Sustainable Convergence: The H Heuristics Concept

How to read this page. Section 1 states the definition and the six dimensions, and what the concept is not. Section 2 asks whether convergence has happened. Section 3 shows that access has converged fastest and what the pace gap is worth. Section 4 asks whether catch-up has been clean. Section 5 scores 151 economies on the six dimensions. Section 6 tests whether progress in one dimension pulls the others along. Section 7 is the pathways model to 2100. Section 8 is priorities, research agenda and how to use the term.

Status of the evidence. The record, typology, latecomer, decoupling and index results come from the H Heuristics Sustainable Convergence report of November 2026 and public data (World Development Indicators, Our World in Data, UN population projections, INFORM). The six-dimension score, the access pace scenario and the interactive model are specific to this page. Scenarios are mechanical extrapolations, not forecasts. The concept is presented as a proposed framework open to testing.

A note on versions. This page follows the six-dimension statement above, in which infrastructure is part of human development and resilience is a dimension. The November 2026 report used a seven-dimension version that listed infrastructure separately; scores here merge the two, and the two versions correlate at 0.998. The phrase "sustainable convergence" has been used before in other settings, largely European Union regional and sustainability policy; a November 2026 title-and-abstract check found about 300 earlier uses. The definition and framework here are H Heuristics's own.

What is new. The distinctive claim is that catch-up and sustainability can reinforce rather than trade off. Renewable electricity powers industrialisation while cutting emissions and air pollution; digital public infrastructure widens participation and makes government cheaper to run; healthier, better-educated workers are more productive; climate-resilient infrastructure protects the assets growth builds. When these links are strong, investment in one dimension raises returns in the others. The concept also rejects "grow first, clean later" on two grounds: many environmental harms are irreversible, and clean technology is now cheap.

What it is not. It is not a call for developing countries to grow less: billions need more electricity, transport and housing, and the energy needed for decent living standards for all, while modest by rich-country standards, is far above what the poorest consume now (Kikstra et al. 2021). It is not a demand that poor countries carry the burden of decarbonisation: the high-income world produced most historical emissions and must cut fastest. It is not the claim that convergence is automatic, and it is not a single blueprint: the right path for Ethiopia is not the right path for Brazil.

Three tests. The definition implies three tests for any strategy or national record. A convergence test: is the gap with advanced economies narrowing, relatively and in time absolutely, across the dimensions that matter? A sustainability test: is progress achieved without crossing environmental thresholds such as safe air quality and a fair share of the carbon budget? A resilience-and-inclusion test: do gains reach the whole population and survive the shocks most likely to undo them? Passing one alone is catching up on the old model, protecting the environment while people stay poor, or coping without advancing.

Relation to Polycrisis Resilience. The two concepts developed by Hunter Hughes form a loop (Hughes 2026a; Hughes 2026b). Polycrisis Resilience is the objective: reducing human and systemic vulnerability to interconnected crises. Sustainable Convergence is a development pathway towards it: it builds the income, infrastructure, technology, health and institutions that make societies less vulnerable. Without convergence, resilience is very hard to achieve: a population without savings, power or functioning institutions has little margin for a flood or a price spike. Without resilience, convergence is fragile: severe shocks can wipe out decades of gains, and in many developing economies income losses after severe shocks are never fully recovered (Cerra and Saxena 2008; Hsiang and Jina 2014).

Measured two ways. Convergence can be measured relatively (income as a share of the high-income level) and absolutely (the dollar gap). Economists distinguish absolute convergence, where poorer economies grow faster regardless of conditions, from conditional convergence, which holds once policies and institutions are comparable (Barro and Sala-i-Martin 1992; Baumol 1986). The concept asks for the relative gap to narrow now and the absolute gap to narrow in time, and it adds a second axis the growth literature leaves out: the environmental cost of the route taken. The record below shows how much of the story each measure captures.

A real but concentrated catch-up. Average income in the Global South rose from 13% of the high-income level in 2000 to 23% in 2024. Most of that came from China and India: without them the figure moved from 18% to 20%. Upper-middle-income economies went from 18% to 36% and lower-middle from 9% to 15%, while low-income countries fell from 4.2% to 3.7%. The share of developing economies growing faster than the rich rose from 39% in the 1990s to 75% in 2000–2010, then fell to 63% in 2010–2019 and 49% in 2019–2024, a period that included the pandemic, price shocks and a sharp rise in borrowing costs.

