The Polycrisis Is a Diffusion Problem

Executive summary

One crisis, one engine, three levers

The global polycrisis looks like a stack of separate emergencies: pollution, warming, poverty, water stress, fragile supply chains. It is not. It is one coupled system, and the escape route runs through a single engine. The same diffusion forces that spread clean technologies and close income gaps also break the feedback loops that keep the crises locked together.

Every intervention that accelerates an S-curve, a catch-up path, or a decarbonization loop attacks all three crises at once. The report builds that argument in four moves. First, it shows the polycrisis is a causal loop, not a list. Second, it shows clean technology has compressed the slow half of its adoption curve into a decade. Third, it shows convergence has delivered the largest poverty reduction in recorded history, and cheap clean energy now makes the next round of catch-up cleaner than the last. Fourth, it shows the three forces compound into a single loop whose only binding constraint is time.

02 · The polycrisis

The polycrisis is one system, not a portfolio of crises

The default framing treats pollution, warming, and poverty as separate problems with separate ministries and separate scorecards. The evidence rejects that framing.

Warming cuts crop yields and outdoor labor productivity, which deepens poverty. Poverty pushes households toward biomass and coal, which worsens pollution. Pollution damages health and reduces earnings, which deepens poverty again. Each crisis is an input to the others. This is a polycrisis cascade: multiple compounding crises where failure in one domain accelerates failure in another.

The scale of the coupling is measurable. Atmospheric carbon dioxide sits near 424 parts per million, against roughly 280 before industrialization. Long-term global mean temperature stands about 1.3 degrees Celsius above the preindustrial baseline, and 2024 became the first calendar year to cross 1.5 degrees, per the World Meteorological Organization. Roughly 7 million people die prematurely each year from air pollution, according to the World Health Organization. About 692 million people still live below the World Bank's extreme poverty line of $2.15 a day. These are not four statistics. They are four positions on the same causal loop.

The implication follows directly. A pollution crackdown that raises energy prices can push households back to cheap solid fuels. A poverty program built on coal-fired growth buys income now and buys warming later. Only interventions that work on the loop itself, rather than on one node of it, compound.

03 · Diffusion of innovations

Diffusion does the heavy lifting

Diffusion of innovations moves in S-curves. Adoption starts slow while costs fall and trust builds, then bends sharply upward as the technology crosses a threshold, then flattens as the market saturates.

Rogers documented the pattern in 1962. It still governs every technology that matters here. Clean technology has compressed the front half of the curve into a decade. Solar photovoltaic costs have fallen roughly 90 percent since 2010, onshore wind levelized costs about 70 percent, and lithium-ion battery pack prices about 90 percent, per IRENA and BloombergNEF. The adoption data follow the cost data. Wind and solar supplied under 2 percent of global electricity in 2010 and roughly 15 percent in 2024, per Ember. Electric vehicles moved from well under 1 percent of new car sales in 2010 to about 20 percent in 2024, per the International Energy Agency. These are textbook S-curves caught mid-bend.

The mechanism is learning-by-doing. Every doubling of cumulative solar deployment cuts unit costs by roughly 20 percent; batteries track a similar learning rate. Diffusion therefore does two jobs at once. It spreads a cleaner technology, and it makes that technology cheaper for every adopter who comes later. That second job is the one that changes the arithmetic of the polycrisis.

04 · Economic convergence

Convergence is the poverty exit ramp

Economic convergence is the empirical regularity that poorer economies grow faster than rich ones once they have the institutions, capital, and technology to absorb. Barro and Sala-i-Martin measured conditional convergence at roughly 2 percent a year.

The growth records since 2000 show the mechanism working at scale. China's income per person multiplied about 13 times from 2000 to 2023; Vietnam's about 11 times; India's more than 5 times; Bangladesh's about 6 times. Extreme poverty fell from about 38 percent of humanity in 1990 to about 8.5 percent in 2024. That is the largest poverty reduction in recorded history. Catch-up growth powered it: adopting proven technologies, plugging into trade, and building the infrastructure to do both.

Energy access sits inside this mechanism. About 685 million people still lack electricity, and 2.1 billion lack clean cooking, per the 2024 SDG 7 tracking report. Every one of those gaps is a convergence gap. Households without electricity cannot mechanize, refrigerate, or study at night; firms cannot run reliable equipment. Closing the gap was once expensive because it meant extending coal grids. Cheap solar, batteries, and mini-grids change that trade-off. Late developers can now leapfrog directly to distributed clean power instead of rebuilding the fossil pathway. Diffusion makes convergence cheaper, and convergence creates the demand base that pulls the next wave of clean technology forward.

05 · The clean transition

The clean transition is the coupling mechanism

The clean transition is the point where diffusion and convergence meet the polycrisis. It decarbonizes energy, which attacks warming. It displaces combustion, which attacks pollution. It extends energy access, which attacks poverty. One transition, three crisis levers.

Capital has begun to follow the mechanism. Global clean energy investment reached roughly 2 trillion dollars in 2024, against about 1.1 trillion for fossil fuels, per the International Energy Agency. Renewable energy employment reached 16.2 million in 2023, per IRENA. Where the transition has run longest, the health dividend shows up in the air quality data: urban fine-particle pollution in China fell roughly 40 percent between 2013 and 2020 as coal growth stalled and clean power scaled. The same pattern is now visible in early form across India, Southeast Asia, and parts of Africa.

