Challenges in measuring the distribution of carbon footprints: the role of product and price heterogeneity

Due to the lack of adequate data with information on household characteristics, their detailed consumption basket and the carbon content of each item in this basket, the distribution of the carbon footprint by households groups within a country is regularly computed using consumption data with the assumption that the carbon footprint from a product category is proportional to the spending on that product. Here, we explore the limitations of this proportionality assumption. We provide a systematic framework for discussing the gap between the true carbon footprint and the carbon footprint imputed with spending, a framework which is valid for any type of footprint. We show that two channels can bias the distribution of the carbon footprint: the heterogeneity in prices paid and the heterogeneity in the carbon intensity of products purchased. We rely on French household budget survey data to illustrate how these channels operate and interplay, and use cooking oils as a case study of the interplay of the two channels. We find that wealthier households pay higher prices. Because of the price heterogeneity, the carbon footprint tends to be overestimated for top income and underestimated at the bottom. We also use consumption of cooking oils as a more-in depth case study of the carbon intensity heterogeneity. This case shows that it can lead to an outcome similar to price heterogeneity.