Relative progress hides absolute distance. Even as the ratio rose, the dollar gap between the average person in the Global South and in a rich country widened from about $35,800 to $44,700 a year. Of 102 developing economies with data, 90 became richer in absolute terms, yet the absolute gap widened for 99 of them, and 24 became richer while falling further behind in relative terms. Absolute improvement is not convergence. This is why the concept is stated in terms of narrowing disparities, and why the pace of the next phase matters more than the direction of the last.

Dimensions have not moved together. Measured against high-income economies of the same year, the average developing economy's technological score rose from 9 to 50 between 2000 and 2022, economic from 48 to 65, human development and infrastructure from 57 to 74 and institutional from 57 to 66. The environmental score hardly moved, from 47 to 49: the one dimension for which the benchmark is a sustainability threshold rather than rich-country practice. Resilience is measured only for 2026. The chart shows each dimension's relative position then and now.

The ledger since 2000. The number of people without electricity in developing economies fell from 1.32 billion to 0.69 billion, without clean cooking from 3.14 billion to 2.09 billion, without basic water from 1.10 billion to 0.71 billion, without basic sanitation from 2.66 billion to 1.52 billion and without internet from 4.82 billion to 2.59 billion, in each case while the population grew. Internet use in the average developing economy went from about 1% to 60%. These are the fastest convergences in the record, and they are the ones with the clearest delivery routes.

What is still behind. Two billion people still cook with polluting fuels and 0.69 billion have no electricity, most of them in Sub-Saharan Africa (IEA et al. 2026; IEA 2025). Where access has barely moved, population growth has absorbed the gains: under each country's own trend, the number without clean cooking in 2035 (1.58 billion) is no lower than in 2030 (1.56 billion). The next decade is a race between access gains and demographic growth, which is why the pace per country, not the global average, is the quantity to track.

The proven pace. For each service the page takes every developing economy that was below 85% access in 2010, computes its annual percentage-point gain to 2022 and reads off the median, the 75th percentile (the proven pace, reached by a quarter of peers) and the 90th. It then asks what happens by 2030 and 2035 if every country below 95% moves at the larger of its own pace and the proven pace. The gap between that and the trend is the pace gap, here expressed in people. These are mechanical extrapolations from past paces and UN population projections, not forecasts.

The record. Between 2000 and 2022, 48 of 100 developing economies, with 78% of the population analysed, converged on a carbon-intensive path (income caught up on the high-income average while CO₂ per person grew at least half as fast as income), among them China, India, Indonesia, Bangladesh and Ethiopia. Only 17 economies with 6% of the population converged with slower or falling emissions, and some, such as Nigeria, owe the fall partly to less gas flaring. Thirty-five economies did not converge at all. Air pollution rose in 39 of the 100. This is not an indictment: most began from very low emissions, and coal was often the cheapest way to expand power quickly.

Latecomers do better. Countries that reached $10,000 of income per person before 1970 emitted a median 4.4 tonnes of CO₂ per person when they crossed that line (19 countries); those that crossed after 2000 emitted 2.9 tonnes (25 countries). Across all 69 crossings the line falls by about half a tonne per decade, though with wide scatter (R² 0.13): the latecomer advantage is real and far from automatic. It is the empirical basis for the idea of a leapfrog decoupling frontier, that developing economies can industrialise, electrify and urbanise without the historical surge in emissions (Hughes 2026a).

Decoupling is possible. Between 2010 and 2022, 34 of 45 growing high-income economies cut their territorial CO₂ emissions, against 12 of 91 developing ones. Solar modules are now about 1,100 times cheaper than in 1975 and new solar and wind are among the cheapest electricity ever built (Our World in Data 2025b; Our World in Data 2025c); the binding constraint on clean catch-up has shifted from the price of technology to the cost of capital, which for clean energy in developing economies is often two to three times that in rich ones (IEA 2024a). Costa Rica shows the path can be taken early: it halted deforestation in the 1980s and built a payments-for-ecosystem-services programme from 1997 (UNEP 2019).

