Global Survivability Under Compounding Twenty-First-Century Risks

A simple measure of survivability is the share of a birth cohort expected to survive to age 65 at current death rates. Worldwide it is 82.5% for women and 74.0% for men. Across countries it ranges from under 50% in Lesotho, Nigeria, the Central African Republic and Chad to over 93% in South Korea, Qatar and the United Arab Emirates. A child born in the lowest-survival countries is less likely than not to reach the age at which a child born in the highest-survival countries is likely to retire.

Child mortality shows the same geography in sharper relief. The under-five mortality rate ranges from under 3 per 1,000 live births in much of Europe and East Asia to about 100 or more in Nigeria, Niger, Somalia and Chad. Sub-Saharan Africa accounted for 60% of all under-five deaths in 2024 in the World Bank series. Maternal mortality is even more concentrated: the UN estimates that sub-Saharan Africa accounted for 70% of the 260,000 maternal deaths in 2023, and that countries in conflict or fragility accounted for 64%.

The avoidable-deaths layer applies the benchmark of section 5 to each country: the under-five deaths each year that would not occur if the country had the mortality rate of the best quarter of its own income group. It is not a map of blame. The largest numbers are in very populous countries — Nigeria, Pakistan, the Democratic Republic of the Congo and India together account for more than half — and the highest rates in small, fragile states. Hover over a country for its profile.

Measured over 2000–2024, the headline shocks of the century have each killed on a vast scale. The Uppsala Conflict Data Program records about 1.8 million deaths in state-based armed conflicts, with peaks in the Syrian civil war, the Ethiopian war in Tigray and the war in Ukraine; organised violence of all kinds killed an estimated 244,600 people in 2025 alone, the highest in decades, driven by violence against civilians in Sudan. The EM-DAT database records about 1.6 million deaths from disasters, dominated by a few catastrophes such as the 2004 Indian Ocean tsunami, the 2008 Cyclone Nargis, the 2010 Haiti earthquake and the 2023 Türkiye–Syria earthquake.

The pandemic was larger still. The WHO estimates 14.9 million excess deaths in 2020 and 2021, and in the World Bank series global life expectancy fell from 72.9 years in 2019 to 71.2 in 2021 before recovering to 73.5 in 2024. Yet over the same 25 years, about 174 million children died before their fifth birthday and, over 2000–2023, about 8.4 million women died of causes related to pregnancy and childbirth. Chronic child deaths alone were more than ten times the pandemic's excess deaths and about fifty times the deaths in conflicts and disasters combined.

The comparison is not meant to diminish shocks. Shocks and chronic losses are linked: wars, disasters and pandemics destroy the clinics, supply chains and incomes that keep children alive, so a large part of the chronic toll in fragile states is the long shadow of shocks. And other chronic threats are larger again — the State of Global Air report attributes 7.9 million deaths in 2023 to air pollution, and Zhao and colleagues associate about 5 million deaths a year with non-optimal temperatures. The point is where survivability is won and lost: mostly in the everyday systems that compound risks wear down.

Grouping countries by the INFORM risk class shows how sharply losses concentrate where systemic threats compound. The 44 high- and very-high-risk countries hold 24.6% of the world's population, but in 2024 they accounted for 3.37 million of 4.86 million under-five deaths, 69%, and in 2023 for 77% of maternal deaths. The 16 very-high-risk countries, home to a tenth of humanity, accounted for 45% of child deaths and 57% of maternal deaths on their own. The 33 countries in INFORM's compound core — in the top third of hazard, vulnerability and lack of coping capacity at once — account for 64% of child deaths.

The gradient runs through every class. The median under-five mortality rate is 4.7 per 1,000 in very-low-risk countries, 8.1 in low-risk, 24 in medium-risk, 37 in high-risk and 74 in very-high-risk countries — a sixteenfold difference. Median life expectancy falls from 81.4 to 61.8 years, a gap of almost twenty years, and median survival to 65 from 90% to 62%. Part of the gradient reflects income, because INFORM's vulnerability dimension includes poverty, so risk class and wealth are not independent; the benchmark in section 5 therefore compares each country with its own income group.

The concentration matters for strategy. A survivability agenda spread evenly across the world would miss most of the losses; one focused on high-risk countries would reach seven in ten child deaths. And because those countries are where population is growing fastest — the subject of this series' report on human exposure to 2100 — the concentration will deepen unless their mortality rates fall faster than their births rise.

INFORM's lack-of-coping-capacity dimension measures the institutions and infrastructure a country can bring to bear on a crisis: governance, disaster risk reduction, communications, physical infrastructure and access to health care. Plotted against survival to 65, it produces one of the tightest relationships in this series. The rank correlation is −0.78 across 191 countries. The median survival to 65 is 90% in countries with very low lack of coping capacity, 82% in the low class, 77% in the medium, 69% in the high and 63% in the very-high class.

The relationship is not destiny. At any given level of coping capacity, survival varies by ten points or more, and some countries far outperform their peers. Many of them combine high exposure to hazards with deliberate investment in primary health care, community health workers, vaccination and disaster preparedness — the same pattern seen in this series' geography of global vulnerability, where Bangladesh and Rwanda stood out as coping better than their income predicts. The outliers in the other direction are often resource-rich economies whose income has not translated into health systems, or countries where conflict has broken them.

Survivability under compound risk is therefore largely a question of whether the systems that prevent deaths are built and then kept running. A clinic that functions in normal times saves children; a clinic that keeps functioning through a drought, a currency crisis or an outbreak saves many more, because those are the moments when mortality spikes. That is the practical meaning of adaptive capacity for survival, and it is the lever examined in the next two sections.

