A major new study from the Potsdam Institute for Climate Impact Research (PIK) offers an unprecedented look at how the carbon intensity of household consumption varies across 88 countries representing about 5 billion people. The findings could help governments design climate policies that are both effective and socially fair, particularly as carbon pricing and the removal of fossil fuel subsidies can increase household costs. Surprisingly, the researchers found that the greatest differences in climate-policy burdens are often not between rich and poor households, but among households within the same income groups. Factors such as vehicle ownership, location and energy use can play a decisive role. The research, published in the Journal of Environmental Economics and Management, is accompanied by an interactive online carbon price calculator.

“Uncertainty around the social impacts of climate policy is a problem for many governments worldwide,” said Leonard Missbach, a PIK researcher and lead author of the study. Governments often lack detailed information about how the costs of climate measures are distributed across their populations and how vulnerable households can be compensated without undermining public support. To address this gap, the researchers combined a vast international dataset with machine learning. Their analysis draws on national household surveys covering 1.7 million representative households, which anonymously reported their spending patterns. Together, the surveys reflect living conditions across countries containing most of the world’s population.

The researchers combined household expenditure data with estimates of the carbon emissions associated with individual purchases. These included direct emissions from products such as petrol and heating oil, as well as indirect emissions embedded in other goods and services. This produced an unusually detailed collection of individual household “carbon footprints”. The team then examined how climate policies that raise the cost of carbon could affect households relative to their incomes. Large differences in these additional burdens within a country can create more cases of financial hardship and make it considerably more difficult for governments to design compensation measures that reach the people who need them most.

One of the study’s central findings challenges traditional approaches to social compensation. Across all 88 countries, differences in the impact of climate policy between rich and poor households were significantly smaller than differences among households within individual income groups. Policies focused mainly on income, such as graduated cash transfers or tax rebates, may therefore fail to protect many vulnerable households. In some cases, the researchers warn, well-intentioned compensation schemes could even increase inequality in how climate-policy costs are distributed. The findings suggest that governments may need to look beyond household income when deciding who should receive support.

Using machine learning, the researchers identified several characteristics that help explain why households with similar incomes can face very different climate-policy burdens. Car and motorbike ownership emerged as important factors, alongside geographical differences such as whether households are located in urban or rural areas. Energy consumption also played a major role, including the fuels used for cooking, lighting and heating, access to electricity grids and ownership of large household appliances. However, the importance of these factors varied widely between countries. Motorcycle ownership was particularly relevant in Niger, Burkina Faso and Togo, while the urban–rural divide was more important in Latvia, Sweden and the Czech Republic. Cooking fuels were especially significant in Nicaragua and India, whereas household appliances helped explain differences in Switzerland and the Philippines.

To make these complex global patterns easier to understand, the research team grouped countries into ten clusters based on similarities in the distribution of household carbon intensity. The approach could encourage governments facing comparable challenges to exchange ideas and learn from one another, although the researchers deliberately stopped short of recommending specific policies for individual countries. PIK researcher and study co-author Jan Steckel said the findings are intended to guide policymakers and help non-governmental organisations better understand the social effects of climate measures. The researchers also note that carbon pricing and the removal of fossil fuel subsidies generate government revenue that can be redirected towards compensation, potentially making social equity easier to achieve than under regulations, bans or limits that do not create additional public funds.

More information: Leonard Missbach et al, The heterogeneous effects of climate policy on households: Evidence from 88 countries, Journal of Environmental Economics and Management. DOI: 10.1016/j.jeem.2026.103382

Journal information: Journal of Environmental Economics and Management Provided by Potsdam Institute for Climate Impact Research (PIK)