From Rio to the Paris Climate Agreement

The convention set an objective and no obligations, because obligations would not have been accepted. States signed a framework precisely because it required nothing of them, and every negotiation since has been about converting the framework into commitments.

Binding targets were accepted only by states willing to bear them, and the largest emitter signed and never ratified. A treaty that obliges some parties and not others cannot address a problem to which every party contributes.

Negotiations were relaunched because Kyoto's targets expired in 2012 and covered a shrinking share of emissions. The process continued because the problem continued, not because any state had changed its assessment of what it would accept.

A small group of leaders including the American and Chinese presidents negotiated a text in a side room and presented it to a conference that refused to adopt it. Great-power bargaining produced the only agreement available and the multilateral process rejected it.

The mandate committed parties to negotiate an instrument applicable to all, which quietly abandoned the firewall between developed and developing obligations. That concession by the developing world was the negotiation's real content and took a fortnight of argument to obtain.

Two states responsible for around forty per cent of emissions announced targets together, which removed the argument each had used to justify inaction. Bilateral agreement between the two that mattered most did what twenty years of multilateral process had not.

The agreement contains no binding emissions targets and no enforcement, which is why one hundred and ninety-five states accepted it. Its architecture is a frank acknowledgment of what states will tolerate, and it was designed around that constraint deliberately.

Entry into force took eleven months, faster than anyone expected, because ratification cost nothing. States join readily where the obligations are ones they have written themselves, so the speed of entry into force measures the agreement's weakness at least as accurately as it measures its appeal.

Emissions reduction is a collective good with no enforcement, so each state gains by others acting and loses by acting alone. That structure is the whole problem, and thirty years of negotiation have been attempts to work around it rather than solve it.

The American Senate had voted ninety-five to nothing against any agreement exempting developing countries, before the protocol was negotiated. Negotiators spent years working toward an outcome their own legislature had unanimously ruled out in advance, which is among the clearest illustrations available of how domestic constraints determine what international agreement is possible.

By 2007 China's emissions exceeded America's, so any agreement excluding developing countries was pointless and any agreement including them was unacceptable to those countries. The deadlock was structural and took eight more years to break.

The two largest emitters could not be compelled and would not accept binding limits, so nothing binding was possible. Copenhagen demonstrated the limit of what a consensus process can extract from states that do not wish to be bound.

Applicable to all meant that China and India would take on commitments, which they accepted in principle because the form remained undefined. Accepting a category while deferring its content is how negotiations of this kind proceed.

Neither state accepted an obligation to the other; they announced parallel intentions, which is a form of coordination that avoids commitment. Great powers coordinate where cooperation is impossible, and the distinction is what made this achievable.

Nationally determined contributions let every state set its own target, converting the problem from bargaining over shares to a series of unilateral declarations. The design avoids the distributional fight that had blocked agreement for twenty-three years.

The United States joined by executive action and withdrew four years later by the same means, then rejoined and withdrew again. An agreement designed around one state's domestic politics is hostage to them, which is the cost of the form chosen.

Common but differentiated responsibilities recognised that the states that had emitted historically were not those that would emit most in future. Building equity into the architecture made agreement possible and made every subsequent negotiation about where the line falls.

Ratification requires domestic legislatures, and climate policy imposes concentrated costs on identifiable industries for diffuse and future benefits. That asymmetry is why the politics is so difficult everywhere and why targets are so often missed.

Public concern peaked around this period, driven by scientific reports and by unusually effective communication, and governments responded to it. Opinion sets the outer limit of what is politically possible, and this was the period when that limit was widest.

Expectations had been raised enormously, with heads of government attending and vast public attention, and the failure was correspondingly damaging. Building anticipation before securing the substance is a recurring error in summit diplomacy, and it converted a poor outcome into a public disaster that set the negotiations back by several years.

The European Union led the coalition that produced the mandate, allied with small island states and least developed countries against the largest emitters. A coalition of the willing and the vulnerable is how the process has usually been moved.

China's targets were driven substantially by domestic air pollution, which had become a serious source of urban discontent. Domestic political necessity produced an international contribution, which is how most climate policy actually originates. Governments act on emissions when their own populations are demanding something else that happens to require the same measures.

The structure was chosen partly so that American participation would not require Senate ratification, which would not have been forthcoming. Domestic ratification constraints in a single state shaped the legal form of an agreement binding a hundred and ninety-five, which tells you a good deal about where the real veto in international law sits.

The five-yearly review cycle creates a recurring domestic moment at which governments must produce a stronger target and defend it. Whether that accountability proves sufficient, with no penalty available beyond the embarrassment of arriving with nothing new, is the experiment the agreement represents, and it is being run in real time on the only atmosphere available.

A secretariat, an annual conference, inventories and reporting requirements created a machine that would meet every year regardless of progress. Establishing a process before agreeing substance is how long negotiations are sustained, and this one has run for three decades.

