Physical consumption data reveals hidden carbon inequalities beyond monetary-based accounting

Understanding how carbon footprints differ across households is essential for designing fair and effective carbon pricing mechanisms. However, such assessments can vary depending on whether carbon footprints are allocated based on the monetary value of goods and services purchased or physical quantities consumed, because expenditure differences may reflect not only differences in consumption volumes but also differences in product characteristics, quality, and prices. Here, using micro-level data from China's Household Survey linked with an environmentally extended input–output database, we compare monetary- and physical-based carbon accounting for consumption categories with information on both expenditures and physical quantities and examine the implications for carbon taxation. We find systematical differences in estimated emissions across household groups, ranging from −9% to +7% at the expenditure-quintile level. Differences are larger within individual categories, reaching −54% to +59% for footwear. These differences substantially alter the estimated distribution of carbon tax burdens across household groups. Physical-based accounting indicates higher carbon tax burdens for the poorest quintile, especially for essential goods such as food and residential energy. The comparison demonstrates how accounting choices shape assessments of carbon tax burdens and climate policy fairness.