The Paris Agreement: How the World's Biggest Climate Promise Became Optional

On 12 December 2015, delegates from nearly 200 countries stood up in a conference hall on the edge of Paris and did something that had failed at every previous attempt for over two decades: they agreed, all of them, to a single deal on climate change. People cried. Diplomats hugged. The French foreign minister brought down a green gavel and the room erupted, because everyone in it understood what the alternative had looked like at every climate summit before this one, which was nothing at all. The Paris Agreement was real, it was legally binding, and it had the signature of basically the entire world on it. A decade on, that room's optimism looks like the high-water mark. Emissions are higher today than they were the afternoon that gavel came down.

What the deal actually said

Strip away the ceremony and the Paris Agreement's core promise was to keep global warming "well below" 2 degrees Celsius above pre-industrial levels, with an aspiration to hold it to 1.5 degrees if at all possible. The mechanism for getting there was where the whole thing quietly gave up its teeth. Instead of a global target handed down and enforced from above, which is roughly what had collapsed at Kyoto and Copenhagen before it, Paris let every country write its own target, called a Nationally Determined Contribution, and update it every five years with the expectation, not the requirement, that each new version would be more ambitious than the last. There is no penalty in the treaty for missing your own number. There is no court, no fine, no sanction. The entire enforcement model is transparency and peer pressure: publish what you promised, publish what you actually did, and let the gap embarrass you into doing better next time. It was, by design, the only version of a global climate deal that enough countries would actually sign. It was also, by the same design, a deal nobody could really be held to.

The number that kept going the wrong way

Global greenhouse gas emissions did not fall after 2015. They rose in nearly every year that followed, with a single sharp dip in 2020 caused by a global pandemic shutting down travel and industry, not by policy, and emissions bounced straight back up the year after. 2023 was confirmed as the hottest year on record. 2024 broke that record again within twelve months, and Europe's Copernicus climate service confirmed it as the first calendar year to average more than 1.5C above pre-industrial temperatures, the exact number the whole agreement had been built around avoiding. That single year doesn't technically break the treaty's own definition of the 1.5C goal, which is measured as a multi-decade average, not one calendar year. But it is very hard to read as anything other than a warning shot arriving early. The UN's own Emissions Gap Report, published every year since Paris, has been strikingly consistent in its verdict: even if every country fully delivers on every pledge currently on the table, the world is still on course for something like 2.6 to 2.9C of warming by 2100. That is real progress against the roughly 4C trajectory the planet was on before Paris existed. It is nowhere near what the agreement actually promised.

The money that arrived two years late, and still wasn't enough

Buried inside the Paris Agreement was a financial promise going back to the 2009 Copenhagen summit: wealthy nations would deliver $100 billion a year by 2020 to help poorer countries adapt to a warming world and build clean energy instead of coal plants. The target was missed for two straight years and only technically met in 2022, and even that figure was inflated by counting loans, not just grants, as "climate finance," which drew sharp criticism from the developing nations the money was meant to help. When COP29 in Baku tried to set the next, bigger goal in November 2024, the number that emerged was $300 billion a year by 2035. Developing nations had gone into that summit asking for $1.3 trillion. The gap between what was asked for and what was agreed became, for many delegates from the countries least responsible for historical emissions and most exposed to their effects, the clearest evidence yet of how little leverage the "please be more ambitious" model actually has.

In, then out, then in, then out again

Nothing captures the agreement's fundamental fragility better than watching a single country walk through its front door twice and out the back door twice in under a decade. The United States announced its first withdrawal in June 2017, which took effect on 4 November 2020, the earliest date the treaty's own rules allowed. President Biden rejoined on his first day in office, 20 January 2021. President Trump signed a second withdrawal on his first day back in office, 20 January 2025. Each flip cost the agreement momentum, credibility and, for the periods the US was out, the participation of the country responsible for the largest share of cumulative historical emissions on the planet. A treaty that depends entirely on voluntary goodwill was always going to be exposed to exactly this kind of whiplash, and it got it, twice, from a single signatory.

Saying "fossil fuels" out loud took eight years

Here is a fact that says more about how carefully worded these summits are than almost anything else: it took until COP28, held in Dubai in December 2023, for a final COP text to explicitly call for "transitioning away from fossil fuels" at all. Not eliminating them. Not even phasing them out on a set timeline. Just naming them, in an agreement about climate change, eight years after Paris. It was reported at the time as historic, and in the narrow sense that the words had never appeared before, it was. It also carried no enforcement mechanism, no date, and no binding requirement on any single country, which is the same design limitation that runs through the entire agreement.

So did it fail

The honest answer sits somewhere less satisfying than a clean yes or no. The Paris Agreement did not deliver the world it promised. Judged purely against its own headline target, 1.5 to 2 degrees, it is failing, and the data for 2023 and 2024 makes that harder to argue around every year. But the counterfactual matters too: the pre-Paris trajectory the world was actually on pointed toward something close to 4 degrees of warming by 2100, a scenario serious enough to threaten food systems and coastlines on a scale the current, still-too-hot 2.6 to 2.9 degree path does not. A voluntary agreement with no enforcement mechanism bent the curve. It did not bend it nearly far enough, and it left itself open to being gutted every time a single powerful country's domestic politics changed. Ten years after that green gavel came down in Paris, the deal's defenders and its critics are, in a strange way, both right.

Sources

• United Nations, "The Paris Agreement" (the treaty's text, the 1.5C/2C goal, and the Nationally Determined Contribution mechanism).
• United Nations Environment Programme, "Emissions Gap Report 2024" (the 2.6 to 2.9C projected warming under current pledges).
• Copernicus Climate Change Service, "2024 is confirmed to be the warmest year on record" (2024 as the first calendar year above 1.5C).
• Reuters, "Trump signs order to pull US out of Paris climate agreement, again" (the 20 January 2025 withdrawal).
• White House Archives, "Statement by President Trump on the Paris Climate Accord" (the original 1 June 2017 withdrawal announcement).
• UNFCCC, "COP28 Agreement Signals 'Beginning of the End' of the Fossil Fuel Era" (the December 2023 fossil fuel language).
• Reuters, "COP29 agrees $300 billion climate finance deal" (the Baku finance goal and developing-nation criticism).
• OECD, "Climate finance provided and mobilised by developed countries in 2013-2022" (the delayed delivery of the $100 billion pledge).