Determinants of multidimensional poverty: the case of the Union of the Comoros

Abstract

Analyzing poverty is a tedious exercise. Today, studies have shown that poverty is a multidimensional phenomenon. It is therefore necessary to take into account several dimensions (monetary and non-monetary) in order to differentiate the poor from the non-poor. The objective of this research is to identify the factors that significantly explain the multidimensional deprivation index. To achieve this, we used data from the Harmonized Survey on Household Living Conditions (EHCVM) in Comoros in 2020. From this base, we selected the various dimensions likely to explain the level of household well-being in order to construct the multidimensional deprivation index using the fuzzy set method. This method enables us to model vague phenomena such as poverty. The index thus created is considered as the variable to be explained in quantile regression. The specificity of this method is that it provides a complete picture of the effects of the different explanatory variables on the values of the variable to be explained, i.e. the multidimensional poverty index. In this sense, quantile regression has enabled us to identify the factors that significantly explain multidimensional poverty in the Comoros. The results show that households living on the islands of Ndzouani or Ngazidja are less affected by multidimensional poverty than those living on the island of Mwali. In addition, households living in urban areas are also less affected than those living in rural areas.  On the other hand, households headed by women are more affected by multidimensional poverty than those headed by men. It should also be noted that households headed by a head with no professional activity are more affected by multidimensional poverty in the Comoros.

Classification JEL: C31, D63

Paper type: Research paper