Disparities in household carbon emissions across urban–rural and regional dimensions: Evidence from large-scale field surveys

Using large-scale household survey data collected in 2023 from 5,585 households across 108 Chinese cities, this study examines urban–rural disparities and regional heterogeneity in household carbon emissions (HCEs) and investigates their underlying driving mechanisms. Survey-based emission accounting supported by input–output analysis is combined with complementary ordinary least squares (OLS) regression and machine learning (ML) models to identify key drivers of per-capita HCEs. The results reveal a pronounced urban–rural gap in per-capita HCEs, with higher emissions in urban households driven mainly by service- and mobility-oriented consumption, while rural households rely more heavily on appliance- and basic energy uses. Regional heterogeneity is evident, with higher per-capita HCEs in the southern coastal and northeast regions and lower levels in the northwest and mid-Yangtze River regions, and interregional disparities narrower than in 2013. Daily consumption and energy-use factors emerge as the dominant drivers of per-capita HCEs, jointly accounting for over half of the total explanatory importance in the ML analysis. Household generational structure reduces per-capita HCEs primarily through larger household size, reflecting scale effects from shared resource use. These findings suggest that policy interventions should prioritize household energy transitions, low-carbon infrastructure provision, and consumption–lifestyle shifts to reduce per-capita HCEs while safeguarding living standards.