The International Energy Agency's Global Energy Review 2026, published on April 20, is the first full audit of the 2025 energy system. Low-emissions sources — solar, wind, nuclear, hydropower, and other renewables combined — contributed nearly 60 percent of the growth in global energy demand. [1] Solar PV alone delivered more than 25 percent of that demand growth, the first time on record that a modern renewable source was the single largest contributor. [2] Natural gas came second at roughly 17 percent; oil at 15 percent; the remainder from solid bioenergy, waste, and coal. Overall demand growth slowed to 1.3 percent. Global energy-related CO2 emissions rose approximately 0.4 percent — the smallest non-crisis increase in decades — reaching a record 38 billion tonnes. [2] The paper's Wednesday feature established the arithmetic. Today's read sharpens what follows from it.

The three most durable findings in the report are the 60-percent headline, the 0.4-percent emissions figure, and an underappreciated datum: advanced-economy emissions rose 0.5 percent in 2025, faster than the 0.3 percent in emerging markets and developing economies. This is the first time outside a post-crisis rebound that advanced-economy emissions rose faster than emerging-market emissions since the 1990s. [2] The reversal is not a narrative artifact. In the United States, a cold winter, higher natural-gas prices that drove gas-to-coal switching in electricity, data-centre demand growth, and industrial expansion combined to push 2025 energy-related emissions up at the second-highest rate since 2000 excluding post-recession years. [2] Europe's numbers were smaller. China's emissions fell outright on renewables and structural industrial shifts. India's were flat for the first time since the 1970s. The United States, as a single country, accounted for a disproportionate share of the advanced-economy uptick.

This is the context in which Environmental Protection Agency Administrator Lee Zeldin delivered Tuesday's Earth Day remarks. Zeldin's message, as the paper covered Wednesday, framed the endangerment-finding repeal as vindication and asked climate skeptics to celebrate. [3] The repeal removes the legal basis on which the federal government has regulated carbon emissions from power plants, vehicles, and industrial sources since 2009. The IEA's 2025 audit describes a global fuel mix in which low-emissions growth is outpacing the fossil stack at the margin. The endangerment repeal is a domestic regulatory choice; the IEA's 2025 numbers describe the global mix the repeal is choosing to exit. The gap between the two is not rhetorical. It is measured.

The avoided-demand arithmetic sharpens the measurement. The IEA estimates that clean-energy deployment since 2019 has avoided more than 35 exajoules of annual fossil-fuel demand in 2025 — roughly 7 percent of global fossil-fuel use, or the combined 2025 energy demand of Latin America. [4] Avoided coal demand alone (approximately 800 million tonnes of coal equivalent) exceeded the entire coal use of India in 2025. Estimated avoided gas demand (over 260 billion cubic metres) equals almost half the global liquefied natural-gas market. These are substitution effects: energy that would otherwise have been delivered by coal, gas, and oil was delivered by renewables, nuclear, EVs, and heat pumps. The substitutions happened because the clean-technology stack, after six years of deployment, has reached the scale at which it can meet marginal demand growth faster than the fossil stack can expand to meet it.

What the report does not argue is the opposite conclusion. Oil, natural gas, and coal demand all grew in 2025. Coal grew at 0.4 percent, down from 1.4 percent in 2024; global coal demand reached roughly 30 million tonnes above the 2024 level. [5] The fossil stack has not contracted. The direction of change has slowed. The composition of 2025's demand growth — 60 percent from low-emissions sources, 40 percent from fossils and other — is the measurable inflection.

The 2025 solar-PV generation increase of 600 terawatt-hours is the single largest electricity-generation increase by any technology in a single year on record. [1] Renewables and nuclear combined added more electricity to global supply than total electricity supply grew. The growth in solar PV output was itself enough to cover more than a quarter of the year's total primary energy demand growth. The deployment trajectory is the rate at which the fossil-stack share of marginal generation is being replaced. At the current rate, the fossil stack's share of marginal growth falls each year; the legacy fossil stack's share of total generation falls more slowly, because the installed base is the lagging variable. The report is explicit about this difference.

Battery storage, electric vehicles, and heat pumps follow the same pattern. EV sales grew at their fastest rate since 2022. Battery storage installations kept pace with solar deployment. Heat-pump sales in advanced economies stabilised after a soft 2024. [1] The reason the 60-percent figure holds is that every clean-technology category moved in roughly the same direction at roughly the same time. This is not a single-category story; it is a portfolio story.

The US case sits inside the report as the test of the policy frame Zeldin delivered on Tuesday. US electricity demand rose 2.1 percent in 2025; the IEA projects 2 percent annual growth through 2030, with roughly half that increase driven by data-centre expansion. [5] Meeting that demand with the fuel mix the EPA is now walking away from regulating is the question the next four years produces a data answer for. The IEA's 60-percent figure is a description of what has already happened globally. Whether the United States's 2.1-percent demand growth is met disproportionately by gas-and-coal (as 2025's gas-to-coal switching began to suggest) or disproportionately by solar-and-wind (as the deployment pipeline, if fully built, would deliver) is the substantive test the IEA's 2027 report will measure.

For X's energy-policy discourse, which has been running the IEA numbers as a rebuttal since Sunday, the 60-percent figure is the single cleanest quantitative rebuttal of the endangerment-repeal framing available. [6] It does not argue the repeal; it measures the global trajectory the repeal walks away from. Zeldin's remarks and the IEA report are, in this sense, not disputing the same facts. They are looking at two different time-scale and geographic frames of the same energy system. The IEA is measuring 2025 globally. Zeldin is setting US policy for the next several years. The gap is not rhetorical. It is the measurable distance between a global direction and a domestic one, captured in the same document.

The paper's position on the IEA's Earth Day timing — the report landed on April 20, the Monday before the day itself — is that the timing was not accidental. The IEA has routinely released its Global Energy Review in mid-to-late March or early April, and for 2026 the window slipped to align with Earth Day. The 60-percent finding would have landed into a different editorial context had it been published in March. Publishing it on the Monday before Earth Day ensured that the data frame entered the editorial week in which the Zeldin speech was scheduled. The data and the speech ran on adjacent calendar days. For readers following both, the gap is the editorial product.

-- DARA OSEI, London