Climate Negotiations as a System: Why COP Outcomes Move in Increments

Last updated: 31 October 2026. COP31 runs 9 to 20 November 2026; check official sources for the latest agenda.

Climate negotiations follow a pattern that frustrates observers and puzzles newcomers: outcomes arrive late, in compromise language, and only slightly beyond the previous round. Treated as a system rather than a series of failures of will, the pattern has a structure. The decision rule is consensus, nearly 200 parties each hold a practical veto, and package deals link unrelated issues. Those features produce text that moves in increments and rarely moves backward. This article describes that structure with the facts of the most recent summits, shows a small numerical model of why the consensus rule sets the ceiling, and lists what to watch at COP31 in Antalya.

The structure of the system

The annual UN climate conference (the Conference of the Parties, or COP) is the decision-making body of the UN climate convention. Three structural features explain most of its behaviour.

1. Consensus decision-making. There is no voting on substance. A text is adopted when no party maintains a formal objection. This gives every party, from small island states to large fossil-fuel producers, a practical veto.
2. Coalitions. Parties negotiate in blocs, such as the group of small island states, the EU, the group of least developed countries and producer groupings. Blocs reduce the number of positions but add their own internal bargaining.
3. Long agendas with linkage. Mitigation, adaptation, finance, loss and damage and transparency are negotiated at once, so concessions in one can be traded for another.

This is the kind of polycentric, many-actor system described in polycentric governance, with the added constraint of a unanimity rule. It is also a common-pool problem: the atmosphere is a shared resource, and each party’s incentive to restrain its own use depends on what others do, a structure analysed in common-pool resource governance and in the older tragedy of the commons.

Why consensus sets the ceiling: a small model

Imagine a negotiated text with an ambition level on a scale from 0 to 10. Each party has a reservation point: the highest ambition it will accept without blocking. Under consensus, a text can be adopted only if its ambition is at or below every party’s reservation point. The achievable ambition is therefore the minimum across parties.

Take five illustrative blocs with reservation points of 9, 8, 6, 4 and 2. The achievable text sits at 2, even though a majority would accept 6 or more. Now allow linkage. A finance commitment that two of the reluctant blocs value raises the fourth bloc’s reservation point from 4 to 6 and the fifth’s from 2 to 4. The new minimum is 4. The text doubled in ambition without any bloc changing its underlying view, because the package changed what each bloc was weighing.

Scenario	Reservation points of five blocs	Achievable ambition
Single issue	9, 8, 6, 4, 2	2
With a finance package	9, 8, 6, 6, 4	4

The numbers are invented to show the mechanism, not to describe any country. The model makes three points that match how negotiators behave. First, the leverage of holdouts is high, so presidencies spend their effort on the least willing parties. Second, linking issues is not a distraction but the main tool for moving the ceiling. Third, because the minimum is what counts, adding more ambitious parties to the coalition changes little, while moving one blocker matters a great deal.

The ratchet: what recent summits show

Agreed language rarely disappears. Later decisions cite earlier ones, which turns each into a floor for the next round. The sequence of headline texts illustrates this.

• Paris, 2015. The Paris Agreement set the aim of holding warming well below 2°C while pursuing 1.5°C, and created a five-year cycle in which countries submit updated national plans (nationally determined contributions).
• Glasgow, 2021. The decision text called for a phase-down of unabated coal power and the phase-out of inefficient fossil-fuel subsidies, the first explicit mention of coal in a COP outcome, in softened wording after late changes.
• Dubai (COP28), 2023. The global stocktake decision called for transitioning away from fossil fuels in energy systems, in what is often called the UAE Consensus.
• Belem (COP30), 2025. The main outcome, called the Mutirao text, combined several contentious negotiating tracks into a single consensus decision and called for tripling adaptation finance by 2035. Reporting by UN News and the Stockholm Environment Institute notes that the final text did not clearly commit countries to phasing out fossil fuels, and referred back to the UAE Consensus rather than adding to it. Stronger draft language was removed in the final hours, and the COP30 presidency said it would pursue a roadmap for transitioning away from fossil fuels separately.

Each step is modest on its own, and critics point out that adaptation finance pledges have a record of falling short: a pledge to double adaptation finance made in Glasgow remains unmet, according to SEI. Defenders of the process answer that consensus text carries legitimacy with every party, which makes implementation more likely than with a text imposed on dissenters. Both observations hold. The pattern is that a rule giving everyone a veto produces slow, durable, widely owned language.

Figure 1. Headline climate texts as a staircase. Each step cites the earlier ones, so language holds even when the pace is slow. The Belem entry notes that the fossil-fuel roadmap was kept outside the consensus text.

