Database: Measuring Social Well-being in Mexico, a Municipalized Approach Using the Human Development Index and Multidimensional Poverty Indicators (2010-2020)
Description
This dataset provides a municipal-level, year-comparable measure of social well-being in Mexico for 2010, 2015, and 2020, expressed as three normalized index scores (Efecto_2010, Efecto_2015, Efecto_2020) on a 0–1 scale, where higher values indicate higher social well-being. The dataset includes 1,877 municipalities with complete information across the three benchmark years and identifies each unit by a harmonized municipal code (cve_propia) and administrative names (nom_ent, nom_mun). The underlying hypothesis is that a transparent integration of municipal Human Development achievements and multidimensional poverty-related deprivations can yield a consistent, interpretable composite measure of social well-being that is comparable across municipalities and across time. Rather than advancing causal claims, the dataset supports descriptive and comparative analysis of how municipal well-being levels are distributed and how they shift between 2010, 2015, and 2020. The index is derived from publicly reported municipal indicators aligned to the national geostatistical municipal framework. Data were cleaned and harmonized to ensure consistent municipal identifiers across sources and years, and to retain only municipalities with complete observations for all three time points. Indicators were transformed to a common direction so that “more” consistently reflects “better” conditions, and then normalized to enable comparability across variables and years. The composite index was constructed through a multivariate weighting strategy that reduces redundancy among correlated inputs, stabilizes the contribution of dimensions, and produces a single well-being score per municipality per year. The final outputs were rescaled to the 0–1 interval to facilitate interpretation, mapping, and cross-sectional comparisons. The distribution of scores suggests (i) wide heterogeneity in municipal well-being within each year and (ii) meaningful temporal change for many municipalities. In aggregate terms, average well-being is higher in 2015 than in 2010, with a modest decline by 2020, indicating non-linear progress over the decade. Users should interpret changes across years as shifts in relative municipal positioning under a consistent measurement framework, not as direct evidence of policy impact. Values close to 1 represent municipalities with comparatively stronger achievements and fewer deprivations, while values close to 0 indicate persistent structural disadvantages. How others can use the dataset. The dataset is suitable for spatial and regional analysis, benchmarking and monitoring, prioritizing municipalities for targeted interventions, and replicating or extending municipal well-being measurement frameworks. It can be directly joined to GIS layers using municipal identifiers and used as an outcome variable in descriptive, comparative, or spatial econometric applications.
Files
Steps to reproduce
The dataset was constructed to provide a year-comparable municipal measure of social well-being in Mexico for 2010, 2015, and 2020. Input data were compiled from publicly available municipal indicators aligned to Mexico’s official geostatistical municipal framework. Two types of information were integrated: (i) municipal human development achievements (health, education, and income components) and (ii) multidimensional poverty-related deprivations (social and basic services dimensions). All inputs were harmonized to a common municipal identifier to ensure consistent matching across sources and benchmark years. Data processing followed a reproducible pipeline. First, raw tables were imported and standardized (consistent variable names, types, and municipal identifiers). Second, quality control rules were applied to remove records with missing or zero values in required fields and to eliminate duplicates. Third, municipal identifiers were harmonized to address administrative inconsistencies across years, and the analytic sample was restricted to municipalities with complete observations for all three time points, yielding 1,877 municipalities in the final dataset. Fourth, variables were directionally aligned so that higher values consistently represent better conditions; deprivation-type indicators were recoded (inverted) prior to aggregation. Fifth, variables were normalized to make them comparable across scales and years, and a multivariate weighting/regularization strategy was applied to reduce redundancy among correlated inputs and to stabilize the composite measurement. Finally, the resulting municipal scores were rescaled to the 0–1 interval and exported as three year-specific index fields (Efecto_2010, Efecto_2015, Efecto_2020). The workflow was implemented in Python and can be reproduced by executing the scripts in sequence: (1) import inputs; (2) clean and harmonize identifiers; (3) recode indicator direction; (4) normalize variables; (5) compute multivariate weights/regularization and aggregate to the composite index; (6) rescale outputs and export the final CSV. The released dataset contains the harmonized municipal code (cve_propia), administrative names (nom_ent, nom_mun), and the three normalized index scores for 2010, 2015, and 2020.
Institutions
- Autonomous University of TamaulipasTamaulipas, Ciudad Victoria