Healthcare Convergence in the Developing World

Across today's developing countries, life expectancy rose from 43.5 years in 1960 to 65.5 in 2000 and 72.0 in 2024; in high-income countries, from 68.4 to 80.4. The gap fell from 24.9 years to 10.5 by 2000 and 8.4 by 2024. Low-income countries started lowest and remain furthest behind, at 65.2 years, but have converged fastest since 2000, adding 11.7 years against high-income countries' 4.4, closing their gap from 22.5 to 15.1 years. The spread of life expectancy across all countries (its standard deviation) fell from 11.7 years in 1960 to 7.1 in 2023.

Convergence has been "β-convergence" in every decade: countries that started with lower life expectancy gained faster. The relationship has strengthened recently, with the correlation between starting level and subsequent gain reaching −0.45 in both the 2000s and the 2010s. But average gains have shrunk as the gap narrowed: the gap closed by 14.4 years in 1960–2000, about 0.36 years a year, and by 2.2 years in 2000–2024, about 0.09 a year. COVID-19 interrupted progress: developing countries lost 1.7 years of life expectancy between 2019 and 2021, more than high-income countries, but had recovered by 2022.

Child mortality tells a sharper story. Under-5 deaths in developing countries fell from 223 per 1,000 births in 1960 to 69 in 2000 and 29 in 2024. But rich countries' rates fell proportionally faster for four decades, so the ratio rose from 5.0 in 1960 to a peak of 7.7 in 2000, before falling to 6.0 in 2024. Low-income countries' rate, 62 per 1,000, is still 13 times the high-income rate of 4.8. Maternal mortality in developing countries fell from 360 per 100,000 births in 1990 to 133 in 2023, still 15 times the high-income rate.

The UHC service coverage index combines 14 tracer indicators across reproductive, maternal and child health, infectious disease, non-communicable disease and service capacity. Globally it rose from 54 in 2000 to 68 in 2015, almost a point a year, then to only 71 by 2023, about 0.4 points a year. WHO and the World Bank project 74 by 2030, far short of the SDG ambition of universal coverage. The early gains came largely from campaigns against infectious disease and for child health; the non-communicable disease component is now the lowest-scoring and slowest-moving.

Low-income countries' index rose from 35 to 47, upper-middle-income countries' from 69 to 81, and high-income countries' from 78 to 85. The gap between the poorest and richest groups has narrowed only from 43 to 38 points. Dispersion across countries has fallen, from a standard deviation of 20.5 points in 2000 to 15.4 in 2023, but 49 countries with 1.8 billion people still score below 60.

The fastest movers show what is possible. Rwanda raised its index from 18 to 59 and Nepal from 18 to 58; India from 31 to 69; Cambodia, Lao PDR, Bangladesh, Sierra Leone, Uganda, Zambia and Tanzania each by about 30 points. Most did it through primary care and community health workers, national insurance or tax-funded schemes, and sustained investment in maternal and child health, often with substantial donor support.

Spending per person has risen three- to four-fold in every income group since 2000. In the median low-income country it rose from $43 to $128 in purchasing-power terms, in lower-middle-income countries from $80 to $275 and in upper-middle-income countries from $302 to $1,169; in high-income countries, from $1,451 to $4,266. In relative terms the gap has barely moved, from 1:34 to 1:33; in absolute terms it has nearly tripled. In US dollars, the median low-income country spends $38 per person a year on health, all sources combined.

Across countries, life expectancy rises with the logarithm of spending: on average, each doubling of spending per person goes with 3.0 more years of life expectancy and 6.9 more points of service coverage. Spending accounts for about three quarters of the variation in both. The relationship is steep at low spending and flat at high spending, which means a dollar buys most health where spending is lowest, and that the marginal value of money moving towards the poorest systems is very large.

The relationship is an association, not a recipe. Income, education, water and sanitation, the HIV epidemic and conflict all shape outcomes alongside health spending. But the scatter shows wide variation at every level of spending: countries at $500 per person range from life expectancies in the low 60s to nearly 78. How money is spent, and who pays, matters as much as how much.

Comparing each country with what its spending predicts picks out the "good health at low cost" systems. Sri Lanka, with a life expectancy of 77.7 on $507 per person, is 7 years above the line; it has run a tax-funded public system with free primary and hospital care for seven decades. Thailand covers 82% of essential services, 13 points more than its spending predicts, while households pay just 10% of costs, the result of its universal coverage scheme of 2002 built on a rural district health network. Viet Nam, Kyrgyzstan, Tajikistan, Algeria, Jordan, Tunisia, Iran, Ecuador and Nicaragua also outperform on both life expectancy and coverage.

