Every Country on Earth Joined the Ozone Treaty. Twenty-Four Have Not Joined Its Climate Amendment.

by Arjun Mehta · September 1, 2026

Science & Technology Writer, The Globe Ledger

There is exactly one environmental treaty that every country on the planet has joined. The Ozone Secretariat's ratification table, current as of 10 August 2026, records 198 parties to the Vienna Convention for the Protection of the Ozone Layer and 198 parties to the Montreal Protocol beneath it. No other multilateral environmental agreement has managed it.

The Protocol is also the rare case where the result is visible in the atmosphere rather than only in a projection. In a release dated 12 December 2025, the World Meteorological Organization reported that the 2025 Antarctic ozone hole reached a maximum extent of 18.71 million square kilometres between 7 September and 13 October, and closed on 1 December. NOAA and NASA called it the fifth smallest since 1992, the year the Montreal Protocol began to take effect, and recorded the hole breaking up nearly three weeks earlier than the average over the past decade. Paul Newman of NASA put the counterfactual plainly in the same release: "This year's hole would have been more than one million square miles larger if there was still as much chlorine in the stratosphere as there was 25 years ago."

The recovery dates the WMO cites are far off. On current projections the ozone layer returns to its 1980 condition around 2066 over the Antarctic, around 2045 over the Arctic, and around 2040 for the rest of the world — the shape of a chemical success story, in which emissions stopped decades ago and the stratosphere is still working through the inventory.

The amendment that changed what the treaty is for

What makes the Protocol interesting now is not the ozone hole but the fact that in 2016 the parties bolted onto an ozone treaty a control schedule for gases that do not deplete ozone at all. Hydrofluorocarbons were the industry's answer to the chemicals the Protocol banned: refrigerants and blowing agents, harmless to the ozone layer, and many of them extremely potent greenhouse gases. The Protocol had, in effect, engineered a climate problem while solving an ozone one. The Kigali Amendment, agreed at the Twenty-Eighth Meeting of the Parties in Rwanda, was the correction.

It entered into force on 1 January 2019, a date the Ozone Secretariat's own frequently-asked-questions note made conditional on ratification by at least 20 parties to the Protocol. The Secretariat's announcement that month put the running count at 65 countries. As of the Secretariat's 10 August 2026 table, 174 parties have ratified — a large majority, but 24 short of the treaty it amends, and short of the London, Copenhagen, Montreal and Beijing amendments, each of which stands at 197.

That gap is the most important fact about Kigali and the least reported. Universality is not automatic. A country that joined the Montreal Protocol in 1990 did not thereby agree to control HFCs; each amendment requires its own ratification, and a party that has not ratified Kigali carries no HFC obligation at all.

What the schedule actually requires

The word to hold onto is phase-down, not phase-out. Under the schedules in the Ozone Secretariat's frequently-asked-questions note on the amendment, developed countries — the non-Article 5 parties — work from a baseline of average HFC consumption over 2011 to 2013 plus 15 per cent of their hydrochlorofluorocarbon baseline. From there they cut to 90 per cent of baseline in 2019, 60 per cent in 2024, 30 per cent in 2029, 20 per cent in 2034, and settle at 15 per cent from 2036.

Developing countries were split into two groups on different clocks. The first group works from average consumption over 2020 to 2022 plus 65 per cent of its HCFC baseline, freezes in 2024, then steps down in 2029, 2035 and 2040 before settling at 20 per cent of baseline in 2045. The second group uses a 2024-to-2026 average plus the same 65 per cent, freezes in 2028, steps down in 2032, 2037 and 2042, and settles at 15 per cent in 2047.

Read the end states rather than the cuts and the design becomes clear. Nobody reaches zero. The schedules converge on a permanent floor of 15 to 20 per cent of baseline consumption — a policy judgement about the applications, medical and high-ambient-temperature among them, where substitutes are hard. It is a treaty that plans to leave a remainder, and says so.

The figure attached to Kigali in almost every summary is a fraction of a degree, and it is worth being precise about which fraction, because the published estimates differ. The Ozone Secretariat's announcement of entry into force in January 2019 said the amendment would cut projected HFC production and consumption by more than 80 per cent over 30 years and could "avoid up to 0.4°C of global warming this century". Its frequently-asked-questions note puts the same idea at up to 0.5°C by 2100. The 2022 quadrennial Scientific Assessment of Ozone Depletion, summarised by the WMO on 9 January 2023, gives a range: the amendment is "estimated to avoid 0.3–0.5°C of warming by 2100 (this does not include contributions from HFC-23 emissions)".

That parenthesis matters more than the range. HFC-23 is largely an industrial by-product rather than a product anyone sets out to make, and the assessment's headline estimate sets it aside. Every one of these numbers is also a difference between two projections rather than a measurement. Quoting any of them without the word "projected" converts a model output into an observation.

What actually polices it

The Protocol has no inspectorate. What it has instead is a global network of atmospheric monitoring stations, and in 2018 that network caught something: scientists reported an unexpected rise in global emissions of CFC-11, a chemical banned outright, beginning after 2012. The Ozone Secretariat's account reports that the Protocol's Technology and Economic Assessment Panel later estimated total undisclosed CFC-11 production at between 320,000 and 700,000 tonnes across 2007 to 2019, running at 40,000 to 70,000 tonnes a year over 2013 to 2018, with a large share of the emissions traced to eastern China.

The response was not sanctions. It was regulation, enforcement and monitoring: in the Secretariat's account, countries including China imposed additional regulatory controls and improved law enforcement against illegal production. Emissions rose until around 2018, then fell to levels substantially below those of 2008 to 2012, and the episode is now expected to cause only a small delay to ozone recovery.

That is the enforcement model in full. A violation on the scale of hundreds of thousands of tonnes was detected by instruments rather than inspectors, attributed by atmospheric science rather than by a tribunal, and closed by national regulation rather than by any penalty the treaty could impose. It worked — and it is not a mechanism anyone designed. It is a by-product of the fact that these molecules are rare, long-lived and measurable in parts per trillion.

The other half of the machinery is money. The Multilateral Fund pays the agreed incremental costs of compliance in developing countries, and is replenished by decision of the parties. Decision XXXV/1 set the 2024–2026 budget at 965 million US dollars, with contributions of 175.2 million dollars in each of the three years, 428,699,680 dollars drawn from contributions still due from the 2021–2023 triennium, and 10,700,320 dollars from interest. Compliance, in this treaty, is something the rich countries buy.

None of which transfers cleanly to the larger problem. The Montreal Protocol worked on a small number of industrially produced chemicals, made by a countable number of firms, for which substitutes existed and could be paid for. Carbon dioxide is a by-product of nearly every economic activity there is, with no drop-in replacement to subsidise. The Protocol's real lesson is narrower than the one usually drawn from it: a treaty can be extraordinarily effective when the thing it regulates is a manufactured product with a supply chain — and even then, nine years after adoption, its most climate-relevant amendment is still waiting on two dozen governments.