The COP29 climate summit in Baku ended this weekend with developed countries agreeing to take the lead in mobilising at least $300 billion per year for developing countries by 2035, replacing the old $100 billion annual climate-finance target. The agreement also sets a broader aspiration of scaling climate finance from public and private sources to $1.3 trillion annually by 2035. On paper, tripling the previous target sounds substantial. Many developing countries responded with anger. India called the deal inadequate, while representatives of vulnerable states argued that the financing remains far below what is required to decarbonise growing economies while adapting to climate impacts they contributed relatively little to creating. The disagreement therefore is not really about whether $300 billion is a large number. It is about who should pay for a global externality when responsibility for creating it, capacity to finance it and vulnerability to its consequences are distributed radically differently across countries.
Climate change is almost a textbook negative externality. Burning fossil fuels creates private benefits through electricity, transport and industrial production while greenhouse gases impose costs across borders and generations. The atmosphere does not distinguish between a tonne of carbon emitted in Manchester, Mumbai or Shanghai. That makes climate policy unusually difficult because governments bear the political cost of reducing emissions domestically while receiving only a fraction of the global benefit.
International cooperation therefore faces a free-rider problem. Every country benefits if others decarbonise. Each has incentives to let others bear more of the cost.
Historical inequality makes that problem much harder. Britain, the United States and other industrialised economies accumulated wealth during more than a century of carbon-intensive development. Many poorer countries are now being asked to reduce emissions before reaching comparable income levels. From India’s perspective, a citizen consuming a fraction of the electricity used by an average American cannot reasonably be assigned an identical mitigation obligation simply because both now inhabit the same climate-constrained world.
This is the logic behind the UN climate principle of “common but differentiated responsibilities and respective capabilities.” Climate change is a common problem, but countries did not contribute equally to creating it and do not possess equal resources for solving it.
Yet historical responsibility cannot be the only variable either. China is now the world’s largest annual emitter, while India is among the largest in aggregate terms even though its per-capita emissions remain much lower than those of the United States. A climate framework based indefinitely only on which countries were wealthy in 1992 increasingly fails to describe the actual distribution of current emissions.
This creates the central climate-finance dispute: should responsibility be determined by historical emissions, current emissions, per-capita emissions, national income or some combination of all four?
Each metric produces different winners.
Historical emissions place greater responsibility on the West.
Current emissions increase China’s burden.
Per-capita measures strengthen India’s argument for additional development space.
Income measures place more responsibility on wealthy Gulf states and advanced economies.
There is no scientifically objective weighting because the final decision is distributive rather than climatic.
The $300 billion agreement therefore is ultimately a political allocation rule.
Developing countries argue that their actual financing needs are much larger. Independent expert estimates associated with the climate negotiations have suggested that emerging and developing economies outside China may require around $1 trillion annually in external climate finance by the mid-2030s, alongside much larger domestic investment. The Baku agreement’s broader $1.3 trillion objective reflects recognition of that scale, but the guaranteed political commitment is much smaller.
This is where the distinction between finance and fiscal transfer becomes important. A trillion dollars of climate investment does not mean rich-country governments need to write a trillion-dollar cheque annually. Much of the energy transition can be financed through private capital if projects generate returns. Solar farms sell electricity. Grid investments generate revenue. Electric transport can attract commercial financing.
Public finance is most valuable when it makes private investment possible.
Suppose a renewable project in India is commercially viable except investors demand a high interest rate because of currency and political risk. A development bank can provide guarantees or concessional financing, lowering the project’s cost of capital. One public dollar can therefore mobilise several private dollars.
That is financial leverage.
Climate negotiations become less useful when they focus entirely on the headline quantity of government transfers rather than the amount of actual low-carbon infrastructure each public dollar produces.
But developing countries have a legitimate concern about counting ordinary commercial loans as climate assistance. A loan needs repayment. If a vulnerable country borrows heavily to rebuild infrastructure destroyed by climate-related disasters, it can emerge more climate-resilient but more indebted.
This creates the climate-debt trap.
Countries that contributed relatively little to historical warming can face hurricanes, floods or droughts, borrow money to recover, then devote increasing portions of future government revenue to debt service.
Climate vulnerability becomes fiscal vulnerability.
This is why grants matter particularly for adaptation and loss and damage. A solar plant can generate revenue and therefore support a loan. A seawall protecting a poor coastal community may generate enormous social benefits without producing a direct cash flow. Private investors cannot monetise avoided deaths or prevented displacement easily.
Mitigation and adaptation therefore require different financing architectures.
Markets are relatively good at financing projects with predictable revenue.
States and multilateral institutions are needed where benefits are public goods.
The political debate also exposes the enormous difference between the cost of climate action and the cost of climate inaction. Hundreds of billions annually sounds extraordinary when treated as a budget line. It looks different when compared with the expected destruction from extreme weather, agricultural losses, sea-level rise, health effects and forced migration over decades.
Climate investment is therefore partly insurance.
The difficulty is the same one governments face with military deterrence or pandemic preparation: successful prevention produces an invisible counterfactual.
A government spends billions strengthening flood defences.
The flood does not destroy the city.
Citizens observe the expenditure.
They cannot observe the disaster that did not happen.
Political systems systematically underinvest in prevention because the benefit is difficult to attribute.
This is one reason climate finance remains far below estimated need despite decades of negotiations.
