Equitable cooperation deepens the solution space for high ambition pathways

Fairness is assessed after the pathway is fixed

Global mitigation scenarios that inform climate ambition typically treat fairness and cooperation as external to the modelling process. Fairness enters afterwards, when analysts compare each region's modelled effort with a principle-based allocation (an 'ex-post' assessment). Least-cost pathways minimise the global cost of mitigation, so mitigation occurs wherever it is cheapest, and this distributional judgement usually stays implicit. The result is a perceived tension between fair effort-sharing and the urgent collective action needed to limit warming.

We ask what changes when fairness is a design constraint during scenario generation. Each region must meet a principle-based 'fair share' of the carbon budget and may cooperate with other regions to do so. We then examine how the pathway to the same climate outcome changes, both when finance for that cooperation is unlimited and when it is held to a minimum.

A framework that starts from any existing scenario

The framework starts from any existing mitigation scenario, which we call the 'source' scenario (here, a globally cost-effective pathway). We split its emissions into two sets. 'Covered' emissions (Kyoto gases excluding land use, international shipping and aviation) are allocated to the twelve model regions by a fairness principle. 'Excluded' emissions follow the source scenario carbon price, which prevents leakage between sectors. This boundary is itself a strong normative choice that shapes regional obligations.

A region whose cumulative emissions exceed its allocation carries a carbon debt. It closes that debt through some combination of deeper domestic gross reductions, domestic carbon removal, and cooperation with other regions. We model cooperation as trade in covered allocations inside the model, so the price of transferred allocation emerges from the model rather than being imposed on it.

Two 'transfer corners' bound the resulting space of pathways. Under 'unlimited transfers', regions may finance mitigation elsewhere instead of cutting further at home. The physical pathway then reproduces the source scenario, and the corner prices the fair-share obligation. This corner reproduces the classical equity-efficiency result and is the analogue of existing ex-post quantifications of fair finance. The 'lowest-feasible transfers' corner is the paper's addition. Here finance between regions is held to the minimum at which the model still solves, starting from the approximately 100 billion USD per year already committed. Meeting regional fair shares then requires shifts in the energy system itself.

We explore this space across six allocation approaches, which are the rules that divide the budget between regions. Equal cumulative per capita (ECPC) gives each person an equal share of cumulative emissions, and capability-adjusted cumulative per capita (CAPC) adjusts that share for regional capability. Each counts responsibility from 1990, 2015 or 2025. We also test cooperation restricted to carbon removal, a ten-year delay in cooperation, and two carbon budgets.

Both budgets run from 2020 and allow overshoot. The 800 Gt CO2 budget is consistent with limiting warming to approximately 2 °C with 67% likelihood, and the 500 Gt CO2 budget with 1.5 °C at 50% likelihood. All scenarios follow the SSP2 'Middle of the Road' narrative, with an SSP1 comparison in the supplementary information.

The climate outcome holds while the fossil exit accelerates

Both transfer corners reproduce the end-of-century net cumulative CO2 of the source scenario, indicating that fairness and ambition are not inherently in tension. Global renewables deployment and electrification are largely indistinguishable across variants, which differ mainly in the pace of the fossil exit. Under lowest-feasible transfers, higher-responsibility regions cut gross emissions faster at home. Global fossil primary energy in 2040 then falls 3–21% below the unlimited-transfers variant, depending on the allocation approach. Unlimited transfers imply cumulative obligations of 10.1–44.8 trillion USD (NPV 2026–2100, 5% discount rate), and lowest-feasible transfers cut these to less than half in each case.

The aggregate cost is small and progressively distributed, with the important caveat that we externalise climate damages. The cost-effective 2 °C source scenario reduces discounted global consumption by approximately 0.8% relative to the no-new-policy reference scenario. Unlimited transfers leave this unchanged, because they redistribute without altering the physical system. Lowest-feasible transfers raise it to at most 1.3%, and lower-responsibility regions gain relative to the cost-effective source scenario.

Model feasibility limits how far finance and domestic effort can substitute for each other. Extending historical responsibility further back therefore eventually scales financial flows rather than the physical transition. Delaying cooperation by ten years roughly triples near-term energy investment under lowest-feasible transfers, from 0.4% to approximately 1.2% above the source scenario (2026–2050). Under the 1.5 °C budget the solution space persists, though with less room to shift energy systems and with differentiation settled increasingly through finance.

Why restricting cooperation to carbon removal raises the cost

International cooperation may in practice run through specific instruments. We therefore also restrict transfers to financing novel carbon dioxide removal: bioenergy with carbon capture and storage (BECCS) and direct air capture with carbon storage (DACCS). Under this restriction the fair-share variants diverge from the source scenario even when finance is unlimited. Transfers barely change between the corners (27.4 versus 27.0 trillion USD NPV), because technical and geological limits, not cost alone, govern how much removal can be transferred. Per tonne of mitigation transferred, removal-only cooperation costs approximately 220 USD, roughly an order of magnitude above the 20–30 USD when all mitigation options are in scope. The global consumption loss relative to the source scenario rises from marginal to just over 0.6%.

Notably, the response remains one of gross emissions reductions first. Roughly two-thirds of the higher-responsibility overdraft is still met through deeper domestic gross reductions, and domestic novel removal covers a further tenth. Global geological storage reaches its 6 Gt CO2 per year injection cap around mid-century, ahead of the source scenario. Regional responses are heterogeneous. Sub-Saharan Africa expands direct air capture at the expense of industrial carbon capture, using its geological storage as an export resource. South Asia, in contrast, develops no additional removal-export capacity under its modelled constraints. These responses diverge from the simple reallocation of removal that fair-share studies outside model frameworks commonly assume.

Fairness as a feature of feasibility

These dynamics emerge only when scenario generation itself accounts for differentiated responsibilities and capabilities, and ex-post assessments of cost-effective scenarios miss them. Under constrained finance, a 'gross emissions reductions first' strategy among higher-responsibility regions releases pressure on lower-responsibility regions. The window for regional differentiation also stays open in 1.5 °C scenarios, where the remaining carbon budget is effectively exhausted before the first model timestep.

The results quantify the aggregate potential of interregional cooperation, both market and non-market. They should not be misconstrued as prescribing carbon market designs or endorsing Article 6 mechanisms to compensate for shortfalls in nationally determined contributions (NDCs). Every variant presumes functional multilateral governance, though this caveat applies equally to any global pathway that returns to 1.5 °C. The allocation approaches are a subset of many valid interpretations of 'common but differentiated responsibilities and respective capabilities'. We also externalise climate damages, which fall disproportionately on the lower-responsibility regions that gain under cooperation. With these caveats, we argue that scenario assessments should routinely explore fair-share variants, lowest-feasible transfers in particular, alongside cost-effective pathways. This matters as the next NDC cycle and the second Global Stocktake approach.