Abstract

Multidimensional poverty measures have become essential tools for monitoring living conditions beyond income. Their usefulness for targeting policies to those furthest behind depends on the ability to produce reliable estimates for small population groups and territories. This presentation describes the experience of the Statistics Division of the Economic Commission for Latin America and the Caribbean (ECLAC) in applying small area estimation (SAE) methods to disaggregate the Multidimensional Poverty Index for Latin America (MPI-LA).
The MPI-LA, calculated for 16 countries in the region using household survey data, combines 12 deprivation indicators to provide a harmonised and regionally comparable measure of multidimensional poverty. When the index is disaggregated simultaneously by population subgroups (such as sex, age, ethnicity, or area of residence) and by major administrative divisions, sample sizes become too small, producing estimates with high coefficients of variation.
To overcome these limits, ECLAC has applied model-based SAE techniques that combine survey data with auxiliary information from population censuses and satellite imagery. The presentation discusses the main methodological challenges of this approach, particularly the need to estimate separate models for each of the 12 MPI-LA indicators while preserving the joint distribution of deprivations at the household level, a requirement given that multidimensional poverty is identified through the accumulation of simultaneous deprivations. It also addresses issues related to model specification and selection, the consistency of auxiliary variables between surveys and censuses, benchmarking to direct national estimates, and the assessment of precision through mean squared error estimation.
Finally, the presentation shows the results obtained, highlighting territorial and population-group disparities in multidimensional poverty, and reflects on the lessons learned for producing granular and comparable poverty statistics that support the commitment to leave no one behind.