Among the poorest, India achieves coverage 11 points above its spending level, and Malawi, Uganda, Lao PDR and the Gambia outperform on both measures; Rwanda, with households paying just 4% of costs through community-based health insurance, is 7.5 points above on coverage. Eighteen developing countries outperform on both counts. (Countries where measured spending collapsed in conflict or crisis, such as Venezuela, Syria and Sudan, are excluded: their outcomes reflect past investment, not current spending.)

At the other end, Nigeria's life expectancy of 54.6 is 13 years below what its spending of $260 per person predicts, and Chad, the Central African Republic and South Sudan are 7–10 years below. A group of southern African countries, South Africa, Botswana, Namibia, Eswatini and Lesotho, spends relatively generously but is 5–12 years below the line, the long shadow of the HIV epidemic. These cases show that spending more is necessary but not sufficient: what matters is spending on effective services that reach everyone.

How a health system is paid for decides whether it protects people. Where costs are paid out of pocket at the point of care, illness becomes a financial shock: households sell assets, borrow, cut food or skip treatment. In the median low-income country, households pay 46% of all health costs out of pocket, down from 56% in 2000; in lower-middle-income countries, 38%; in high-income countries, 17%. Across countries, the out-of-pocket share is closely tied to financial hardship (correlation 0.47): in Bangladesh, where households pay 79% of costs, a quarter of people spend more than a tenth of their budget on health.

Globally, the number of people spending over 10% of their budget on health rose from 588 million (9.6%) in 2000 to 1.04 billion (13.5%) in 2019. The rates are highest in middle-income countries (15–16%); in low-income countries they are lower, 7.4%, not because care is affordable but because many people cannot pay and go without. Financial protection has been moving in the wrong direction even as coverage expanded.

The most striking trend is in who funds the poorest systems. In the median low-income country, the government's share of health spending fell from 34% in 2000 to 20% in 2023, while external aid rose from 9% to 28%; government health spending is just 0.9% of GDP and 5.5% of the government budget, against the 15% African governments pledged in the 2001 Abuja Declaration. Of developing countries with 1 million people or more, 14 meet that benchmark, and South Africa is the only one in Africa. Aid has bought enormous gains, but it is volatile, often disease-specific, and now falling, which makes domestic public financing the foundation that convergence needs.

The third edition of Disease Control Priorities defined an essential universal health coverage package of 218 cost-effective interventions, and a highest-priority subset of 115. Watkins and colleagues estimate that delivering the full package at 80% coverage would cost $79 per person a year in low-income countries and $130 in lower-middle-income countries, in 2016 dollars (about $100 and $165 today), requiring additional investment of 8.0% and 4.2% of national income; the highest-priority package would cost about half as much. About two thirds of the cost is for mortality-reducing interventions, and the largest share for non-communicable disease and injury, the conditions current systems handle least well.

Against these costs, current spending falls far short. In the median low-income country, total health spending from all sources is $38 per person, and in the median lower-middle-income country $85. Of 66 low- and lower-middle-income countries, 53, home to 3.4 billion people, spend less in total than the package would cost; the combined shortfall is roughly $285 billion a year. In the median low-income country, the full package would cost 11% of GDP, more than twice what the country currently spends on health from all sources, and twelve times what its government spends.

This is why affordability is a sequencing question. A highest-priority package of maternal, child, infectious-disease and core chronic-disease care costs roughly half as much and is within reach of many lower-middle-income countries that raise public spending towards 3% of GDP. For most low-income countries, even the priority package requires both more domestic revenue and continued, better-aligned external support for a decade or more. The simulator in section 9 shows how long it would take a country to afford each package under different choices.

In 2013, the Lancet Commission on Investing in Health argued that a "grand convergence" in health, bringing infectious, child and maternal deaths in poorer countries down to the levels of the best-performing middle-income countries, was achievable within a generation. Its 2024 successor report set a goal of halving premature death by mid-century. The data show both the promise and the problem. Many countries have converged at historic speed; Malawi, Rwanda, Zambia and Uganda have added around 20 years of life expectancy since 2000.

But on current trends, the targets are slipping away. Sixty countries remain above the SDG target for child mortality, 63 above the maternal mortality benchmark, and 49 below 60 on service coverage. The life-expectancy gap closed by 2.2 years in 24 years; at that pace, closing the remaining 8.4 years would take about 90 years. The slowdown is partly arithmetic, since the easiest gains come first, but it is also financial and institutional: coverage stalled after 2015, government health spending in low-income countries fell as a share of budgets, and aid is now falling.

