The top 10% of health-care consumers account for 48% of health care's carbon emissions

Health systems are large emitters, and that has been established for years. What has not been established is who inside them is doing the emitting. National totals treat a country's health care as one undifferentiated block of carbon, which makes the obvious policy response — decarbonise hospitals — the only response on the table.

A study published in The Lancet Planetary Health breaks the block apart. Researchers estimated health-care carbon footprints for different income groups within 121 countries from 2005 to 2017, combining consumer expenditure surveys, national health expenditure data, and a global multiregional input-output model [s1].

The distribution

By 2017, the top 10% of high-spending health-care consumers accounted for 48% of the world's health-care carbon footprint — 1,128 megatonnes of CO2 equivalent [s1]. The bottom 50% accounted for less than 10% of the total [s1].

The concentration at the very top is sharper still. The top 1% contributed 2,857 kg CO2e per capita, more than eight times the global per-capita average and nearly 66 times the figure for the bottom half [s1].

This is not simply a rich-country versus poor-country split, though high-income countries did maintain consistently high footprints throughout [s1]. Upper-middle-income countries drove a substantial share of the growth, with total health-care emissions rising from 181 Mt CO2e in 2005 to 760 Mt CO2e in 2017 — more than a threefold increase in twelve years [s1].

The finding underneath the headline

The more consequential result is not the ratio but the elasticity. By estimating the relationship between health expenditure and embodied carbon, the authors found that health-care spending carries higher marginal emissions among high-spending groups than among low-spending groups [s1].

In plain terms: an additional dollar of health care bought by someone already consuming a great deal of it emits more than an additional dollar bought by someone consuming very little. The carbon intensity of care is not flat across the spending distribution.

That has a direct implication for the standard objection to demand-side climate policy in health, which is that reducing health-care consumption means denying people care. If the marginal emissions of the highest-spending decile are elevated relative to everyone else's, then the emissions reductions and the access expansions are not competing for the same territory.

What the scenarios show

The authors ran scenario analyses on that logic. Demand-side controls targeting carbon-intensive overuse of health care among the top 10–20% of consumers, applied while simultaneously advancing ambitious universal health coverage, delivered a 25% to 40% reduction in carbon footprints — and eased costs rather than raising them, without compromising health outcomes as modelled [s1].

That is a large number and it deserves the qualifier the authors attach to it. It is a modelled scenario, not an implemented policy, and it depends on the assumption that the consumption being removed at the top is overuse rather than needed care.

What the model cannot see

The methodology is worth understanding, because it determines what the result means.

Multiregional input-output modelling traces emissions embodied in economic flows. It works outward from what was spent, through supply chains, to the carbon those supply chains released. It is the standard tool for footprinting at this scale, and it is the only tool that could cover 121 countries and twelve years at all.

It is also blind to clinical content. The model knows that a high-spending consumer purchased a carbon-intensive bundle of health care. It does not know whether the imaging in that bundle was indicated, whether the admission was avoidable, or whether the patient was very ill. "Overuse" in the scenario analysis is an expenditure-derived category, not a chart-reviewed one. Distinguishing low-value care from expensive necessary care is a clinical question this method cannot answer, and any policy built on the finding would have to answer it separately.

The data also end in 2017 — a real constraint for a sector reshaped by the pandemic that followed. Whether the concentration held, tightened or loosened through the 2020s is outside this study's window.

The work was funded by the National Natural Science Foundation of China, the Youth Innovation Team of the China Meteorological Administration, and the Tsinghua-Rio Tinto Joint Research Center for Resource Energy and Sustainable Development [s1].

Why it changes the conversation

Health-system decarbonisation has been framed almost entirely as a supply-side engineering problem: cleaner energy, lower-carbon anaesthetic gases, less waste, better procurement. Those remain necessary.

This paper argues that a demand-side lever exists alongside them, and that it points at the opposite end of the equity distribution from where climate-health policy usually points. Most climate-health research concerns harms falling on the poorest populations. This one concerns emissions generated by the highest-consuming ones — and finds that expanding essential care to the bottom half of the world's health-care consumers is compatible with the sector cutting its footprint substantially [s1].