The Field Guide · No. 54
Life expectancy at birth: what the average counts
Life expectancy at birth is an average over a hypothetical group of newborns, so deaths in infancy and childhood pull it down hard, and it is not a prediction for any one person.
Updated
Life expectancy at birth is an average, and the UN's demographers define it as "the average number of years of life expected by a hypothetical cohort of individuals who would be subject during all their lives to the mortality rates of a given period." The number imagines a group of newborns who live their whole lives under one period's death rates. It summarizes that period. It says little about any particular baby born in it.
Because every birth counts, babies who die in their first year pull the average down hard. The UN's 2024 methodology report describes the first phase of the mortality transition as slow early growth, followed by a period of "accelerated improvement, especially in the mortality of infants and children". It also notes that saving a child's life adds more years to life expectancy than saving the life of an older person. The US life table for 1900 to 1902, which covers ten states and the District of Columbia, shows the effect. Life expectancy at birth was 49.2 years. A child who had reached its first birthday could expect 55.2 more years, six more than a newborn, and a 20-year-old could expect 42.8 more, which takes them to almost 63. A figure near 30 can likewise come from a great many early deaths, and it does not mean that adults were expected to die around 30.
Life expectancy also comes in two versions. Period life expectancy uses the death rates of a single year, or of a group of years, and holds them fixed for the rest of a person's life. Cohort life expectancy follows people born in the same year, using observed death rates for past years and projected rates for the years ahead. The ONS publishes both for the UK, and says period figures tend to be lower. In an explainer last revised in January 2023, it reports a gap of around 8 to 10 years between the two for cohorts born from 1950 to 2000 in England and Wales.
When you meet the number, check three things. Is it at birth or at a later age? Is it a period figure or a cohort figure? Which years does it cover? The ONS tables run by single year of age, from 0 to 100, so a figure at age 50 answers a different question from a figure at birth. The ONS also says cohort figures "do not consider any person-specific factors such as lifestyle choices."
Period and cohort, side by side
Official life tables for past years are generally period tables, because they use observed death rates and give an objective way of comparing trends over time. Cohort figures for past years usually need projected rates as well, which the ONS says makes them involve an element of subjectivity.
| Period | Cohort | |
|---|---|---|
| Death rates used | One year, or a group of years | Observed rates for past years, projected rates for later years |
| Future death rates | Held fixed for the rest of life | Observed and projected changes built in |
| Who the figure describes | The death rates of a given area and time period | People sharing a year of birth |
| Where it is used | Official life tables for past years | Regarded by the ONS as the more appropriate measure of how long a person of a given age would be expected to live |
What to remember
- Life expectancy at birth averages over a hypothetical group of newborns. It is not a prediction for any one person.
- Deaths in infancy and childhood weigh heavily in the average, so a low figure does not mean adults were expected to die young.
- Period life expectancy holds one year's death rates fixed. Cohort life expectancy builds in observed and projected change, so its figure tends to be higher.
- Before comparing two numbers, check whether each is at birth or at a later age, and whether it is a period or a cohort figure.
From the record
Life expectancy at birth: The average number of years of life expected by a hypothetical cohort of individuals who would be subject during all their lives to the mortality rates of a given period.
Asked often
Does a life expectancy of 30 mean most people died at 30?
No. It is an average over everyone born, including babies who died in their first year. In the US life table for 1900 to 1902, drawn from ten states and the District of Columbia, life expectancy at birth was 49.2 years, yet a 20-year-old could expect about 43 more years. A figure near 30 can come from many early deaths, and on its own it does not tell you when the adults in that group were expected to die.
What is the difference between period and cohort life expectancy?
Period life expectancy uses one period's death rates and holds them fixed. Cohort life expectancy follows a birth year, using observed rates for past years and projected rates for later ones. Period figures tend to be lower. An ONS explainer last revised in January 2023 reports a gap of around 8 to 10 years for cohorts born from 1950 to 2000 in England and Wales.
Is life expectancy a prediction for me?
No. It is an average for a group. The ONS says these averages do not consider person-specific factors such as lifestyle choices.
Read the news better, every morning.
We send you the day's real progress in science, medicine and beyond, told plainly and sourced. Free, forever.
Free for good. One click to leave.