An avoidable death needs a benchmark. Setting it at the level of the richest countries would count almost every child death as avoidable but would not be achievable soon. This analysis uses a more demanding test of realism: the under-five mortality rate already achieved by the best quarter of each country's own World Bank income group — 44.5 per 1,000 for low-income countries, 18.4 for lower-middle-income, 9.4 for upper-middle-income and 3.3 for high-income countries. A country above its peer benchmark has avoidable deaths equal to its deaths multiplied by the share of its rate above the benchmark.

On that basis, 2.37 million of the 4.86 million child deaths in 2024 — 49% — were avoidable. Meeting the SDG target of 25 per 1,000 everywhere would avert a similar number, 2.26 million; matching the upper-middle-income median would avert 3.3 million, and the high-income median 4.2 million. The avoidable deaths are highly concentrated: 1.81 million (76%) are in high- and very-high-risk countries, and five countries — Nigeria, Pakistan, the Democratic Republic of the Congo, India and Niger — account for 61%. Nigeria alone, with an under-five mortality rate of 116 against a lower-middle-income benchmark of 18, accounts for about 720,000.

For maternal deaths the avoidable share is larger. Against the same income-peer benchmark (maternal mortality ratios of 225, 75, 19 and 4 per 100,000 live births), 147,000 of the 247,000 maternal deaths reported for 2023 at country level — 59% — were avoidable; meeting the SDG target of 70 everywhere would avert 70%. These are the deaths that skilled birth attendance, emergency obstetric care, family planning and functioning referral systems prevent, and they are the deaths that conflict and fragility multiply.

The world has made enormous progress against child mortality this century: the global under-five mortality rate roughly halved after 2000. But the pace has slowed. Among the 190 countries in this analysis, the median annual rate of reduction fell from 3.4% in 2000–2015 to 2.6% in 2015–2024, and the UN reports that the global pace of decline has slowed by more than 60% since 2015. At the same time births are shifting towards the countries where mortality is highest, so a slower decline in rates can mean a stall or even a rise in deaths.

Applying four mortality paths to the UN's projected births gives very different totals for 2025–2050. If every country's rate stayed at its 2024 level, deaths would rise to 5.4 million a year by 2050, and 137 million children would die over the period. If each country continued its 2015–2024 rate of reduction, deaths would fall to 3.2 million a year and the total would be 103 million. If every country reached the best quarter of its income group by 2035 and then continued its trend, the total would be 69 million; reaching the SDG target of 25 by 2030 would bring it to 60 million.

The differences are the stakes. Continuing recent trends rather than stalling saves about 34 million children; reaching the peer benchmark by 2035 saves about 34 million more; reaching the SDG target, about 43 million more than trend. In every path, most deaths remain in today's high-risk countries — between 61% and 76% — because that is where births are concentrated. The tool below shows each path, and lets the year in which the peer benchmark is reached vary.

Four priorities follow. The first is to count the chronic losses as part of the polycrisis. Planning for pandemics, wars and disasters is necessary, but the largest and most avoidable losses of this century are the deaths of children and mothers in places where compound risks have hollowed out basic systems. A survivability agenda that does not centre them misses most of what is at stake. The second is to close the gap to peers: half of child deaths are avoidable at levels that countries of the same income already reach, and the interventions are known and cheap relative to the lives they save.

The third is to shock-proof the systems that prevent deaths. Mortality spikes when shocks hit — when conflict closes clinics, drought brings malnutrition, outbreaks overwhelm wards and currency crises cut drug supplies. Keeping primary health care, immunisation, nutrition and maternal care running through shocks — through contingency financing, pre-positioned supplies, community health workers and social protection that scales up automatically — turns adaptive capacity into survival. The pandemic showed the cost of the opposite: life expectancy fell by 1.6 years, and routine services were disrupted around the world.

The fourth is to concentrate effort where losses concentrate. Seventy-six per cent of avoidable child deaths are in high-risk countries, and five countries account for 61%. Those countries are also where births are growing fastest, and many are affected by conflict, where the UN finds most maternal deaths now occur. Peace, and the protection of health care in conflict, are therefore survival measures in the most literal sense. None of this competes with preparing for the century's great shocks; the same systems that prevent everyday deaths are the ones that limit the toll of shocks when they come.

Under-five deaths and mortality rates, maternal deaths and ratios, survival to 65 (average of women and men) and life expectancy are from the World Bank World Development Indicators (updated 8 October 2026), drawing on UN IGME, the UN Maternal Mortality Estimation Inter-agency Group and UN WPP. Country risk is from the INFORM Risk Index Mid-2026. Projected births are UN WPP 2024 medium-variant figures via Our World in Data. Conflict deaths are UCDP state-based battle deaths and disaster deaths EM-DAT totals, both via Our World in Data; pandemic excess deaths are the WHO estimate for 2020–2021. Maternal deaths summed at country level (247,000) differ slightly from the global estimate (260,000) because of coverage.

Avoidable deaths equal a country's deaths × (1 − benchmark ÷ its rate) where its rate exceeds the benchmark. The peer benchmark is the 25th percentile of the latest rate within the country's World Bank income group (INFORM's classification); alternatives are the SDG targets and income-group medians. Projections multiply projected births by a mortality path, which reproduces 2024 deaths to within 2%. Paths: stall (2024 rate), trend (each country's 2015–2024 average annual rate of reduction, floored at zero), peer (log-linear convergence to the benchmark by the chosen year, then trend) and SDG (convergence to 25 by 2030, then trend).

All cross-country patterns are associations; risk, income, peace and health spending are correlated and not separately identified. Projections are scenarios, not forecasts. The map uses Natural Earth 1:50m boundaries via world-atlas; boundaries do not imply endorsement. Related reports: the geography of global vulnerability, human exposure to the polycrisis and the polycrisis bill.