New Metric Reveals Carbon’s Deadly Price — In Lives
A new analysis extends the classic DICE climate–economy model to directly account for heat-related mortality—and puts a stark number on carbon’s human cost. The study introduces a “mortality cost of carbon” (MCC): in 2020, each additional metric ton of CO₂ is linked to 2.26×10⁻⁴ expected deaths, implying roughly one excess death per 4,434 tons emitted between 2020 and 2100. Incorporating these mortality damages boosts the 2020 social cost of carbon from $37 to $258 per ton and flips the “optimal” policy from slow cuts to full decarbonization by 2050. Under that path, the model projects warming near 2.4 °C and 74 million fewer deaths this century.
How deadly is a ton of carbon dioxide? A study by R. Daniel Bressler answers with a new metric: the mortality cost of carbon. By weaving temperature-driven death risks into the DICE-2016 integrated assessment model, the paper estimates that one additional ton emitted in 2020 causes 2.26×10⁻⁴ expected excess deaths over 2020–2100. Flip that number and the picture sharpens: about one death per 4,434 tons of CO₂—roughly the lifetime emissions of 3.5 average Americans, or 12.8 people worldwide. Scale it up and the stakes become immediate: a one-million-ton cut in 2020 emissions would avoid an expected 226 deaths.
Accounting for those lives dramatically changes the economics. The 2020 social cost of carbon jumps from $37 to $258 per ton when mortality is included, making many faster, deeper cuts appear cost-effective. The model’s optimal policy pivots from gradual reductions after mid-century to steep, near-term cuts culminating in net-zero by 2050. In that optimal path, warming peaks near 2.4 °C rather than 3.5 °C, and the model projects 74 million lives saved compared with the baseline. The timing matters because heat-related mortality accelerates nonlinearly with temperature; as global heat climbs beyond ~2 °C, small extra warming inflicts disproportionately higher death risks, especially late in the century.
The MCC is intentionally simple: it reports deaths rather than dollars and avoids discounting or valuing lives. The study also recomputes the SCC—which must monetize and discount mortality—by calibrating welfare loss per lost life year and, crucially, valuing lives equally regardless of income or location. Sensitivity tests show policy and MCC magnitudes depend on the warming path: on the optimal path, the MCC roughly halves to ~1.07×10⁻⁴ deaths per ton, implying one death per ~9,318 tons.
There are caveats. The mortality function focuses on temperature-related deaths and omits other channels (disease, conflict, floods, food insecurity) unless embedded in source studies; adaptation is modeled in reduced form; and other known DICE limitations remain. Those gaps likely mean the central estimates are conservative for total harm and could push optimal policy toward even faster cuts if fully included. Still, by translating carbon into lives at the margin—and showing how that reshapes costs, benefits, and timelines—the paper gives policymakers, firms, and households a clearer yardstick for the human stakes of their emissions decisions.