Polycrisis Resilience: The H Heuristics Concept

The 21st century is characterised by an increasingly interconnected set of global challenges. Climate change, environmental pollution, biodiversity loss, poverty, geopolitical instability, demographic pressures, technological disruption and economic inequality are not isolated problems. They interact, reinforce one another and create cascading vulnerabilities across societies, economies and ecosystems. This interconnected condition is increasingly understood through the concept of the polycrisis: a situation in which multiple crises interact in ways that produce consequences greater than the sum of their individual effects.

Addressing such conditions requires more than conventional crisis management. It requires a systematic effort to strengthen the underlying capacities of humanity to prevent, withstand, adapt to and recover from interconnected threats. H Heuristics defines this broader approach as Polycrisis Resilience. It is concerned not only with responding to crises after they emerge but with reducing the structural vulnerabilities that allow crises to develop, interact and escalate. It emphasises the role of economic development, technological progress, innovation diffusion, institutional capacity, environmental improvement and global convergence in reducing human exposure to interconnected risks. The central objective is to improve humanity's ability to navigate an increasingly complex global risk environment while preserving and improving long-term human survivability.

The definition encompasses both the reduction of existing vulnerabilities and the development of new capabilities that allow societies to manage future disruptions. It also recognises that resilience is not distributed equally. Countries, regions, communities and populations have vastly different access to the resources, technologies, infrastructure and institutions necessary to withstand interconnected crises. Consequently, strengthening global resilience requires particular attention to vulnerable populations and developing economies. The distribution of the capacities is measured in section 5: the mean low-income country is at the 22nd percentile of composite capacity and the mean high-income country at the 76th.

Provenance of the words. The word polycrisis is attributed in secondary sources to Edgar Morin in the 1990s and was popularised in the 2020s by the economic historian Adam Tooze and by the Cascade Institute, whose 2022 paper (Lawrence, Janzwood and Homer-Dixon) defines a global polycrisis as crises in multiple global systems that become causally entangled in ways that significantly degrade humanity's prospects. The World Economic Forum's Global Risks Report 2023 used the term for a cluster of related global risks whose compounding effects exceed the sum of their parts. Resilience, in the UN disaster-risk terminology of 2017, is the ability of a system exposed to hazards to resist, absorb, accommodate, adapt to, transform and recover from their effects, preserving and restoring essential functions.

Provenance of the phrase. A search for a settled definition of "polycrisis resilience" found discussions of polycrisis and of resilience as a response, and loose uses of the compound in commentary, but no authoritative definition. The H Heuristics series has used the phrase since 11 October 2026 (report 2026-21, Polycrisis Resilience in the Twenty-First Century) and has developed it through several reports since, among them 2026-79, 2026-81, 2026-83 and 2026-84. This page states the H Heuristics definition and does not claim priority over other uses of the words.

What the distinction is for. A term earns its place by changing what is measured or done. Here three things change. The unit of analysis is the set of interacting crises, not a single hazard or a single response phase. The object of policy is a set of five capacities developed in advance, systematically, not an emergency capability. And the test is the reduction of cascading effects and long-term consequences, which is why the series measures transmission, scarring and the stability of resilience across shocks and not only the size of one shock's damage.

Convergence. The definition's emphasis on global convergence rests on the observation that vulnerability is largely a function of where a country stands in development. Report 2026-63 finds that countries vulnerable in 2000 have closed their human-development gap at about 0.3% a year, a half-life of two centuries, and that the countries that avoided reversals closed it faster. The speed of convergence is therefore the main variable, which is why the proposition says accelerating.

Innovation diffusion and sustainable development. The technologies that reduce vulnerability, clean power, connectivity, early warning, diagnostics, exist; the gap is in their spread. Report 2026-75 finds, in US data, that technologies embedded in infrastructure diffuse about five times more slowly than free-standing ones and that no low-income country in its data has passed 50% internet use; report 2026-78 finds that high-income countries account for 57% of Claude.ai usage against 0.2% for the five published low-income countries, with income elasticity of 0.94. On the sustainable-development side, report 2026-77 finds that clean electricity is gaining about one percentage point a year against 2.3 needed for a fully clean grid by 2050.

