This study examines how alternative weighting schemes shape the measurement and diagnostic interpretation of multidimensional energy poverty in Mexico. Using ENIGH 2022 microdata, we construct a Multidimensional Energy Poverty Index from eight binary indicators of energy-service deprivation and compare four weighting rules within a common Alkire¿Foster framework: a hierarchical normative scheme (HIE), two empirical schemes based on Multiple Correspondence Analysis (MCA) and Principal Component Analysis (PCA), and equal weights (EW). The findings show that weighting choices are not neutral for the resulting diagnosis. While the four indices capture broadly similar patterns of household energy-service deprivation, they produce different classifications of households as energy poor once the identification threshold is applied. Incidence is the most sensitive aggregate, whereas intensity remains comparatively stable. HIE yields the most conservative diagnosis, while PCA and EW classify broader shares of households as energy poor with implications for the potential scale of policy intervention. These results suggest that multidimensional energy poverty measures should not be used mechanically for policy targeting. Instead, weighting choices should be made transparent and accompanied by sensitivity analysis, particularly in middle-income countries where high electricity coverage may coexist with persistent deprivations in clean cooking, thermal comfort, basic appliances and digital connectivity.
