ex
Remaining life expectancy
How many additional years does someone who has reached age x expect under a given mortality schedule? It does not adjust for quality of health.
Expected years still to live
Research notes November 2026 Interactive edition
Living longer.
Living better?
We’ve added years to life. How much life have we added to those years? Explore the evidence, the economics, and the very human choices behind longevity.
A life, measured differently01—100
Healthy yearsHealth gap1 dot = 1 year
73.3
Years of life expectancyGlobal · at birth · 2023
62.8
Years in full-health equivalentHealthy life expectancy · 2023
10.5
Years of equivalent health lossLife expectancy minus HALE
WHO Global Health Estimates, released October 2026. Population averages, not a prediction for any individual. Source ↗
01 / The evidence
More years.
Different starting lines.
Longer lives are one of humanity’s great achievements. But where you live still shapes how many years you can expect. Compare six populations across nearly four decades.
Life expectancy at birthBoth sexes · years · 1986–2023
At a glance
2023
19862023
Shaded area: 2020–2021 pandemic years. Select up to six populations.World Bank / UN population estimates ↗
World Bank series updated 8 October 2026. This series and WHO’s headline figures use different estimation methods; values are not interchangeable.
The long view
Global life expectancy rose substantially between 1986 and 2023.
Progress can reverse
The pandemic interrupted decades of gains. A rising long-term trend does not make each new year a certainty.
Averages hide inequality
National averages mask differences in income, sex, healthcare access, and living conditions within each population.
A historical snapshot: sex, income & geography
In WHO’s 2015 estimates, global life expectancy was 73.8 years for females and 69.1 for males. Twenty-nine countries—all high-income—reached at least 80 years, while 22 countries—all in sub-Saharan Africa—were below 60. Africa also recorded a 9.4-year gain from 2000 to 2015. These are historical estimates from the 2016 release, not current rankings. Source ↗
These differences reflect interacting social, biological, economic, and healthcare conditions. They should not be read as an individual forecast or a simple causal effect of income.
Beyond child survival
It’s not just about
surviving childhood.
A survival curve plots the share of a cohort still alive at each age. A curve shifting right indicates better survival to older ages. People who reach older ages have gained years too. In France, a 65-year-old’s expected age at death rose from about 76 in 1816 to 87 in 2023.
The comparisons here are period life-table estimates conditional on reaching an age, not observed birth-cohort curves. Remaining life is not found by simply subtracting age from life expectancy at birth. OWID / Human Mortality Database ↗
Expected age at death · France · rounded
181676 years
202387 years
The healthspan gap
Longer doesn’t always
mean healthier.
Life expectancy counts years. Healthy life expectancy (HALE) adjusts them for time lived in less than full health. By 2023, total years had nearly recovered to 2019 levels. Healthy years were still behind.
WHO’s 2026 release reports 2019 and 2023 estimates. Read the source ↗
62.8healthy-equivalent years
+10.5equivalent years of health loss
020406080 years
Healthy life expectancyLife expectancy − HALE
The gap is a population-level equivalent of health lost. It does not mean that illness is confined to the final years of life.
02 / The metrics
What is a healthy
year worth?
Not all extra years bring the same health benefit. These measures help compare outcomes and interventions. Each answers a different question.
Q
QALY · health gained
A quality-adjusted life year combines duration and health-related utility. Five years at a utility of 0.8 contribute four QALYs.
Years × health utility
D
DALY · health lost
A disability-adjusted life year measures disease burden: years lost to premature death plus disability-weighted years of ill health.
YLL + YLD
One QALY represents one year in full health; one DALY represents one healthy year lost. Their utility systems and reference assumptions differ, so they are not automatically one-for-one equivalents. Definitions ↗
Imagine a treatment adds years of life. Adjust its benefit and extra cost relative to the same comparator.
0.1 year20 years
−0.5 · below death on utility scale1 · full health
$0$300,000
Changing currency loads a new example; it does not convert money. Utility describes a health state, not the value of a person.
A small, hypothetical population: premature deaths and separate nonfatal cases. No discounting or age weighting.
0 · no health loss1 · full health loss
Illustrative incidence-based calculation
The health gain
1.50
additional QALYs per person
1.5 years × 1.00 utility
Incremental cost-effectiveness ratio
$33,333 / QALY
Below the illustrative benchmark range
Compared with ICER’s $100,000–$150,000 per QALY benchmark range.
Simplification: constant utility in added years, no quality change during existing years, and no discounting. A full appraisal also tests uncertainty, harms, equity, and affordability.
The disease burden
30.0
healthy-equivalent years lost
20.0 YLL + 10.0 YLD
How it adds up
For real burden estimates, use appropriate reference life tables and validated disability weights. WHO’s current YLD estimates use prevalence methods; this toy example uses cases × duration × weight.
UK · NICE technology appraisals
£25,000–£35,000 / QALYStandard threshold range implemented in April 2026. Appraisal context and modifiers matter. Source ↗
US · ICER health-benefit benchmarks
$100,000–$150,000 / QALYA value-assessment benchmark, not a single US government funding rule. ICER also uses evLY measures. Source ↗
03 / The choices
Finite budgets.
