Abstract

National estimates often conceal substantial territorial inequalities, limiting the ability of the Sustainable Development Goals (SDGs) monitoring framework to identify populations and communities left behind. As discussions on future development monitoring frameworks gain relevance, incorporating subnational statistics becomes essential to better capture territorial disparities. Although household surveys remain a primary source of social statistics, their sample sizes frequently prevent reliable estimation at lower administrative levels or for specific population groups. Small Area Estimation (SAE) methods offer an alternative by combining survey data with population censuses, satellite imagery and other auxiliary sources to produce geographically disaggregated statistics.
This presentation highlights recent developments led by the United Nations Economic Commission for Latin America and the Caribbean (ECLAC) in supporting national statistical systems to produce disaggregated social indicators using SAE methodologies. Applications across several countries cover monetary and multidimensional poverty, labour market statistics, and a growing set of indicators related to gender equality and reproductive health. These applications draw on different model-based approaches, adapted to the characteristics of each indicator and the available data sources. ECLAC also maintains a regional SAE portal that is regularly updated with subnational estimates, contributing to the dissemination and accessibility of geographically disaggregated statistics.
Drawing on these experiences, the presentation discusses how SAE can contribute to making subnational inference a more systematic component of future development monitoring frameworks for the SDGs. Committing SAE within national statistical systems could help move beyond national estimates, providing a more comprehensive picture of development inequalities and strengthening the statistical foundations of future global and regional monitoring agendas.