There is also a profound trust problem. Developed countries originally promised in 2009 to mobilise $100 billion annually by 2020 and only reached that level years late. Developing governments therefore reasonably discount new promises. A commitment to $300 billion in 2035 has less value if recipients believe future governments can reinterpret accounting rules or delay delivery.
Finance commitments therefore possess a credibility premium.
A smaller commitment delivered predictably can be more useful for investment planning than a larger number repeatedly missed.
This is why climate finance needs stronger measurement. How much is grants? How much concessional lending? How much ordinary market-rate finance? How much represents money that would have been provided as development assistance anyway?
Without common accounting, governments can increase reported climate finance partly through reclassification rather than additional resources.
Additionality matters again.
If Britain relabels an existing development project “climate finance”, the reported total rises while the recipient receives no additional fiscal capacity.
The same scrutiny should apply to private finance claims. If a profitable solar investment would have occurred without government intervention, counting the entire project as “mobilised” by climate policy exaggerates policy impact.
The relevant metric should be counterfactual investment created.
How much additional clean infrastructure exists because the financing mechanism existed?
That is harder to measure than headline commitments but economically more meaningful.
India’s anger at COP29 therefore has substance even though New Delhi also has incentives to maximise external financing. India needs enormous investment in electricity generation, transmission, storage, transport and industrial decarbonisation while simultaneously expanding energy consumption for a population whose average income remains far below Western levels.
The country cannot simply stop consuming more energy.
Development requires more electricity.
The objective is making the marginal unit of development progressively less carbon-intensive.
This distinguishes India from mature economies where total energy demand can plausibly stabilise or decline. India is building cities, factories, railways and housing that do not yet exist. Climate policy therefore determines the technological composition of new capital rather than merely replacing old capital.
That creates an enormous opportunity.
Building clean infrastructure initially can be cheaper than constructing fossil-intensive systems and replacing them twenty years later.
But upfront capital requirements are high.
This is why the cost of capital matters so much for developing-country decarbonisation. Renewable technologies such as solar have relatively low operating costs but require significant initial investment. A project financed at 12% interest can be far more expensive than the same physical project financed at 4%.
Rich countries therefore possess a structural climate advantage simply because capital is cheaper.
Climate finance can correct part of this inequality by reducing financing costs rather than paying for every solar panel directly.
This is perhaps the most productive direction for international policy: guarantees, multilateral development-bank reform, foreign-exchange risk instruments and concessional lending capable of turning projects currently considered too risky into investable assets.
The World Bank and other multilateral development banks could therefore matter more than annual COP bargaining suggests. Their balance sheets allow relatively small amounts of government capital to support much larger lending volumes.
Reforming them could create climate finance through institutional leverage rather than relying entirely on politically difficult annual budget transfers from rich countries.
That approach also makes climate policy easier to defend domestically in donor states. Taxpayers may resist transferring hundreds of billions abroad permanently. They may be more willing to capitalise institutions that provide loans, guarantees and investments producing measurable infrastructure.
But concessional finance cannot replace grants completely.
The poorest and most climate-vulnerable countries simply cannot borrow their way out of every disaster.
That is where climate justice becomes economically concrete rather than rhetorical.
A small island state can contribute almost nothing to global warming and still face existential costs from sea-level rise.
There is no plausible polluter-pays principle under which that state should finance the entire adaptation burden itself.
Some international transfer is therefore justified if countries accept that greenhouse gases create cross-border damages.
The harder question is how large.
COP negotiations struggle because there is no global government capable of taxing emissions and redistributing revenue automatically.
Every transfer remains voluntary enough that national politics constrains it.
American voters ask why money should go abroad.
Indian voters ask why they should pay for emissions accumulated before they became wealthy.
European voters face high energy and tax costs.
African governments ask why development should be constrained by a climate problem they contributed little to producing.
Every argument contains some validity.
Collectively they create insufficient finance.
This is the global collective-action trap.
The optimal global policy can be obvious while no individual government has sufficient incentive to finance its share.
That is why climate agreements advance through imperfect compromises rather than one economically optimal solution.
The Baku agreement is one.
$300 billion is much more than $100 billion.
It is much less than developing countries wanted.
It does not solve the financing gap.
It creates another institutional benchmark against which future governments can be judged.
The real test will therefore not be whether diplomats can announce a larger number in 2035.
It will be whether the number changes actual investment decisions between now and then.
Do grids get built?
Does clean electricity become cheaper?
Do vulnerable countries receive adaptation finance before disasters rather than loans afterwards?
Does private capital flow into emerging economies at lower risk premiums?
Do countries gain enough financing that climate commitments become compatible with development rather than perceived as constraints imposed by wealthy states?
Those are the outcomes that matter.
Climate diplomacy has spent decades negotiating tonnes of carbon and billions of dollars.
The underlying problem is simpler.
Developing countries need to become much richer.
The world simultaneously needs them to industrialise through technologies far less carbon-intensive than those wealthy countries used.
That transition costs money today but reduces enormous costs tomorrow.
The Baku argument is therefore not fundamentally about generosity.
It is about who finances the capital stock of a low-carbon global economy when the countries that need to build the most infrastructure are often the countries where capital is most expensive.
Until that financing problem is solved, climate targets will continue outrunning the institutions capable of delivering them.
COP29 has increased the number.
It has not yet closed the gap.