How it is built. Seventeen public indicators, scored 0 to 100 as distance to a benchmark: for economic, technological, human development, infrastructure and institutional indicators the benchmark is the high-income median (for example life expectancy 80.5 years, internet use 92%); for the environment it is a sustainability threshold, because rich countries are not a sustainable model (WHO air-quality guideline of 5 µg/m³, about 2 tonnes of CO₂ per person, low carbon intensity, fully renewable electricity); resilience is the INFORM vulnerability and lack-of-coping scores, benchmarked to the high-income median. The six dimension scores (human development includes infrastructure) are combined by a geometric mean. Economies need at least half the indicators in each dimension.

What it shows. Scores rise with income; income explains 81% of the variation among developing economies. The largest gaps are in technology (low-income 10 against high-income 96) and the economy (30 against 98). The environment is the exception: low-income economies score highest (69) because they emit little and generate much of their power from hydropower, upper-middle-income economies lowest (44) and high-income economies only 50. No income group is yet on a path that combines high capabilities with environmental sustainability, so environmental convergence has to come from both ends.

Convergence and sustainability. Capabilities and sustainability move together overall (correlation 0.50, because resilience rises strongly with capabilities, 0.88) but capabilities and the environment pull in opposite directions (−0.43). Splitting 103 developing economies at the median of each gives four groups: 36 sustainable convergers, 35 trapped, 16 with capability without sustainability (economies dependent on coal, oil or gas) and 16 sustainable, not yet converging: low-emitting, largely rural economies with the chance to grow without ever becoming high emitters.

The hypothesis. The original Sustainable Convergence monograph describes a Compounding Convergence Flywheel: simultaneous investments in clean energy, digital infrastructure, human capital and climate resilience strengthen one another, giving a combined result greater than the sum of sectoral efforts (Hughes 2026a). Once a rural district has clean reliable power, telecom towers run without diesel, clinics refrigerate vaccines, schools use lights and computers and farmers run mills and cold stores; each raises incomes, which raises demand for power. The idea has roots in balanced growth, the big push and multisolving; the monograph's combined multiplier of 2.35 to 3.25 times for integrated over siloed investment is a hypothesis, not a finding.

What the evidence supports. Place-based studies find complementarities: electrification in rural South Africa raised women's employment (Dinkelman 2011); fast internet raised employment in African areas connected first (Hjort and Poulsen 2019); mobile money helped Kenyan households cope with shocks and lifted many out of poverty (Jack and Suri 2014; Suri and Jack 2016); early nutrition raised adult earnings (Hoddinott et al. 2008). They also show complementarities must be present: grid connections in rural Kenya produced few measurable gains in the short run because households lacked appliances, skills and markets (Lee et al. 2020). One dimension does not pull the others along automatically; the links work when the pieces are put in place together.

The national test. Most pairs of dimensions move together positively, weakly: the average partial correlation is +0.06 and 8 of the 15 pairs are positive. The strongest links are technology with institutions (+0.33) and human development with infrastructure (+0.22) and institutions (+0.22). Faster gains in capabilities were not associated with environmental deterioration; if anything the reverse (+0.17). If gains were unrelated, about 11% of economies would beat the median on five or six of the six dimensions by chance; 14% did, while 17% advanced quickly on at most one.

Reading it honestly. Bundling exists but is far from the strong, automatic compounding a literal flywheel implies. National averages over two decades can hide strong local complementarities, and the indicators available for every country are crude. The evidence neither confirms nor rules out multipliers of the size proposed; it suggests complementarities are real, uneven and dependent on design. The practical implication is integrated planning: power projects planned with telecom operators, clinics built with electricity, schools with water.

Five high-leverage combinations. The table lists where the evidence for complementarity is strongest. Coordinated investment also has a fiscal logic: clean power cuts fuel import bills and exposure to price shocks, digital systems widen the tax base and cut leakage, healthier populations pay more tax and need less emergency spending, and resilient infrastructure cuts repair costs after disasters. The savings can finance further investment: the macroeconomic version of the flywheel.

