Life expectancy is a statistical measure of the average number of years a person is expected to live under a specified mortality pattern. The most commonly quoted measure is life expectancy at birth, which represents the average number of years a hypothetical newborn would live if the age specific mortality rates observed during a particular period remained constant throughout that person's lifetime. It is therefore a period based demographic statistic rather than a personalised prediction of an individual's lifespan.
Life expectancy can also be calculated at other ages, such as age 1, age 20, age 50 or age 65. Life expectancy at birth is strongly affected by deaths occurring during infancy and childhood, whereas life expectancy at older ages provides information about survival among people who have already reached those ages. The measure is normally calculated separately for males and females because mortality patterns differ between the sexes in most populations.
Life expectancy is derived from life tables, which organise mortality information by age and calculate the probability of surviving from one age to another. A life table typically begins with a hypothetical cohort and applies observed age specific death rates to determine how many members of that cohort would survive to successive ages. The resulting survival schedule can then be used to calculate the average number of years remaining at different ages.
The calculation does not require following an actual group of people from birth until death. Instead, it applies the mortality rates observed during the relevant period to a synthetic population. This distinction explains why life expectancy can change relatively quickly when mortality conditions change. If deaths from a major disease decline or infant mortality falls substantially, the period life expectancy estimate can rise even though the underlying biological lifespans of individuals have not suddenly changed.
Life expectancy differs from maximum lifespan, average age at death and individual longevity. Maximum lifespan refers to the longest observed or potentially achievable human survival, while life expectancy is a population average based on mortality probabilities. The average age at death in a particular year can be influenced strongly by the age structure of the people who die and is not necessarily equivalent to life expectancy at birth.
An individual's expected lifespan also depends on personal circumstances that a population statistic does not directly predict, including health, occupation, behaviour, socioeconomic conditions and medical history. Life expectancy therefore describes mortality conditions for a population rather than determining how long any particular person will live. A person born in a population with a life expectancy of 80 years is not individually expected to die exactly at age 80. Some people die much earlier, while others survive substantially longer.
Life expectancy has increased dramatically in many parts of the world over the last two centuries, although the timing and magnitude of improvement have differed between countries and periods. Major contributors have included reductions in infectious disease, improved sanitation and clean water, better nutrition, vaccination, antibiotics, safer childbirth, improved medical treatment, reduced occupational hazards and improved living conditions.
Declines in infant and child mortality have been particularly important historically because deaths at very young ages exert a large effect on life expectancy at birth. More recently, reductions in mortality from cardiovascular disease, some cancers and other chronic conditions have contributed to longer survival in many populations. Improvements are not necessarily continuous. Wars, epidemics, pandemics, famines, environmental disasters, health crises and changes in disease patterns can temporarily reduce life expectancy.
Life expectancy can also reveal inequalities within populations. Average national life expectancy can conceal differences associated with sex, income, education, occupation, geography and other social and demographic characteristics. People living in different regions of the same country can experience substantially different mortality risks and therefore different life expectancies. The measure is consequently useful for population comparison but should not be interpreted as evidence that all members of a population experience the same health conditions.
International comparisons also require attention to the period covered, data quality, definitions and methods. A life expectancy estimate may be based on observed deaths and population estimates, while some estimates for countries with incomplete registration systems rely on demographic models. Differences in statistical methods can therefore affect reported figures.
Life expectancy is closely connected to health policy and population ageing because longer survival changes the number of people reaching older ages. An increase in life expectancy at birth can result from lower mortality at young ages, older ages or both. Increasing life expectancy at age 65 is particularly relevant to pensions, social care and retirement planning because it indicates how long people who have already reached retirement age may expect to live under prevailing mortality conditions.
Longevity is not identical to healthy life expectancy. Healthy life expectancy attempts to estimate the number of years expected to be lived in good health or without specified disability, depending on the definition and measurement method. A population can therefore experience increasing life expectancy without experiencing an identical increase in years lived free from illness or disability.
Life expectancy statistics have important limitations. Period life expectancy assumes that the mortality rates measured during the reference period remain constant, which rarely happens over an actual lifetime. Cohort life expectancy instead follows or projects mortality conditions for an actual birth cohort and can therefore incorporate changing mortality rates, although complete cohort calculations require data or assumptions about future mortality. Estimates can also be affected by incomplete death registration, uncertain population denominators, migration and statistical modelling.
Temporary events can have large effects on period life expectancy without permanently changing long term survival trends. Consequently, life expectancy should be interpreted with its reference year, population, sex, geography, measurement method and definition in mind rather than treated as a timeless characteristic of a country.
Life expectancy is one of the central indicators used in demography because it summarises mortality conditions into a single interpretable measure while remaining connected to detailed age specific death rates. Rising life expectancy generally indicates improved survival, but the reasons for change must be investigated through age specific mortality patterns. A rise caused mainly by reduced infant mortality has a different demographic interpretation from a rise caused mainly by declining mortality among people aged 70 and over.
Life expectancy is therefore both a headline population indicator and an output of a much more detailed mortality system. It is useful in population projections, health planning, actuarial calculations, international comparisons and studies of demographic transition, but it should always be distinguished from an individual's personal forecast of lifespan.
