ECLAC experiences on disaggregating SDG indicators: from poverty to unemployment

Poverty rate and labor market statistics are of major importance when addressing social and economic challenges. These statistics are typically derived from complex survey data. However, when it comes to small geographical areas or domains, survey samples often fail to provide reliable estimates. To address this issue, the Economic Commission for Latin America and the Caribbean (ECLAC) has put forth a standardized approach that utilizes area-level models, taking into account the binomial structure of poverty and the multinomial natura of occupation statuses. Within a Bayesian framework, we define a both a linkage and a sampling model that enable the simultaneous estimation of a diverse set of indicators of interest. This innovative approach proposed by ECLAC aims to overcome the limitations associated with traditional survey sampling methods, particularly in smaller areas where the precision of estimates may be compromised. By adopting this standardized approach, ECLAC aims to enhance the accuracy and reliability of poverty and labor market statistics at both national and subnational levels. The utilization of area-level models and the incorporation of Bayesian techniques allow for more reliable estimations, thus enabling policymakers, researchers, and practitioners to make informed decisions and develop effective strategies to address labor market challenges in Latin America and the Caribbean region.