Energy share in total consumption (regression-adjusted)

Figure 1: Energy share in total consumption (regression-adjusted)

Non-Technical Summary

Carbon taxes on households and firms are a key tool for reducing greenhouse gas emissions. However, their costs are unevenly distributed across the population. These inequalities can undermine public support for climate policies, as demonstrated by France’s Yellow Vest protests.

Using detailed French data on households, workers and firms, the paper first documents large differences in exposure to fossil fuels across locations. Rural households spend 2.8 times more on fossil fuels than Parisians (Figure 1). This is mainly because they live in larger homes and rely more heavily on cars. By contrast, the share of spending devoted to fossil fuels is relatively similar across income groups. Location also shapes exposure through employment as rural workers are more likely to work in emissions-intensive sectors such as agriculture and manufacturing. The firms employing rural households emit 2.7 times more greenhouse gases than in Paris. Rural households are therefore particularly exposed to carbon taxation, both through their energy consumption and through its potential effects on employment and wages.

The paper then develops a model in which households differ in income, wealth, housing and location, and can move between regions. Mobility is crucial: if households could move freely, geographical differences in the burden of carbon taxation would largely disappear, because people could relocate in response to higher energy costs, wages and housing prices. However, the data show that moving is costly. Homeownership, borrowing constraints and relocation costs prevent many households from easily changing location. Some households are effectively “trapped” in places where the carbon tax is particularly costly.
 
Calibrating these mobility frictions to match the data, the model finds that a carbon tax reducing emissions by 10% generates welfare losses 56% larger in rural areas than in Paris: around −3.1% compared with −2%. Geography therefore creates substantial differences in the cost of the transition, even among households with similar incomes.

This has important implications for the design of climate policy. Income-based compensation alone is not enough when households cannot easily move away from high-cost locations. Transfers based on income can reduce inequality between rich and poor, but they do not fully compensate rural households. In the model, transfers that account for both income and location are much more effective at equalizing the costs of the transition. When households also value the reduction in climate damages, appropriately designed transfers can make the transition beneficial for households across income and geographic groups.

Geographic exposure and mobility constraints also play an important role in determining how households are affected. Accounting for these factors can therefore help inform the design of carbon taxation and the targeting of compensatory measures. Understanding and addressing these distributional effects will be particularly important as Europe prepares to implement its new carbon market, the EU-ETS 2, in 2028.

Keywords: Carbon Tax, Inequalities, Revenue Recycling, Spatial and Macroeconomic Models, Migration

Codes JEL : C61, E62, H23, Q43, Q58, R13.