The feedback loops behind the increments

A reinforcing loop of ambition

Falling costs of renewables and batteries raise what governments can accept, which expands deployment, which lowers costs further. When this loop is strong, reservation points rise on their own, and negotiators codify what is already happening. This is why mitigation language tends to follow reality rather than lead it.

A balancing loop of resistance

Parties whose economies depend on fossil-fuel production or use face concentrated near-term costs and diffuse long-term benefits. Their resistance increases when proposals threaten revenue, a response described in policy resistance. Where infrastructure has long lives, the resistance reflects real carbon lock-in, not only bargaining posture.

A delay between decisions and effects

The five-year plan cycle, multi-year negotiating processes and decades-long asset lives create delays much longer than election cycles. Feedback arrives late, which makes course correction hard, the general problem examined in delays in systems thinking.

COP31: the structure of this year’s round

COP31 is scheduled for 9 to 20 November 2026 at the Antalya Expo Center in Antalya, Turkiye. The arrangement is unusual: Turkiye hosts and its environment minister, Murat Kurum, serves as COP President, while Australia’s climate and energy minister, Chris Bowen, serves as President of Negotiations. The split resulted from a compromise reached at COP30 after both countries had bid to host. Under the UN’s rules a host must be agreed by consensus, and without agreement the conference would have defaulted to Bonn, where the convention’s secretariat sits.

The split presidency is itself a case study in the consensus system: two competing claims were resolved by dividing the role rather than by one side winning. Its effect on the negotiations remains to be seen. The leaders’ summit is scheduled for days 3 and 4, 11 to 12 November, a change from earlier COPs that held summits at the start or beforehand.

What to watch

1. Bundling. Which issues are packaged together, and which are split off into separate processes. The COP30 pattern was both: a single consensus decision plus items taken outside it.
2. Adaptation finance. Whether the tripling goal gets a delivery pathway and reporting, since earlier targets lacked them.
3. The fossil-fuel transition. Whether the separate roadmap process the COP30 presidency announced produces anything that enters the formal texts.
4. Procedure. Whether talks run into overtime, as at COP30, where negotiators worked past midnight, and which parties announce reservations.
5. Observer access and civil-society participation. Concerns about this have been raised by NGOs, and the decisions on it shape who can inform the negotiation.

These are indicators to read, not predictions. The outcome depends on bargaining that has not happened yet.

Reading outcomes without hype or despair

Because the consensus rule sets a ceiling, the headline text is a poor measure of what is happening. Three better questions:

• Is the language new, or a restatement? New wording that all parties accept is a real step. Restating earlier wording keeps the floor but adds nothing.
• Is there an implementation mechanism? Targets with finance, reporting and deadlines behave differently from targets without.
• What happens outside the text? Coalitions of the willing, national policies and market changes can move faster than the consensus text, and they feed the first loop above.

This reading is consistent with how adaptive governance is framed in adaptive policy design: build in review points so that a slow decision rule is complemented by faster experiments elsewhere.

Frequently asked questions

Why do climate negotiations take so long?

Because decisions need consensus among nearly 200 parties, many issues are linked, and delays between decisions and effects reduce the pressure to resolve quickly. Overtime sessions are common as deadlines force bargaining.

Could voting replace consensus?

It could change what gets adopted, but supporters of consensus argue that texts adopted over objections have weaker legitimacy and compliance. Critics argue the rule lets a few parties dilute outcomes. The convention’s own rules on voting have long been unsettled, which is why consensus remains the practice.

What does the split presidency at COP31 mean?

Turkiye hosts and presides over the conference, while Australia’s minister leads the negotiations. It resulted from a compromise at COP30 when both had sought to host. The practical effects on outcomes are not yet known.

Does a weak text mean nothing happened?

Not necessarily. The text sets a floor and the real changes often come from national policy and market shifts, which the text may follow. But a weak text with no mechanisms does leave a gap between stated goals and delivery.

Conclusion

Climate negotiations move in increments because the decision rule makes the least willing party the ceiling, because package deals shift that ceiling slowly, and because agreed wording ratchets. That structure explains both the slow pace and the stickiness. For anyone following COP31, the useful habit is to look at structure first: what is bundled, who presides, which items sit outside the text, and whether there is a mechanism to deliver. Headlines about success or failure tell you less than those details.

See the glossary for climate negotiations, polycentric governance and common-pool resources.

Sources and further reading

• 2026 United Nations Climate Change Conference (Wikipedia), for COP31 dates, venue and presidency arrangement. Confirm with the official UNFCCC site.
• United Nations News, report on the COP30 outcome, November 2025.
• Stockholm Environment Institute, “COP30: collective effort, constrained ambition”.
• UNFCCC, The Paris Agreement.