Human trade-offs.
Healthcare decisions happen at two scales: how societies share limited resources, and how people act on what they know. Neither is settled by a single number.
The allocation sandbox
Where would you put
the next million?
This hypothetical model uses £5,000 per QALY for prevention and £20,000 for treatment. Move the budget split to see the arithmetic.
All treatmentAll prevention
Healthcare budget£1,000,000
Prevention
£500k50% of the budget100.0 QALYs gained
Treatment
£500k50% of the budget25.0 QALYs gained
Total modelled health gain125.0 QALYs
Illustrative assumptions, not empirical prices or a recommended allocation. Returns are linear with no capacity limits, delays, uptake barriers, or discounting. Real prevention is not always cheaper or more effective than treatment. Urgent needs, equity, timing, and the people reached can change the decision.
The other bottleneck is us
Knowing better
isn’t doing better.
Our attention and decisions often favour what feels easier now. Better systems can make useful choices easier to understand and carry out.
Evidence is about associations, not destiny. The studies below do not establish that any single bias causes a particular loss of lifespan.
“I’ll deal with it tomorrow.”
Present bias & procrastination. Immediate effort feels more concrete than a distant benefit. A study following 3,525 Swedish university students linked procrastination to several worse health outcomes nine months later; it was observational. A practical design response: make the next step small and specific. Study ↗
“It probably won’t happen to me.”
Optimism & risk misperception. Feeling safe can make a future risk easy to dismiss. As a decision aid, compare absolute risks and time horizons rather than relying on a reassuring story. This is an illustrative behavioural pattern, not a quantified effect on lifespan here.
“I thought I understood the instructions.”
Health literacy. Confusing information can make care harder to navigate. In a cohort of 3,260 older US adults, inadequate health literacy was associated with higher mortality after adjustment for several factors. Plain language and checking understanding can reduce friction; the study does not by itself prove causation. Study ↗
04 / Follow the evidence
Open the footnotes.
Observed data and illustrative models are labelled separately. Research checked 3 November 2026.
- WHO · Global Health Estimates, 2000–2023 ↗October 2026 release. Headline life expectancy and HALE figures for 2019 and 2023.
- World Bank · Life expectancy at birth ↗Indicator SP.DYN.LE00.IN. Embedded annual series for six populations, 1986–2023; update 8 October 2026.
- Our World in Data · Life expectancy at every age ↗France, 1816 and 2023. Period estimates from Human Mortality Database; rounded values.
- WHO · DALY methods and sources ↗YLL and YLD definitions. See also NICE’s QALY glossary.
- UK Parliament · NICE threshold update ↗Department of Health and Social Care answer, 16 April 2026: £25,000–£35,000 per QALY.
- ICER · Health Benefit Price Benchmark ↗$100,000 and $150,000 per QALY or evLY gained, as defined in ICER’s framework.
- Heart · Cost-effectiveness of statins in older people ↗2024 UK modelling study. Standard and higher-intensity treatment have different comparators.
- Johansson et al. · Procrastination and health ↗JAMA Network Open, 2023. Observational study of Swedish university students.
- Baker et al. · Health literacy and mortality ↗Archives of Internal Medicine, 2007. Prospective cohort of older US adults.
- WHO · Historical progress and inequality ↗May 2016 release. Historical context: rapid gains from 2000–2015 and persistent sex and income gaps.
Data notes, definitions & limitations
What this page measures. Life expectancy is a population average under a mortality schedule. Period life expectancy applies one period’s age-specific death rates to a hypothetical cohort. Cohort life expectancy follows an actual birth cohort’s mortality experience and may require projections for ages not yet observed. Neither is an individual expiry date.
Two sources, kept separate. The country chart uses the World Bank’s annual SP.DYN.LE00.IN series (both sexes), downloaded 3 November 2026. The headline and HALE comparison use WHO’s October 2026 release. Source revisions and modelling choices can produce different estimates for the same year. The line chart’s vertical scale begins at 35 years; it does not begin at zero. Lines connect annual estimates. No post-2023 projection is shown.
Reading healthy years. HALE weights years for health loss across a population. Life expectancy minus HALE is an equivalent loss, not the number of years every person spends ill or the location of those years in a life.
Calculator assumptions. QALYs are incremental added years multiplied by a constant utility, with no effect on the comparator’s existing years and no discounting. Utilities may be negative for health states valued below death. For nonpositive QALY gains, the cost-per-gain comparison is not reported. DALYs use a simplified incidence example with distinct premature deaths and nonfatal cases. The allocation sandbox uses invented constant costs per QALY; it is not a forecast.
What numbers leave out. A favourable ratio does not guarantee funding. Budget impact, evidence quality, distribution of benefits, disability rights, ethical concerns, and access all matter. This is an educational explainer of population metrics, not a personal medical assessment.