Reconstructing the Human Development Index at the Individual Level from Survey Microdata: Development and Validation of a Methodology

arXiv:2610.35893v1 Announce Type: cross Abstract: This study proposes and validates a methodology for reconstructing the Human Development Index (HDI) from individual-level survey microdata, enabling disaggregated estimates for demographic subpopulations while maintaining coherence with the official HDI. Data from 40,185 respondents in 21 countries from the European Social Survey (ESS7, 2014) were analysed, combined with World Bank indicators for life expectancy, gross national income per capita, and expected years of schooling. The health dimension was estimated from national life expectancy, adjusted by self-rated health; income, by a lognormal multiplier calibrated by the Gini coefficient; and education, by combining observed and expected years of schooling. The dimensions were aggregated by geometric mean, weighted by sampling weights, with confidence intervals estimated via the delta method. Among the 39,706 individuals with valid data (98.8%), the reconstructed HDI showed high agreement with the official 2014 HDI, with a mean absolute error of 0.0099 and a Pearson correlation of 0.917. Consistent differences across subgroups evidenced the methodology's convergent validity and its capacity to produce disaggregated estimates of human development, expanding the HDI's potential for analyses of socioeconomic and demographic inequalities.

Source: arXiv cs.LG | 2026-10-28