The Distributional Effects of Carbon Taxation: Lessons from the French Yellow Vests Movement

About the Author

Dr. Thomas Douenne

After completing his Ph.D. at the Paris School of Economics, Dr. Douenne began as an Assistant Professor in Economics at the University of Amsterdam. His main research field is Environmental Economics, which he approaches through both theoretical and empirical methods. In particular, Dr. Douenne has studied the economic consequences of environmental disasters (mostly theoretically), the distributive effects of carbon taxation (empirically), and issues related to public support for climate policies. The ultimate goal of his research is to understand and find solutions to overcome the obstacles to the implementation of ambitious and effective environmental policies.

Public support for carbon taxation is often lacking

Despite being advocated by economists as an effective tool to reduce greenhouse gas emissions, carbon taxation is still struggling to find its place in the policy mix of most countries. While France had introduced a carbon tax since 2014, the increasing rates of the tax met with strong public opposition when fuel prices peaked in October 2018.[1] The protests of the “Yellow Vests” quickly compelled the government to abandon the planned tax increases, leaving a lot of uncertainties about the future of French climate policies.

Carbon taxation alone is regressive

The Yellow Vests raised concerns about the negative impact of the tax on households’ purchasing power, and in particular the burden it imposed on the least well-off. In a recent paper (Douenne, 2020), I estimate the monetary distributional effects of the last increases of French taxes on fossil fuels.[2] I find that before revenue-recycling, the policy was indeed regressive, imposing a higher burden on the poorest households. For example, I find that as a share of their income, the increase in taxes was 2.6 times higher for households in the bottom income decile as compared to those in the top decile. The ultimate distributional effect of a carbon tax policy depends, however, on the use of its revenues. In the case of the French policy, the increases in the carbon tax rate were concomitant to reductions in labor and capital taxes, following the so-called “double dividend” strategy. While this strategy aimed to stimulate the economy by using the carbon tax to reduce distortionary taxes, the gains from these tax cuts mainly favored high-income households—and in particular the very top of the income distribution (Ben Jelloul et al, 2019)[3]—making the fear of the Yellow Vests all the more legitimate.

Redistributing its revenues equally leads to a progressive policy

In my paper, I investigate what the distributional effects would have been if the French government had instead redistributed tax revenues through equal lump-sum transfers.[4] This strategy known as the carbon tax and dividend has recently been defended by many economists in order to “maximize the fairness and political viability of a rising carbon tax” (The Wall Street Journal). My results (see Figure 1) confirm that the policy would indeed have been progressive in this case and beneficial to the first seven income deciles.

Figure 1: Average net transfers after equal lump-sum transfers, by income decile.

Note: Transfers are expressed per consumption unit to adjust for household size. As lower (resp. higher) income households emit less (more) CO2, the taxes they pay on their carbon emissions are lower (higher) than the transfers they receive, hence positive (negative) net transfers. 

Source: Douenne, 2020.

Horizontal distributional effects are more difficult to deal with

Given this result, equal lump-sum transfers could appear as a straightforward policy response to the distributional effects of carbon taxation. Unfortunately, things are not that simple. While carbon taxation generates transfers along the income dimension (what we generally refer to as “vertical distributional effects”), it also generates transfers between households for a given income level (called “horizontal distributional effects”). In the case of the French policy, I find that transfers within income groups were in fact much larger than transfers between income groups. Assuming as before that the tax revenue would have financed equal transfers (i.e. a tax and dividend policy), Figure 2 shows that 25% of households in the bottom income decile would have lost more than the median household in the top decile.

Figure 2: Distribution of net transfers after equal lump-sum transfers, by income decile.

Note: Transfers are expressed per consumption unit to adjust for household size. The figures read as follows: after equal lump-sum transfers, 25% of households belonging to the bottom income decile are expected to lose more than 32€ annually, while 50% of households belonging to the top income decile are expected to lose less than 28€ annually. 

Source: Douenne, 2020.

In analyzing the determinants of this heterogeneity, I find that households’ past investments—such as their location (urban vs. peri-urban or rural) and the type of energy they use for heating—are critical factors, but unexplained heterogeneity remains very important, which makes it difficult for the State to identify the losers of carbon taxation. Consequently, if governments want to avoid a large proportion of poor households losing out in the short run, they must allow greater redistribution through a more progressive scheme.[5] In the longer run, both the environmental success of the reform and the mitigation of its distributional effects require households to replace their polluting equipment. Given the links between the carbon externality and other market failures—such as credit constraints, imperfect information, or bounded rationality—it may be desirable to implement additional cost-effective policies than the carbon tax alone to accelerate this transition and reduce its cost for the least well-off (Stern & Stiglitz, 2017).