The model. Starting from Global South income of $13,675 per person (2021 PPP) and 23.9 Gt of CO₂ a year, high-income income of $58,377 growing 1.2% a year and UN medium-variant population projections, the Global South closes a share λ of the income gap each year on top of the 1.2% base growth, and its carbon intensity (CO₂ per dollar) falls at a rate you set. High-income emissions (a fixed 179 Gt of CO₂ over the century) fall linearly to zero by 2050. Cumulative emissions are compared with the roughly 1,050 Gt left for a 50% chance of 2°C and 130 Gt for 1.5°C from 2025 (Forster et al. 2025).

The three presets. Stalled convergence (λ = 0): no catch-up, intensity falling 2% a year; cumulative world emissions about 1,830 Gt. Fossil catch-up (λ = 2%): the Global South reaches 74% of high-income income by 2100 with intensity falling only 1.2% a year until 2050, then 4%; about 3,080 Gt. Sustainable convergence (λ = 2%): the same income gains with intensity falling 4% a year, ramping to 10%; about 820 Gt, inside the 2°C budget. The difference between the second and third is the entire value of the clean route.

What it cannot say. It is a transparent accounting, not a forecast: emissions per dollar are assumed to fall on a schedule, not derived from energy-system or cost models, and the 1.5°C budget (about 130 Gt) is already beyond reach in every path. Income convergence is assumed to follow a constant-share rule. Intensity falls of 4 to 10% a year are fast but have been observed in rich economies for individual decades. The model shows what has to be true, not that it will be: the feasibility case rests on falling clean technology costs, capital costs and policy, discussed in section 8.

Make the cheapest path the cleanest. The cost of capital, often two to three times higher for clean energy in developing economies than in rich ones, is now the main wedge; guarantees, local-currency lending and better project preparation narrow it (IEA 2024a). Finish the access agenda with distributed, low-carbon technology: universal electricity, clean cooking, safe water and sanitation underpin every other dimension. Build digital public infrastructure: open identity, payment and data-exchange systems lower the cost of reaching people and widen markets (UNDP 2023).

Invest early in people: nutrition in the first 1,000 days, primary care and foundational learning pay returns for decades (Hoddinott et al. 2008; World Bank et al. 2022). Build resilience in from the start: infrastructure designed for the future climate, early warning, adaptive social protection and disaster finance protect gains from shocks (Hallegatte et al. 2019; Global Commission on Adaptation 2019). Measure and learn: track the six dimensions and the pace gap annually, publish country scores with uncertainty ranges and test the flywheel with comparison groups.

Limits and honest nulls. The scorecard is a first version: indicators are those available for nearly every country, so AI adoption, education quality, biodiversity, water stress and financial resilience are missing; resilience is measured once (2026) so it cannot yet be tracked over time; environmental scores use territorial emissions, which flatter rich importers; benchmarks and floors involve judgement. The latecomer effect is weak (R² 0.13), the national flywheel evidence modest and the pace scenarios mechanical. Grid connection in rural Kenya showed little short-run effect without complements. Report what fails as prominently as what works.

Data and method. Income, access, health, governance, emissions and air-quality indicators: World Bank World Development Indicators and Worldwide Governance Indicators (2000 to 2024, interpolated within series and carried forward at most three years); CO₂ and solar and electricity cost series: Our World in Data; resilience: the INFORM Risk Index 2026 (vulnerability and lack of coping capacity); populations and projections: UN World Population Prospects 2024 medium variant. Developing economies are low-, lower-middle- and upper-middle-income economies in the current World Bank classification; the index covers economies with more than one million people. The pace scenario uses developing economies below 85% access in 2010 to set the median, 75th and 90th percentile paces (2010–2022), then projects each economy forward at its own pace (trend) or at the larger of its own and the 75th-percentile pace if it is below 95% (proven pace), multiplied by projected population; negative own paces are kept. The pathways model reproduces the November 2026 report's accounting; the presets match its published results.

Cite. H Heuristics (2026). Sustainable Convergence: the H Heuristics concept. H Heuristics Digital Reports, 2026-88. Cite the concept as "Sustainable Convergence (Hughes, 2026)", a concept developed by Hunter Hughes, Founder of H Heuristics, and cite evidence by the report that supplies it. The underlying report is Hughes, H. (2026), Sustainable Convergence, H Heuristics Research Report, November 2026 (researchreports.hheuristics.com). Country boundaries: Mike Bostock's world-atlas.