The policy conclusion is uncomfortable for ministries that still file clean energy under environment. The transition is a development strategy. It is the fastest available route to universal energy access, the cheapest new power in most markets, and the only growth model that does not reimport pollution into the lungs of the people it lifts out of poverty.

06 · The compound loop

The DR-D-C-T loop compounds

Diffusion, convergence, and the clean transition are usually analyzed as three separate literatures. They are three positions on one loop.

Clean technology diffuses down its cost curve. Cheap clean technology lets poorer economies converge without the pollution bill of the old pathway. Convergence expands the markets and supply chains that make clean technology cheaper still. The resulting transition de-escalates all three polycrisis nodes. Each turn of the loop lowers the cost of the next turn.

The loop also raises what this analysis calls the D-coefficient: an economy's absorptive capacity for change. Grids, workforces, and institutions that absorbed the last wave of clean deployment can absorb the next one faster. Capacity compounds the same way costs do. The binding constraint is time. Every year the loop runs slower than it could, the world adds to what this analysis calls the P-integral: cumulative pollution exposure integrated over time. Emissions this year do not vanish next year; they accumulate in the atmosphere and in lungs. The P-integral is the one cost in this system that never discounts.

↻ each turn lowers the cost of the next turn

07 · The counter-case

The honest counter-case

The counter-case does not dispute the mechanism. It disputes the speed. Policies currently on the books point toward warming near 3 degrees Celsius by 2100, per the UNEP Emissions Gap Report 2024. The loop is running, and it is still running too slowly.

The obstacles are real. Clean capital concentrates in rich countries and China, while emerging markets pay the highest financing costs precisely where new demand is largest. Grids and permitting move at bureaucratic speed while solar factories move at manufacturing speed. Critical minerals for batteries and grids are geographically concentrated, and their processing is more concentrated still. Rebound effects can eat part of the efficiency dividend as energy gets cheaper. And convergence reduces poverty on average while inequality within countries can still widen.

None of these obstacles refutes the loop. Each one is itself a diffusion problem. Capital must diffuse into emerging markets. Grid capacity must diffuse into lagging regions. Mineral supply must diffuse out of chokepoints through substitution and recycling. The counter-case argues for accelerating the same three forces, with more finance and harder industrial policy, not for a different theory of change.

08 · Implications

Measure curves, fund the bend, build capacity

The loop reframes what to build, measure, and fund.

Measure adoption curves, not pledges. A country's announced targets matter less than whether its solar, storage, and EV S-curves are bending on schedule. Fund the middle of the S-curve in emerging markets, where a dollar buys more adoption and more convergence than anywhere else. Treat energy access as a convergence instrument, not a charity line item; every gigawatt of distributed clean power is development policy and climate policy delivered together. Build absorptive capacity before the wave hits: grid codes, workforce training, procurement standards, and the institutions that let the D-coefficient compound.

The cheap path is the fast path. Clean technology is now the low-cost option in most markets, and delay no longer buys a cheaper transition; it buys a larger P-integral, more locked-in warming, and more avoidable deaths. Treating its symptoms one at a time will not solve the polycrisis. Diffusion, convergence, and the clean transition running as one system will outrun it.

09 · Sources & notes

Sources and notes

Figures are rounded to the precision the underlying sources support. Chart series marked "approximate" interpolate between published data points for readability; the direction and order of magnitude are the claim, not the decimal.

1. IPCC, AR6 Synthesis Report (2023) — climate system state and warming attribution.
2. World Meteorological Organization, State of the Global Climate 2024 — 2024 as the first calendar year above 1.5°C.
3. NOAA Global Monitoring Laboratory, Mauna Loa CO₂ trends — atmospheric CO₂ concentration.
4. World Health Organization, Air pollution — premature mortality estimates.
5. World Bank, Poverty, Prosperity, and Planet (2024) — extreme poverty counts and shares.
6. IRENA, Renewable Power Generation Costs in 2023 — solar and wind cost declines.
7. BloombergNEF, Lithium-ion battery price survey (2024) — battery pack price decline.
8. Ember, Global Electricity Review 2025 — wind and solar share of global electricity.
9. International Energy Agency, Global EV Outlook 2024 — electric vehicle sales shares.
10. International Energy Agency, World Energy Investment 2024 — clean versus fossil investment.
11. IRENA and ILO, Renewable Energy and Jobs (2024) — renewable energy employment.
12. IEA, IRENA, UNSD, World Bank, WHO, Tracking SDG 7: The Energy Progress Report 2024 — electricity access and clean cooking gaps.
13. UNEP, Emissions Gap Report 2024 — policy-implied warming trajectory.
14. Barro, R. and Sala-i-Martin, X., "Convergence," Journal of Political Economy 100(2), 1992 — conditional convergence estimates.
15. Rogers, E., Diffusion of Innovations, Free Press (1962; 5th ed. 2003) — the S-curve framework.
16. Ministry of Ecology and Environment of China, Air quality reports — urban PM2.5 trends, 2013–2020.