A Multidimensional Approach to Poverty

The study of poverty in Morocco has always been based on the so-called monetary approach advocated by the World Bank. Indeed, the Direction de la Statistique (DS) uses consumer spending alone to characterize this phenomenon. However, due to its one-dimensional character, this approach has proven to be limited. In recent years, a universal consensus has been reached: Poverty is to be treated as a multiple deprivation. This approach stipulates the importance of taking both monetary and nonmonetary dimensions into account when measuring and analyzing poverty.This research utilizes raw data from the 2004 General Population and Housing Census (Recensement général de la population et de l’habitat, or RGPH) to approach and measure multidimensional poverty in an urban setting in one of Morocco’s largest cities, Marrakech. This article’s precise goal is to construct a composite poverty indicator (CPI) for each household. This indicator allows us to map multidimensional (nonmonetary) poverty across the six urban districts of the city of Marrakech.Our methodology relies on the complementary use of several advanced statistical methods: Multiple correspondence analysis (MCA), two classification methods (ascending hierarchical classification (AHC) and a partition method) and stochastic dominance. Our results highlight the stark disparities between the different districts of the city in 2004 with regard to non-monetary poverty. Our detailed case study reveals a discrepancy of 18.4 percentage points between Guéliz, the district with the smallest poor population, and Annakhil, the district with the greatest number of poor inhabitants. These results also differ significantly, in terms of incidence, from those obtained by the Haut Commissariat au Plan (HCP) using a monetary approach and illustrated by their poverty map for the same year, 2004.