A Proactive Strategy for Mitigating the Twenty-First-Century Polycrisis

Executive Summary

Every serious strategy for the polycrisis names the same four elements. The list has never been the problem. What is missing is an account of how the four differ — and they differ in the one way that determines which instrument can deliver them.

This series has already examined what to deploy, which constraint binds, why it matters for growth, when to buy it, and where in a system to put it. Each takes one element and examines it closely. None asks the question this report asks, which is what happens when they are combined — and specifically, whether the governance machinery available at global scale can deliver all four, or only some.

The answer is that it can deliver one of them well. The international system's characteristic instruments — negotiated targets, pledge-and-review, global goals, summit cycles — are designed for a global public good with a free-rider problem, which describes climate mitigation and describes almost nothing else on the list.

Those four numbers describe three distinct failures, not one. The first two are the signature of a coordination failure: an identical externality priced fifteen-fold differently, and a counter-subsidy seventy times larger than the price signal. The third is not a coordination failure at all — nobody free-rides on a sea wall — but a financing one. The fourth is neither: it is the gap between what a state can declare and what it can operate, and no amount of external money closes it directly.

1. The Same List, Every Time

Convergence on the right answer has not produced the right outcome, which suggests the difficulty lies somewhere other than in identifying what to do.

Read a dozen strategies for managing systemic risk this century and the contents barely vary. Decarbonise the energy system. Adapt to the warming already committed. Sustain development so that populations are less exposed and states better resourced. Strengthen the institutions that have to run all of it. The list is not contested, and it is not wrong.

It is also not a strategy. A list of desirable objectives becomes a strategy only when it says something about relationships: what must happen before what, which items reinforce one another, which compete, and — the question this report is concerned with — which piece of machinery is capable of delivering each one. Those questions are answered rarely, and when they are answered it is usually by assuming that a single approach works for everything on the list.

That assumption is the subject of this report, and the claim is that it is false in a specific and consequential way. The four elements are not four instances of the same kind of thing. They differ in their economic geometry — in who receives the benefit and in whether the good can be supplied by anyone other than the beneficiary — and that geometry determines which instrument can deliver them. Combined strategies underperform because they apply one instrument to four different geometries.

2. Four Goods, Four Geometries

Two questions, asked of each element, sort them into three different problems requiring three different remedies.

The first question is who captures the benefit. A tonne of carbon not emitted benefits everyone and no one in particular; the atmosphere does not distinguish between the country that paid and the country that did not. A sea wall benefits the people behind it and nobody else. This is the difference between a global public good and a largely private or local one, and it determines whether free-riding is possible.

The second is whether the good can be supplied from outside. A solar farm can be financed, built and operated by foreign capital and foreign contractors. So, to a large extent, can a drainage system. A competent procurement function, a credible auction that bidders believe will be honoured, an agency whose staff turn up and whose decisions stick — these cannot be delivered by an external party at all. They can be assisted, advised and financed at the margin, but they are produced domestically or not at all.

Read the quadrants and three distinct problems appear. Where the benefit is global and the good is externally suppliable, the difficulty is coordination: everyone would rather someone else paid. Where the benefit is local and the good is externally suppliable, there is no coordination difficulty whatsoever — but there may be a financing one, if the party who captures the benefit lacks the means to buy it. And where the good cannot be supplied externally at all, the difficulty is capability, and money is not the operative variable.

3. Mitigation: The Coordination Problem, Made Visible

The free-rider structure is textbook. What is less often noticed is that it leaves a legible fingerprint in prices.

Mitigation is the case the machinery was built for, and even here the record is mixed in a way that is directly attributable to the geometry. Because the benefit is non-excludable, no country can capture the return on its own abatement, and the collectively rational price is not individually rational to charge. The result is not an absence of carbon pricing — there is a great deal of it — but an absence of any common level.

A fifteen-fold spread in the price of an identical global externality is what the absence of a binding collective mechanism looks like when it can be measured. And the price signal such as it is competes against a much larger one running the other way.

Two responses to this structure are worth distinguishing, because only one of them is a genuine solution.

The first is coercion in place of agreement. The EU's Carbon Border Adjustment Mechanism entered its definitive period on 1 January 2026, extending a domestic carbon price to imported goods. It works on its own terms, and it is contentious precisely because it substitutes unilateral extraterritorial reach for a negotiated common price. It manages the free-rider problem rather than resolving it, and it does so in a way that transfers the political cost to countries with the least capacity to absorb it.