Adaptive capacity and preventative intervention. The proposition's remaining terms concern how much shocks cost and how early money is spent. Report 2026-84 finds that growth lost to shortfalls is 0.9 points a year against 0.04 for volatility itself, that shortfalls are largely permanent and that 65% of the loss is idiosyncratic and not disaster-linked; report 2026-80 finds that early-warning returns are large in assets where assets are and in lives where hazards are, but arrive in lumps. Report 2026-74 finds that donors spent 8 cents on prevention for every dollar spent on response in 2015–24, against published returns to prevention of 2 to 87 dollars per dollar.

The matrix makes explicit that no single capacity does all four jobs and that most capacities contribute to more than one. Economic buffers absorb and finance recovery; technological monitoring prevents and warns; institutions coordinate every function. A country strong in one cell and weak in the neighbouring one is exposed where the two meet, which is the mechanism by which a shock in one system cascades into another.

The definition's insistence on systematic development is a statement about sequencing and balance: capacities developed separately, in separate ministries and budgets, leave gaps at the interfaces. Report 2026-85 finds that earlier progress in water and sanitation, power and income predicts faster later progress in other systems, and that lopsided development is associated with worse survival in the pandemic (−0.33 standard deviations of life expectancy against trend per standard deviation of imbalance).

The matrix is also an audit tool. A national or sectoral plan can be read against it: which of the twenty cells does it fund, which does it leave empty, and does any capacity carry a function alone? The scorecard in section 5 measures the five capacities; the functions are not yet measured separately because the data for prevention and recovery in particular are thin.

Each capacity is the mean percentile rank of two to four indicators among the 145 countries with all five. Mean composite capacity is the 76th percentile in high-income, 51st in upper-middle, 33rd in lower-middle and 22nd in low-income countries. The gaps are widest for technological capacity (82 against 11) and human capacity (79 against 14) and narrowest for environmental capacity (66 against 36), which includes clean cooking and water and sanitation but also lower particulate exposure in poorer, less industrial countries.

The capacities are highly correlated (0.59 to 0.92), so the first principal component carries 81% of their variance, the next four 9%, 6%, 3% and 1%. This contrasts with the exposure side measured in report 2026-82, where the four exposure domains are close to independent. The asymmetry matters: exposures differ by geography and cause, but capacities accumulate with development. That is why the definition frames capacity-building as an integrated development approach, and why it speaks of convergence.

Income does not determine capacity. Relative to the capacity income predicts, Estonia, Finland, Burundi, the Kyrgyz Republic, Rwanda, The Gambia, Madagascar and Uruguay score 12–13 percentile points higher; Equatorial Guinea (−29), Gabon (−22), Bahrain, Eswatini, the Republic of Congo and Angola (−17 each), Djibouti and Qatar (−14) score lower, mostly resource-dependent economies with high income and thinner institutions, health and technology. The weakest capacity is economic in 39 countries, human in 38, environmental in 36, institutional in 21 and technological in 11.

Composite capacity correlates −0.92 with INFORM's lack-of-coping score, −0.83 with INFORM's overall risk, −0.75 with its vulnerability score and −0.82 with the unified exposure score of report 2026-82. These are strong, and partly built in: INFORM's coping score uses governance and infrastructure indicators that overlap with the capacities, and the exposure score includes technological indicators, so the correlations show consistency, not independent confirmation.

Against realised outcomes the relationships are weak: −0.14 with growth lost to shortfalls and −0.15 with growth volatility (report 2026-84), +0.02 with life expectancy against trend in 2021 and −0.04 with composite resilience in the 2020 episode (report 2026-83). This is consistent with the finding of report 2026-83 that resilience in one shock is almost unrelated to resilience in the next: shock outcomes depend on the shock. It is also a limit on what a capacity scorecard can claim.