The faster convergers point the way: countries that sustained public investment in primary care and community health, pooled financing to reduce out-of-pocket payments, and steadily expanded a defined package. Convergence does not require rich-country spending. It requires that the money a country has, domestic and external, flows through public, pooled channels to effective, prioritised services that reach everyone.

The atlas shows each country's latest life expectancy, its gain since 2000, under-5 mortality and the UHC service coverage index, and its health spending per person in purchasing-power dollars.

The "health per dollar" layer shows years of life expectancy above or below what the country's spending predicts. Countries flagged as in crisis, where measured spending collapsed, are shown but should be read with caution. The out-of-pocket layer shows the share of health spending paid directly by households.

Hover over or tap a country for its card, which shows outcomes, coverage, spending, financing and package costs. The sortable table lists all 160 countries with at least 1 million people; filters pick out the good-health-at-low-cost group, the aid-dependent and the household-financed.

The simulator starts from each low- or lower-middle-income country's current government and external health spending per person (US dollars) and its GDP per person. It raises government health spending in a straight line from today's share of GDP to the target share over ten years, grows GDP per person at the chosen real rate, and keeps external funding at today's level, halves it, or phases it out over ten years.

It then compares pooled public funding (government plus external) with the cost of the highest-priority package (about $50 per person in low-income countries and $83 in lower-middle-income countries, 2023 dollars) or the full essential package ($100 and $165). Package costs are held constant in real terms; in reality they would rise somewhat with wages, and coverage would expand gradually.

The result is the first year in which pooled funding covers the package, and the gap per person today. It is a stylised affordability calculation, not a forecast: it ignores efficiency gains, cost escalation, and the private spending that would shift to the public system as coverage grows.

Pay through public, pooled channels. Every high performer finances most care through taxes or mandatory insurance pooled across the whole population, so that the healthy subsidise the sick and the rich the poor. Raising government health spending towards 3% of GDP, from 0.9% in the median low-income country, and reducing out-of-pocket payments for priority services, protects households and lets systems plan. Health taxes on tobacco, alcohol and sugary drinks raise revenue and reduce disease at the same time.

Build from primary care and define the package. Sri Lanka, Thailand, Rwanda and Ethiopia built networks of primary care and community health workers before expanding hospital care. Explicit, costed benefit packages, starting with the highest-priority interventions, let countries spend scarce money on what does most good, and expand coverage step by step. Strategic purchasing, paying providers for services delivered and outcomes, not just inputs, gets more care per dollar.

Invest in people and medicines, and make aid build systems. Health workers are the binding constraint in many countries: WHO projects a shortfall of about 11 million by 2030, concentrated in low- and lower-middle-income countries. Pooled procurement and essential-medicines lists cut drug costs. Digital health records, payments and data systems help small budgets reach further. And external aid, now falling, should move from parallel disease programmes towards co-financing national systems, with clear paths to domestic ownership.

Outcomes: World Bank WDI life expectancy at birth (SP.DYN.LE00.IN), under-5 mortality (SH.DYN.MORT) and maternal mortality (SH.STA.MMRT), 1960–2024, population-weighted by current World Bank income group (developing = low, lower-middle and upper-middle income). σ-convergence: standard deviation across countries; β-convergence: slope of annual gain on starting level by decade. Coverage: WHO UHC service coverage index (UHC_INDEX_REPORTED), 2000–2023, global and income groups as published. Financial hardship: WHO SDG 3.8.2 (FINPROTECTION_CATA_TOT_10), global and country.

Spending: WHO Global Health Expenditure Database via WDI: current health expenditure per person (PPP and US$), and shares from government, external sources and out of pocket; medians by income group. Health per dollar: residual from a regression of life expectancy (and the UHC index) on the log of spending per person (PPP) across 1-million+ countries; crisis-affected countries (spending down 40%+ since 2015, or in conflict) are flagged. Package costs: Watkins et al. (2020), $79 and $130 per person (2016 US$) uplifted to 2023 dollars by US CPI (×1.27); the highest-priority package is taken as half.

Caveats: aggregates use current income classifications, so groups are not fixed sets over time; spending data have gaps and revisions, and US-dollar comparisons move with exchange rates; regressions are associations, not causal estimates; the package comparison sets total current spending, which includes services outside the package, against a package cost at 80% coverage; the simulator is stylised. Life expectancy reflects many factors beyond health systems.