The second is more interesting and is the reason mitigation has advanced despite everything above. A mitigation action that produces a saleable product stops being a public good at the margin. A solar farm sells electricity. Its investor captures a private return and does not need a treaty, a pledge or a price to justify the outlay. Cost reduction, in other words, does not merely make abatement cheaper; it moves abatement out of the quadrant where free-riding governs and into one where ordinary commercial incentives do. That is the deepest reason the deployment argument made in an earlier report in this series has purchase where three decades of target-setting did not.

4. Adaptation: The Category Error

Adaptation has no free-rider problem, which makes its chronic underfunding a puzzle — until one notices that it is being managed with instruments designed for a problem it does not have.

Nobody free-rides on a flood defence. The country that builds it captures essentially all of the benefit, which by the logic of Section 2 should make adaptation the easier of the two climate agendas to finance: the incentives are aligned, the beneficiary is identifiable, and no international agreement is required for a state to act in its own interest.

The observed pattern is the reverse.

The explanation is not that adaptation's returns are worse — they are, if anything, better, and the evidence is assembled in an earlier report. It is that adaptation lacks the feature which rescued mitigation. A sea wall has no revenue line. It produces avoided losses, which accrue diffusely, arrive stochastically, and appear in no one's accounts as income. There is nothing to sell, so private capital has nothing to price, so the good depends on public finance — and the public in question is, disproportionately, one of the poorest.

That diagnosis has an uncomfortable corollary for how adaptation is currently governed. Adaptation has been folded into the same architecture as mitigation: global goals, pledge cycles, headline finance targets announced at summits. Those instruments address free-riding. Adaptation does not suffer from free-riding. Applying them produces the observable result — a great deal of declaratory activity and very little money — because the instrument is not attached to the actual failure.

5. The Good That Cannot Be Imported

Institutional resilience is not underfunded so much as un-purchasable — and because everything else runs through it, it sets the rate at which the rest of the strategy can proceed.

The clearest evidence that capability is a distinct constraint is the gap between what states can declare and what they can operate. Writing a national adaptation plan requires analysis, drafting and a cabinet decision. Running a multi-hazard early-warning system requires observation networks, forecasting, maintenance budgets, standing staff, communication channels and a chain of local actors who respond when the alert fires.

The same asymmetry appears wherever a policy requires sustained administration rather than a decision. Carbon pricing coverage concentrates in wealthy and upper-middle-income economies with the fiscal and administrative capacity to design, monitor and enforce a price, and is thin across the low- and lower-middle-income world — not principally because the idea is contested there, but because operating an emissions trading system is a demanding standing function.

This is what makes institutional resilience different in kind from the other three. It is not a fourth item that competes for the same budget. It is the rate limit on the conversion of money into installed capability, which means it multiplies or divides the return on everything else on the list. External finance arrives at a state's absorptive capacity and stops.

5.1 The honest difficulty

It would be convenient to conclude that capability should therefore be built first and everything else second. The record does not support that, and the reason is instructive. Capacity building as a discrete line item — workshops, advisers, seconded experts, strategy documents — has a poor track record precisely because the capability in question is tacit and organisational rather than informational. Officials generally know what a competent procurement process looks like.

What builds the capability is doing the thing, repeatedly, with real money and real consequences: an auction run four times rather than once, a permitting process that processes a pipeline, an agency with a standing budget that outlasts a political cycle. Capability is a by-product of sustained delivery, not a precondition assembled in advance of it. The practical implication is not to sequence capability ahead of deployment but to couple them — to choose deployment programmes partly for the institutional muscle they build, and to protect them long enough for that muscle to form.

6. Complementary in Output, Competitive in Input

The four reinforce one another in what they produce and compete for what they consume. Strategies advertise the first relationship and budget as though the second did not exist.

Table 1 — How the four elements interact

Relationships are analytical judgements drawn from the evidence in Sections 3 to 5 rather than measured elasticities. The pattern — reinforcement in output, competition in input — is the point.

Both columns are true, and strategies routinely cite only the first. The complementarities are real: a grid that stays up during a flood is at once mitigation, adaptation and development, and the case for double-duty investment is genuine. But complementarity in output does not relieve competition in input. Four programmes drawing on one procurement office, one planning ministry and one fiscal envelope do not become cheaper because their benefits overlap.

This is where the tension becomes concrete rather than rhetorical. Roughly 600 to 700 million people still lack reliable access to electricity, and per-capita emissions across most of those economies are a small fraction of the OECD average. A government weighing rapid access against least-cost decarbonisation is not being timid; it is resolving a genuine conflict between two items on the same list, under a shared constraint. A strategy that has not acknowledged the conflict offers that government nothing.