So the evidence supports a structural claim: vulnerability falls as the five capacities rise, the capacities rise together, the distribution is very unequal and the weakest capacity differs by country. It does not support the claim that a high capacity score protects a country in a given crisis. The concept's own wording is careful on this point: the aim is to reduce cascading effects and long-term consequences, and those are measured over time and across crises, which the series does through transmission (report 2026-65), scarring and pooling (2026-84) and spillovers between systems (2026-85).

The explorer recomputes a composite from the five capacity percentiles under the weights chosen. It lists the fifteen lowest-scoring countries, the countries that move ten places or more against equal weights and the capacity gap of the chosen country: the difference between its weakest capacity and its strongest, and against the median of its income group.

Weights express judgement about which capacity matters most for a given purpose: a health funder, a development bank and a climate fund would weight differently, and the definition does not prescribe weights. Report 2026-82 found that unified rankings are sensitive to weights; the same is true here, though less so, because the five capacities are highly correlated and a change of weights moves ranks less.

The explorer measures capacity levels at the national level, the first step in operationalising the definition. A fuller scorecard would add the functions of section 4, within-country distribution and capacity measured against specific crisis channels; the research agenda in section 9 lists these.

One sentence. Polycrisis Resilience is the systematic development of human, institutional, technological, economic and environmental capacities to prevent, absorb, adapt to and recover from interconnected global crises while reducing their cascading effects and long-term consequences.

One paragraph. The world faces crises that interact: climate, economic, technological, geopolitical and social stresses reinforce one another and cascade. Polycrisis Resilience, a concept of H Heuristics, treats the response to this condition as an integrated development agenda: accelerate convergence, spread innovations, build sustainable and adaptive systems, intervene early, and attend first to the countries and people with the least capacity. Its test is whether cascades and long-term consequences fall.

How to cite. H Heuristics (2026). Polycrisis Resilience: The H Heuristics Concept. H Heuristics Digital Reports, 2026-86. https://digitalreports.hheuristics.com/reports/polycrisis-resilience-concept/ . Cite the definition as "Polycrisis Resilience (H Heuristics, 2026)" and the evidence by the number of the report that supplies it.

Capacity indicators (2020 or the latest year): human = life expectancy (WDI SP.DYN.LE00.IN) and gross secondary enrolment capped at 100 (SE.SEC.ENRR); institutional = WGI government effectiveness and the World Bank Statistical Performance Index (via report 2026-62); technological = the H Heuristics diffusion index (2026-75), the IMF AI Preparedness Index (2023) and internet users (IT.NET.USER.ZS); economic = log GDP per person, PPP (NY.GDP.PCAP.PP.KD), reserves in months of imports and interest as a share of revenue (inverted; via 2026-62); environmental = access to clean cooking (EG.CFT.ACCS.ZS), mean basic water and sanitation access (SH.H2O.BASW.ZS, SH.STA.BASS.ZS), PM2.5 exposure inverted (EN.ATM.PM25.MC.M3) and the renewable share of final energy (EG.FEC.RNEW.ZS).

Scoring: each indicator is converted to a percentile rank among countries with data; a capacity is the mean of its available indicators (at least half required); the composite is the mean of the five capacities; the 145 countries with all five and at least one million people are used. Correlations are Spearman rank correlations. The principal-component shares are from the correlation matrix of the five capacities. The income prediction is a linear regression of the composite on log GDP per person.

Caveats: the scorecard is a first operationalisation; capacities are national and relative; the environmental capacity includes outcomes (particulate exposure) as well as inputs; indicators have uneven coverage and years; the comparison with INFORM and the unified exposure score uses overlapping indicators; the correlations with realised outcomes are single-episode or single-period; and the glossary and matrix are conceptual statements, not findings.