Reconstructing Subnational Labor Indicators in Colombia: An Integrated Machine and Deep Learning Approach

Abstract

This study proposes a unified multi-stage framework to reconstruct consistent monthly and annual labor indicators for all 33 Colombian departments from 1993 to 2025. The approach integrates temporal disaggregation, time-series splicing and interpolation, statistical learning, and institutional covariates to estimate seven key variables: employment, unemployment, labor force participation (PEA), inactivity, working-age population (PET), total population, and informality rate, including in regions without direct survey coverage. The framework enforces labor accounting identities, scales results to demographic projections, and aligns all estimates with national benchmarks to ensure internal coherence. Validation against official departmental GEIH aggregates and city-level informality data for the 23 metropolitan areas yields in-sample Mean Absolute Percentage Errors (MAPEs) below 2.3% across indicators, confirming strong predictive performance. To our knowledge, this is the first dataset to provide spatially exhaustive and temporally consistent monthly labor measures for Colombia. By incorporating both quantitative and qualitative dimensions of employment, the panel enhances the empirical foundation for analysing long-term labor market dynamics, identifying regional disparities, and designing targeted policy interventions.