7. What a Proactive Strategy Requires

Proactive means acting before compulsion. That is possible only where the instrument fits the good — otherwise the outcome is pledge-and-fail or need-and-underfund, whatever the intent.

Table 2 — Matching the instrument to the geometry

7.1 Stop applying one template to four problems

• Reserve pledge-and-review machinery for the goods that actually have a free-rider problem. It is a sound instrument in that quadrant and close to inert outside it.
• Move adaptation out of the pledge architecture and into the financing architecture, where its actual constraint lives: concessional windows, risk pools, and lending instruments that price avoided loss.
• Stop reporting institutional capability as a finance number. It is not one, and reporting it as one obscures the only variable that matters.

7.2 Exploit the complementarities; budget for the competition

• Prefer investments that serve more than one element at once — reliable clean power being the clearest, since it is simultaneously mitigation, an input to protective services, and a development good.
• Budget the shared constraint explicitly. Publish how much administrative and procurement capacity each programme consumes, not only how much money, and stage programmes against that constraint rather than against the fiscal one alone.
• Where two elements genuinely conflict — energy access against least-cost decarbonisation being the standing example — say so, and resolve it deliberately rather than by leaving both in the plan.

7.3 Build capability by doing

• Choose deployment programmes partly for the institutional muscle they build: repeated auctions over one-off procurements, standing pipelines over pilot projects.
• Protect those programmes across political cycles for long enough that the organisational learning accumulates. Capability is a by-product of continuity, and continuity is the thing most easily withdrawn.

7.4 Conclusion

The reason strategies for the polycrisis converge on the same four elements is that the elements are correct. The reason they disappoint is that agreement on a list has been mistaken for possession of a strategy. What is missing is not ambition or analysis but a match between instrument and problem — and the mismatch is systematic rather than incidental, because the machinery available at global scale was designed for one of the four and has been asked to carry all of them.

The corrective is neither a new institution nor a larger number. It is a sorting exercise. Establish, for each element, who captures the benefit and whether anyone but the beneficiary can supply it; read off the failure mode; and attach the instrument that addresses that failure rather than the one the calendar happens to offer. Mitigation needs coordination and, more powerfully, cost reduction that makes coordination unnecessary. Adaptation needs finance and a manufactured revenue line, and needs to be released from a pledge architecture that cannot help it. Institutions need continuity of delivery, which is the one thing that cannot be pledged at all.

Proactivity, on this reading, is not a disposition but a consequence. A state or a system acts before it is compelled when the instruments available to it can actually move the thing it is trying to move. Where they cannot, no amount of foresight substitutes, and the result is what the record already shows: a great deal of declaration, and a strategy that remains a list.

References

Every quantitative claim above is attributed inline. The principal sources are collected here.

• World BankState and Trends of Carbon Pricing 2025 — 80 instruments worldwide, 43 carbon taxes and 37 emissions trading systems, and the price range across jurisdictions.
• IMFFossil Fuel Subsidies Data: 2025 Update — US$725bn explicit and US$6.7tn implicit in 2024, against roughly US$100bn raised by carbon pricing globally.
• European CommissionThe Carbon Border Adjustment Mechanism entered its definitive period on 1 January 2026.
• IEAWorld Energy Investment 2025 — clean-energy investment of about US$2.2 trillion a year.
• UNEPAdaptation Gap Report 2025 — US$26bn of international public adaptation finance against US$310–365bn of assessed need; 172 of 197 countries with a national adaptation plan.
• UNDRREarly Warning Systems Reach New Heights, but Critical Gaps Jeopardize Global Progress — 119 countries, about 60 per cent. See Early Warnings for All.
• UNDRRGlobal Assessment Report on Disaster Risk Reduction 2022: Our World at Risk — institutions built for discrete hazards against systemic risk.
• World Economic ForumGlobal Risks Report 2026 — the interconnection of the risk landscape.
• World Bank / ESMAPTracking SDG 7: The Energy Progress Report 2025 — the scale of the remaining electricity-access deficit.
• United NationsThe Sevilla Commitment — the 2025 Financing for Development outcome on debt, multilateral bank reform and domestic resource mobilisation.
• H HeuristicsEarlier reports in this series, whose single-element arguments this report takes as given: deployment as mitigation infrastructure; the adaptation finance and delivery gates; adaptation and convergence; the economics of prevention; and adaptation read topologically.