Emissions trading, joint implementation and the clean development mechanism introduced market instruments into environmental treaty-making. Those mechanisms worked unevenly, and the clean development mechanism in particular was criticised for financing projects that would have happened anyway. They nonetheless established the template that carbon markets around the world still use.

The plan created two negotiating tracks running in parallel, one under the convention and one under the protocol, which produced years of procedural complexity. Architecture chosen to accommodate a disagreement tends to institutionalise it instead, and the two tracks then had to be merged again at Durban four years later before any progress became possible.

Consensus rules meant a handful of objecting states could block adoption, so the conference took note of the accord instead. Procedural workarounds of that kind preserve a text without giving it status, and the accord's targets were nonetheless submitted.

Setting a deadline of 2015 and an implementation date of 2020 gave the negotiation a schedule and a focal point. Deadlines concentrate multilateral negotiation more reliably than any substantive argument, because they convert an open-ended process into one where delay has a cost, and everyone in the room can count the remaining conferences.

The announcement gave every other negotiator a floor to build on and removed the standard excuse for delay. Establishing a focal point a year before a negotiation is worth more than concessions made during it, because everyone else's position is then constructed in relation to a number that already exists.

What is binding is procedural: submitting a contribution, reporting under a common transparency framework, and returning every five years with a stronger one. The ratchet is the mechanism, and whether it works is the only question that matters.

Transparency, common reporting and a global stocktake substitute peer pressure and information for enforcement. It is the most ambitious test yet of whether a regime can work through reputation alone, and the evidence so far is mixed.

The atmosphere is not any state's territory, so managing it required a form of collective authority the society of states had rarely attempted. Extending governance to a global commons is among the harder things the society has tried.

Differentiating obligations by development status wrote a formal inequality into the treaty, which the states bearing the obligations increasingly resented. What fairness requires between historic and current emitters has no agreed answer and blocks progress repeatedly.

Mitigation, adaptation, finance and technology were established as the four pillars, which acknowledged that poorer states needed support rather than only obligations. Broadening the agenda beyond emissions was the price of keeping the developing world engaged at all, and finance in particular has been the issue on which every conference since has come closest to breaking down.

Small island states objected that an agreement negotiated among the powerful and presented for acceptance denied them any role. The tension between effectiveness and inclusion runs through every multilateral negotiation and was unusually visible here.

Durban preserved the convention's principle of differentiation while committing everyone to act, holding two positions the society could not reconcile. Living with an unresolved principle is often the condition of continuing to negotiate at all.

Two great powers settling the essentials between themselves and presenting them to the wider process is the concert method applied to the atmosphere. International society still works this way whenever it works quickly, and the tension between that method and the formal equality of a hundred and ninety-five parties is never resolved, only managed by presenting the result as a contribution rather than a decision.

Nearly universal participation was achieved by abandoning obligation, which is a trade international society makes often. Whether broad membership with weak duties is preferable to narrow membership with strong ones is the permanent argument in this field, and Kyoto and Paris are now the two standing exhibits on either side of it.

Nearly every state accepted a common framework for a shared problem, which is a real achievement of international society whatever the emissions data show. Universality was obtained and the substance was deferred to the members' own consciences.

Climate was established as a legitimate subject of international negotiation, which was not obvious in 1992 and had taken scientists twenty years to achieve. Getting an issue onto the agenda is a distinct achievement from resolving it.

Kyoto established that binding international targets were the model for climate policy, and its failure discredited that model. Paris succeeded eighteen years later partly by abandoning the approach that Kyoto had made seem obligatory, which suggests that the most useful thing a failed treaty does is demonstrate which design cannot work.

The plan's language distinguished commitments for developed countries from actions for developing ones, a formulation negotiated word by word. In this field the terminology is the substance, because it determines what each party has accepted.

Copenhagen was understood as a failure so complete that it forced a rethinking of the whole approach. The reassessment it forced, away from binding targets negotiated collectively and toward pledges each state sets for itself, is what made Paris achievable six years later. Failure did the work that success would not have.

The phrase legal instrument or agreed outcome with legal force was drafted in the final hours to allow both readings. Deliberate ambiguity about legal character kept the process alive and postponed the argument to Paris.

The announcement changed what participants believed was possible in Paris a year later, and expectations of success are partly self-fulfilling in negotiations. Shifting the expectation was arguably the announcement's largest effect, since negotiators who believe agreement is coming behave differently from those preparing to explain a failure, and that difference was visible throughout the following year.

The 1.5 degree reference, inserted at the insistence of small island states, became the target everything is now measured against. A number negotiated as an aspiration acquired an authority that has structured the entire subsequent debate.

Paris established net zero as the organising objective of climate policy worldwide, adopted since by states, cities and companies that were not parties. The agreement's largest effect has been on what everyone understands the goal to be.