VoxEU Column Environment

Carbon and inequality: From Kyoto to Paris

The COP21 conference faces a severe problem when it comes to funding climate adaptation in developing countries. This column examines novel strategies to increase the funding. The strategies are based on high individual carbon emitters wherever they are in the world, rather than according to the responsibilities of high-emitting countries. To this end, a global distribution of individual income and CO2e emissions is constructed. 

The 21st United Nations Climate Change Conference is being held in Paris. One key issue to be debated relates to the financing of climate adaptation in developing countries, i.e. how to finance investments in infrastructure and human capital to make societies more resilient to the impacts of climate change. Climate adaptation is currently under-funded: the needs range from €100bn to €300bn a year according to the United Nations (UNEP 2014), but less than €10bn is currently dedicated to climate adaptation funds (OECD 2015).

Our new study (Chancel and Piketty 2015) examines novel strategies to increase the volume of funding for climate adaptation in developing countries. In these strategies, efforts are determined according to the emissions of high individual carbon emitters wherever they are in the world, rather than according to the responsibilities of high-emitting countries. In order to do so, we construct a global distribution of individual income and CO2e emissions (CO2 and other Green House Gases).

Our methodology

Our methodology is based on the pioneering work of Chakravarty et al. ( 2009), further refined along three lines: we aim at better representing top individual incomes and emitters; we focus on consumption-based emissions to better represent responsibilities associated to climate change (for example, CO2e emissions due to the production in China of smartphones used in Europe are attributed to Europeans, not to Chinese); and we also provide dynamic estimates ranging from the Kyoto Protocol in 1998 to 2013.

We start by reconstructing a global distribution of income, using data from Lakner and Milanovic (2015), obtained from income and consumption surveys. One issue with survey data is that it often misrepresents top incomes. We follow Anand and Segal (2014) and combine Lakner-Milanovic estimates with fiscal data from the World Top Income Database (WTID 2015) in order better represent inequality at the top of the distribution. In effect, we regress existing top 1% income shares (from fiscal sources) on decile shares (from survey sources) so as to predict top income shares in countries without fiscal data. We stress that our estimates should not be seen as definitive values for the world income distribution, but as a first attempt to combine global income distributions with top incomes data.

We then attribute CO2e emissions to each income group using consumption-based CO2e data provided by Peters and Andrew (2015). For each country, we use a simple income-CO2e elasticity model and allocate national CO2e emissions to income groups assuming different income-CO2e elasticity values. We use a central value of 0.9 as suggested by a review of country-level income-CO2e elasticity studies.

Trends in the global distribution of carbon emissions (1998-2013)

Our results show that global CO2e emissions inequalities between individuals decreased from the Kyoto Climate Protocol in 1998 to 2013, due to the rise of top and mid-income groups in developing countries and the relative stagnation of incomes and emissions of the majority of the population in industrialised economies. Income and CO2e emissions inequalities, however, increased within countries over the period. Global CO2e emissions remain highly concentrated today: the top 10% emitters contribute to 45% of global emissions, while the bottom 50% contribute to 13% of global emissions. The top 10% emitters live on all continents, with one third of them from emerging countries (Figure 1). 

Figure 1. Breakdown of top 10, middle 40 and bottom 50% CO2e emitters

Note: Among the top 10% global emitters, 40% of CO2e emissions are due to US citizens, 20% to the EU and 10% from China.

Source: Chancel and Piketty (2015).

Such results are moderately sensitive to elasticity choices we make (Table 1). Higher within-country elasticities imply higher global CO2e concentrations: assuming an elasticity of 0.7 within countries, the top 10% of global emitters are responsible for 40% of total emissions. The figure rises up to over 50% with an elasticity of 1.1. We also provide in our online data files (available here) a scenario in which elasticities vary across countries. In all cases, the top 1% world emitters (about 70 million individuals out of 7 billion) pollute approximately as much as the bottom 50% world emitters (3.5 billion individuals out of 7 billion).

Table 1. CO2e concentrations for different elasticity values

Note: Assuming an income-CO2e elasticity of 1.1 within countries, the top 1% emitters are responsible for 19% of global emissions in 2013.

Source: Chancel and Piketty (2015).

Our estimations show that the top 1% richest Americans, Luxemburgers, Singaporeans, and Saudi Arabians are the highest individual emitters in the world, with annual per capita emissions above 200tCO2e. At the other end of the pyramid of emitters lie the lowest income groups of Honduras, Mozambique, Rwanda and Malawi, with emissions 2,000 times lower, at around 0.1tCO2e per person per year. In the middle of the world distribution of emitters (between 6 and 7tCO2e per year) lie groups such as the top 1% richest Tanzanians, the Chinese 7th income decile, the French 2nd income decile and the 3rd German income decile.

Middle and upper classes of emerging countries increased their CO2e emissions more than any other group within the past 15 years, with cumulated per capita growth rates over 30%, about twice the world average (17%) and much higher than growth rates for the majority of the population in industrialised countries. This led to a reduction in the global dispersion of CO2e emissions – especially between the middle of the income distribution and the top (Figure 2). However, the inequality of CO2e emissions increased between the bottom of the distribution and the middle. While these trends, if continued, are positive from an income point of view (emergence of a global middle class), they constitute a real challenge for future global CO2e emissions levels.

