number of dimensions and the nature of the underlying continuum. The question of the dimensionality of poverty, one versus many dimensions, has to be resolved in applying specific multidimensional methods, like factor analysis, where the one-dimensional solution is a special case of the multidimensional procedure. The question of the nature of the continuum concerns the relationship between the items in each dimension. Two kinds of relationship are considered here: homogeneous and hierarchical. In this paper, the interest of the Rasch model for verifying the hierarchical and cumulative nature of the relationship between the items is underlined. After presenting the main characteristics of the model, and its adjustment for testing poverty, an application confirming the multidimensional nature of poverty is performed on a Luxemburgish dataset (PSELL-3).