A system that protects fossil-fuel profits while accepting preventable deaths deserves to be challenged at its foundations. We cannot build a climate strategy around governments deciding, eventually, to do enough.
Our position: organized, nonviolent pressure for a rapid fossil-fuel phaseout, with food, water, healthcare and energy access protected throughout the transition. Survival should come before profit.
Urgency demands honest numbers. The default example assumes 10 million disruption deaths; that is not a predicted toll. The emissions-to-mortality link is an unfitted proxy, and each study covers only part of the harm. Change the inputs and see how much the answer depends on them.
Read these as conditional mortality comparisons. A crossing cannot establish that a tactic is justified. The model does not estimate the consequences of an attack on infrastructure.
Conditional mortality crossing
Unknown
Enter an assumed disruption toll.
Disruption threshold by 2100
—
Cumulative disruption deaths that would cancel the assumed climate benefit by this date.
Intervention minus reference
Unknown
Positive = more deaths under intervention.
CO₂ emissions avoided—
Warming avoided—
Climate burden avoided—
Cumulative comparison
Move over the chart to inspect a year.
Calendar year × intervention start
Hover or tap to inspect a cell. Click a row to use that start year in the main charts. With keyboard focus on the map, use arrow keys to inspect and Enter to select.
Every row holds the cut, the initial toll, recovery and the mortality response fixed; only the start year moves. The toll is not recomputed for later population or vulnerability. Counting always begins in 2026 and stops at the column’s year, so for late starts part of the toll falls beyond the horizon.
Challenge the assumptions.
Each cell gives the first year-end when the intervention catches up and remains at or below reference mortality through the selected horizon. These are conditional sensitivity tests, with no probabilities assigned. Delay, ramp, persistent shortages and other effects follow your settings.
Crossing years by assumed initial disruption toll and avoided climate mortality share
— = no sustained crossing by the horizon. “Never higher” = the intervention never accumulates more deaths than the reference. “Equal” = the two never differ.
A crossing is the first year-end when an initially positive difference reaches zero or below and stays there through the horizon. The headline reports that year and promises nothing beyond the horizon. Counts are end-of-year, with no financial discounting. Background mortality common to both scenarios cancels out.
What COVID can tell us
Global fossil CO₂ emissions fell 5.4% in 2020 and rebounded the next year. The pulse preset applies that percentage for one year from your selected start. The avoided CO₂ stays out of the atmosphere after the rebound, but it is small against the accumulated stock. This is a scale illustration, not a reconstruction of the pandemic. Global Carbon Budget 2021.
A separate multi-pollutant study projected about 0.01 ± 0.005°C less warming by 2030 from the pandemic response. That is a modeled figure, not a measurement, and it includes aerosol and other effects this CO₂-only pulse omits, so it is not a calibration target here. Cleaner air improves health quickly; losing cooling aerosols can add warming. Forster et al. (2020).
The harm these curves leave out
Each item below is a mechanism the model does not compute, not one shown to be zero.
Warming follows IPCC’s approximately linear cumulative-CO₂ relationship. Methane, aerosol changes, ocean transients, tipping dynamics, crop losses and failing health services are not modeled.
The temperature-to-mortality power law is an unfitted proxy applied to the selected reference; it is not that study’s response function. Reference mortality stays fixed when the stylized climate pathway changes, so this is a sensitivity experiment, not a coupled reproduction of any study.
No source counts every climate death. The World Bank reference covers 69 countries and five health risks. The academic temperature studies omit other pathways. Institutional affiliation alone does not establish the direction or size of bias.
Carleton options keep net cold-related benefits and use your constant population to turn rates into counts. “Without adaptation” removes modeled protection from income growth and adaptation; it is a structural scenario, not an upper bound. Bressler’s annual timing is reconstructed from one published cumulative endpoint.
Disruption deaths and recovery are inputs. Food, medicine, water, energy substitution and allocation need their own models; imported annual trajectories can carry their outputs.
Background mortality common to both scenarios cancels. Changed demographics, competing risks, life-years, the distribution of harm and ecological losses are not simulated. The sensitivity grid assigns no probabilities.
Sources and provenance
IPCC AR6 WGI, SPM D.1.1: best estimate 0.45°C per 1,000 GtCO₂, likely range 0.27–0.63°C. This supports the carbon-to-warming step only; the mortality scaling and the disruption toll are assumptions of this model.
Bressler and Carleton share underlying evidence. Do not add their estimates together or read their spread as independent samples. The README documents data extraction, coverage, interpolation and every formula.
Check individual years
Cumulative excess deaths in millions at each year-end. The CSV export has every year.
The political argument
Whose lives does the economy protect?
Our starting point is anti-capitalist: an economy exists to secure the conditions of life. Fossil-fuel profits cannot take precedence over that obligation. Who bears foreseeable harm, and who decided they would, belongs at the center of climate politics.
01 / THE CONCEPT
Social murder
In 1845, Friedrich Engels described a society that placed workers where they would die early, knew that they would, and let the conditions stand. He called it social murder. Here the term asks who is responsible when preventable deaths are accepted as a cost of keeping an economic order running.
That is a moral and political argument. The mortality curves cannot establish liability, criminal or otherwise.
The IPCC’s 2023 synthesis finds, with high confidence, that vulnerable communities which contributed least to climate change bear a disproportionate share of its impacts.
This model does not count deaths by class, race, occupation or disability. An aggregate total cannot say how fairly the burden is shared. That is a reason to demand the accounting, not to skip it.
A rapid phaseout must include income and transition support for the workers whose jobs end, reliable essential services, and a public say in the decisions. Protection should not depend on ability to pay or usefulness to an employer.
These are demands about how society should be organized. The calculator has not quantified them, and it cannot.
Three kinds of claims. Three things to scrutinize.
Published evidence
Mortality totals or rates from named studies. Each covers particular places, causes and years, by a particular method.
Modeling assumptions
The disruption toll, timing, emissions pathway and mortality response that construct each comparison. Change one and the result can change.
Political judgment
The argument for putting human needs before fossil-fuel profits. A mortality estimate does not attribute deaths to an economic system or settle which actions are justified.
Phase out fossil fuels. Keep people alive.
Organize nonviolent pressure with affected workers and communities. Demand a timetable for emissions cuts, secure livelihoods, accessible essential services and public accountability. Publish the assumptions alongside every chart, including results that challenge the argument.