Figure 2. How did CO2e emissions grow from Kyoto to Paris for different groups of emitters? 

Note: the group representing the 2% lowest CO2e emitters in the world, saw its per capita CO2e emissions level decrease by 12% between 1998 and 2013.

Source: Chancel and Piketty (2015).

Our estimates also show that within-country inequality in CO2e emissions matters more and more in explaining the global dispersion of CO2e emissions. In 1998, one third of global CO2e emissions inequality was accounted for by inequality within countries. Today, within-country inequality makes up 50% of the global dispersion of CO2e emissions (Figure 3). It is then crucial to focus on high individual emitters rather than high-emitting countries.

Figure 3. World CO2e emissions inequalities: Within and between country importance

Note: in 2008, the within-country component of the Theil index was of 0.35 and the between-country component of 0.40, i.e. between-country inequalities contributed to 53% of total inequalities - as measured by the Theil index.

Source: Chancel and Piketty (2015).

Prospects for an equitable adaptation fund

The new geography of global emitters calls for climate action in all countries. While developed and developing countries already engaged in mitigation efforts, contributions to climate adaptation funds remain almost entirely financed by developed nations, and for the most part by Europe (with more than half total contributions). If it is necessary to increase the volume of adaptation finance from developed countries, our study shows that upper income groups of emerging countries, who benefited from income growth and resulting CO2e emissions growth over the past decades, could also participate in such funds. With the contributions of South Korea, Mexico or Columbia to the Green Climate Fund, emerging and developing countries are committing to finance adaptation and broke the standard developed-developing countries divide which prevailed so far. However, their contributions remain symbolic at this stage (less than 1% of all global adaptation funds) and the equity logic behind adaptation funding remains unclear.

The study suggests novel strategies to increase global climate adaptation funding, in which individual CO2e emissions are the basis for contributions. In order to better align these contributions to the new distribution of high emitters, we first examine the implications of a global progressive carbon tax to raise €150 billion required annually for climate adaptation (Table 2). In Strategy 1, all emitters above world average emissions (i.e. all individuals emitting more than 6.2t per year) contribute to the scheme in proportion to their emissions in excess of this threshold. North Americans would contribute to 36% of the fund, versus 20% for Europeans and 15% for China. In Strategy 2, the effort is shared by all top 10% emitters in the world (i.e. all individuals emitting more than 2.3 times world average emissions), again in proportion to their emissions in excess of this threshold. North Americans would then pay 46% of the tax, versus 16% for Europeans and 12% for China. In Strategy 3, the effort is shared by all top 1% emitters in the world (i.e. all individuals emitting more than 9.1 times world average emissions). North Americans would then contribute to 57% of efforts, versus 15% for Europeans and 6% for China. In these strategies, the share of Europe would decrease in proportion, but increase in absolute terms. In Strategy 3, the most favourable to Europeans, the volume of finance coming from Europe would reach €23 billion, about four times its current contributions.

Table 2. Who should contribute to climate adaptation funds?

Note: North Americans represent 46.2% of global emissions released by individuals who emit 2.3 times more than the global average. Individuals who emit more than 2.3 times average emissions (14.3 tCO2e per year) belong to the top 10% emitters.

Source: Chancel and Piketty (2015).

We also discuss possible implementations via country-level carbon and income taxes or via a generalised progressive tax on air tickets to finance the adaptation fund. A tax on air tickets has already been implemented in a handful of countries and is currently used to finance development programmes. Taxing all business class tickets in the world at a rate of €180 and all economy class tickets at a rate of €20 would yield €150 billion required for climate adaptation every year. This latter solution might be easier to implement but less well targeted at top emitters.


Anand, S., Segal, P. (2014), "The Global Distribution of Income", in Handbook of Income Distribution, Elsevier, Amsterdam.

Chakravarty, S., A. Chikkatur, H. de Coninck, S. Pacala, R. Socolow and M Tavoni (2009), "Sharing global CO2 emission reductions among one billion high emitters", Proceedings of the National Academy of Sciences 106, 11884–11888.

Chancel, L. and T. Piketty (2015), "Carbon and inequality: From Kyoto to Paris. Trends in the global inequality of carbone emissions (1998-2013) & Prospects for an equitable adaptation fund", Paris School of Economics, Paris.

HBS (2015), "Climate Funds Update", Heinrich Böll Stiftung.

Lakner, C. and B. Milanovic (2015), "Global income distribution: From the fall of the Berlin Wall to the Great Recession", World Bank Economic Review (advance access).

Peters, G.P. and R. Andrew (2015), "Consumption and Production GHG data from GTAP-CICERO Input Output tables".

UNEP (2014), The Adaptation Gap Report 2014, United Nations Environment Programme (UNEP), Nairobi.

WTID (2015), World Top Income Database.

6,509 Reads