Public support for policies depends on citizens’ beliefs

Horizontal distributional effects thus raise concerns about the fairness and political prospects of a carbon tax and dividend policy. Yet, although imperfect, one may wonder whether this progressive tax policy would meet with public support if it were proposed today in France. This is the question we investigate in a study co-authored with Adrien Fabre (Douenne & Fabre, 2020b) based on a survey carried out shortly after the Yellow Vests movement. From a large sample representative of the French population, we show that although a carbon tax and dividend would be financially beneficial for 70% of households, it would be supported by only 10% of them. Respondents have overly pessimistic beliefs about the outcomes of the policy: the vast majority believe that they would lose out financially and that the policy would be environmentally ineffective and regressive. Our econometric analysis also shows that beliefs causally influence policy approval, so that convincing people of the actual impacts of the policy could lead to majority support. However, pessimism seems to be well entrenched, with respondents being much less receptive to new information when it provides arguments in favor of the policy. Looking at the determinants of this asymmetry, we find that political views and identity shape beliefs, leading those most opposed—such as the Yellow Vests—to process new information more pessimistically. Thus, suspicion against carbon taxation leads people to more pessimistic views about its outcomes, which further reinforce their rejection of the policy.

Lessons from the Yellow Vests

With climate change awareness and concerns now widespread in many countries, policy makers have a great opportunity to implement ambitious climate policies.[6] Nevertheless, the design of environmental tax reforms presents us with major challenges. First, implementing carbon taxation will necessarily create winners and losers, and there is no simple solution to decide on the right way to share the burden. Second, even if citizens were to agree on a given distribution of efforts, their skepticism towards policies in general and carbon taxation in particular leads them to pessimistic beliefs that undermine their support, including for policies that would make them better off. The ecological transition is unlikely to occur without public consent. To move forward with ambitious climate policies, governments must learn from the Yellow Vest movement and its call for greater transparency, deliberation, and redistribution. French President Emmanuel Macron has since launched a Citizens’ Climate Convention, gathering 150 citizens randomly selected to learn, debate, and propose solutions to meet France’s climate goals. By creating a bridge between citizens and climate experts, we may hope that this convention will succeed in its mandate to bring forward effective solutions supported by the people to combat climate change “in a spirit of social justice”.


  • Mahdi Ben Jelloul, Antoine Bozio, Thomas Douenne, Claire Leroy, The 2019 French budget: impacts on households, Note IPP, 2019.
  • Pierre C. Boyer, Thomas Delemotte, Germain Gauthier, Vincent Rollet, & Benoît Schmutz. Les déterminants de la mobilisation des Gilets jaunes. Revue économique, 71, 2020.
  • Dominique Bureau, Fanny Henriet, & Katheline Schubert. Pour le climat : une taxe juste, pas juste une taxe. Les notes du conseil d’analyse économique, (50):12, 2019.
  • Climate Leadership Council, Economists’ Statement on Carbon Dividends, 2019.
  • Thomas Douenne. The vertical and horizontal distributive effects of energy taxes : A case study of a French policy. The Energy Journal, 41(3) 2020.
  • Thomas Douenne & Adrien Fabre. French attitudes on climate change, carbon taxation and other climate policies. Ecological Economics, 169(C), 2020a.
  • Thomas Douenne & Adrien Fabre. Yellow Vests, Pessimistic Beliefs, and Carbon Tax Aversion, 2020b.
  • Nicholas Stern & Joseph E. Stiglitz. Report of the High-Level Commission on Carbon Prices. Technical report, Carbon Pricing Leadership Coalition, 2017.


[1] The French carbon tax is a tax on fossil fuels. From 7€/tCO2 in 2014, it has gradually increased with the objective of reaching 86.2€/tCO2 in 2022. Since 2018, its level remains fixed at 44.6€/tCO2 since all planned increases have been abandoned following protests by the Yellow Vests. For more information on the determinants of the Yellow Vests movement, see Boyer et al (2020) (in French) or the more synthetic summary in Section 2 of Douenne & Fabre (2020b) (in English).

[2] I estimate the effects of the policy changes from 2016 to 2018, i.e. the last increases before the Yellow Vests protests. This primarily includes an increase in the carbon tax rate from 22 to 44.6€/tCO2, as well as an additional 0.026€/liter increase for the diesel tax.

[3] Ben Jelloul et al (2019) estimate the distributive effects of the tax and social security reforms of the French 2018 and 2019 budget, including measures that were taken in response to the Yellow Vests movement.

[4] These transfers are only indexed to the size of households, using the number of consumption units as defined by Eurostat.

[5] See for instance the proposal of the French Council for Economic Analysis in Bureau et al (2020).

[6] For an analysis of attitudes towards climate change and climate policies in France after the Yellow Vests movement, see Douenne & Fabre (2020a) based on the same survey as Douenne & Fabre (2020b).