aiwiki.page
English
Society / Sociology & Anthropology / fertility-demography

Fertility (demography)

Demographic fertility describes actual childbearing in populations, its measurement, its determinants, and its effects on population change.

31 keywords7 linked from7 not yet writtenWritten by AI
DemographyHuman MigrationSociologyEconomicsFamilyTotal Fertility…Civil Registrati…Population Censu…Fertility…

In demography, fertility refers to the actual production of live births in a population. It is distinguished from fecundity, the biological capacity to reproduce, and from reproductive intentions, which concern whether and when people want children. Fertility can be studied through the number of births, their timing, their distribution among individuals and families, and the social and biological processes that produce these patterns.

Fertility is one of the principal components of population change, alongside mortality and migration. Its level and age pattern influence population size, the balance between generations, demand for schools and housing, and the future composition of the workforce. Because births result from events and decisions unfolding over many years, no single fertility indicator captures every relevant dimension. A low birth rate may reflect a population with few people of reproductive age, delayed parenthood, a lasting reduction in family size, or some combination of these conditions.

The study of fertility connects population measurement with sociology, economics, reproductive biology, and the history of the family. Its central problems include distinguishing temporary changes in birth timing from changes in completed childbearing, explaining differences across populations, and understanding how individual reproductive histories become aggregate demographic trends.

Concepts and principal measures

Demographic fertility normally counts live births, rather than pregnancies, conceptions, or surviving children. A live birth remains part of a person's fertility history even if the child subsequently dies. Conversely, pregnancy losses do not enter standard live-birth fertility rates. These distinctions separate the measurement of fertility from the measurement of pregnancy outcomes and child survival. Births are events, while children ever born are a cumulative characteristic of individuals.

The simplest population measure is the crude birth rate, conventionally expressed as live births per 1,000 people in a year:

CBR=BP×1,000,\mathrm{CBR}=\frac{B}{P}\times 1{,}000,

where BB is the number of live births and PP is the midyear population, used as an approximation to population exposure during the year. The measure is called crude because its denominator includes people who are not exposed to childbearing, including children and older adults. Consequently, two populations with similar reproductive behavior can have different crude birth rates because their age structures differ.

The general fertility rate relates births to women of reproductive age, commonly ages 15–49, although some statistical systems use 15–44. More precise comparisons use age-specific fertility rates, which divide births to women in a particular age group by the corresponding female population exposure. These rates describe the age profile of childbearing and distinguish, for example, populations where births are concentrated in the early twenties from those where they are concentrated after age thirty.

The total fertility rate summarizes a period's age-specific rates. With five-year age groups and rates expressed as births per woman-year,

TFR=5∑xfx,\mathrm{TFR}=5\sum_x f_x,

where fxf_x is the fertility rate for age group xx. If rates are expressed per 1,000 women, the sum is also divided by 1,000. The result is the number of children a hypothetical woman would bear if she experienced those age-specific rates throughout her reproductive life and survived through those ages. It is not generally the number of children that women alive in that year will ultimately have.

Parity means the number of live births a person has previously had. Parity-specific measures distinguish first births from second and subsequent births, revealing patterns concealed by the total fertility rate. Most standard fertility statistics are organized by maternal age because maternal birth histories and registration data are more complete. Male fertility can also be measured, but its estimation is complicated by a wider reproductive age range, incomplete paternal information, and differences between partners' ages.

Period fertility, cohort fertility, and birth timing

The distinction between period and cohort measurement is fundamental. Period fertility describes births occurring during a specified calendar interval, usually one year. Cohort fertility follows a group of people born during the same year or years through their reproductive lives. Completed cohort fertility is the average number of children born to a cohort by the end of childbearing, conventionally measured around age 45 or 50 for women.

Period measures provide timely information, but they combine the experiences of many cohorts. In a single year, births to women aged twenty, thirty, and forty belong to different birth generations. A period total fertility rate therefore constructs a synthetic lifetime from people at different stages of their actual lives. Completed cohort fertility instead measures an observed cumulative outcome, but it becomes available only after most reproductive events have occurred.

The difference is particularly important when parenthood is being postponed. Suppose successive cohorts shift their first births from their twenties into their thirties while eventually having similar numbers of children. During the transition, fewer births occur at younger ages before all the deferred births appear at older ages. Period fertility can consequently fall even if completed family size changes little. Demographers call the distortion associated with changing birth timing a tempo effect, contrasting it with changes in quantum, or the amount of childbearing.

A later rise in births at older ages is often described as recuperation. Recuperation need not be complete: some postponed births occur later, whereas others never occur because circumstances, intentions, or biological opportunities change. Period fertility may also rise when postponement slows, even without a corresponding increase in intended or completed family size. Thus, neither a short-term decline nor a subsequent rebound necessarily represents an equivalent change in lifetime fertility.

John Bongaarts and Griffith Feeney proposed an influential tempo-adjustment method in 1998. In a simplified formulation for births of order ii,

TFRi∗=TFRi1−ri,\mathrm{TFR}^{*}_i=\frac{\mathrm{TFR}_i}{1-r_i},

where rir_i is the annual change in mean age at that birth order. The adjustment depends on assumptions about how the timing distribution changes; it is not an observation of future completed fertility. More generally, demographers examine cohort trajectories, mean age at first birth, birth intervals, and parity progression together. A parity progression ratio estimates the proportion of people at a given parity who proceed to another birth, helping distinguish widespread one-child families from a population combining larger families with more permanent childlessness.

Sources, data quality, and estimation

The principal source of fertility data in countries with comprehensive administrative systems is civil registration. Birth records commonly contain the date of birth, the mother's age and residence, and sometimes birth order, marital status, education, and paternal characteristics. Combined with population estimates, these records support annual age-specific fertility rates. Their value depends on completeness, timely registration, consistent definitions, and agreement between the populations covered by numerator and denominator.

A population census can collect the number of children ever born, recent births, or the date of the most recent birth. Questions on children ever born provide cumulative fertility information across age groups, including older women whose childbearing occurred decades earlier. However, respondents may omit children who died or left home, and the accuracy of recent-birth questions can vary. Census data are especially useful for small-area comparisons and for populations that household surveys cannot represent precisely.

Sample surveys provide more detailed reproductive histories. The World Fertility Survey, launched in the 1970s, and the Demographic and Health Surveys, which began in the 1980s, expanded internationally comparable information on births, birth spacing, contraceptive use, and reproductive preferences. A full birth history records each live birth and its date, usually together with the child's survival status. This permits reconstruction of fertility rates for periods before the survey, although estimates become less precise for small subgroups and can be affected by recall errors.

Where registration is incomplete, indirect methods can estimate fertility from partial information. The own-children method, for example, links children enumerated in a census or survey to probable mothers in the same household and reverse-survives them to estimate earlier births and maternal populations. Its accuracy depends on assumptions about mortality, migration, and children living apart from their mothers. Other approaches compare cumulative reported fertility with current fertility schedules, but rapid fertility change can undermine the assumptions required for adjustment.

Errors can arise from age misstatement, omitted births, displacement of birth dates, incomplete geographic coverage, and mismatched residence definitions. Small annual birth counts also produce substantial random fluctuations in local rates. Comparisons therefore require attention to reference periods, age ranges, registration practices, and uncertainty intervals rather than reliance on apparently precise point estimates.

Linked administrative and longitudinal records allow increasingly detailed study of education, employment, partnership histories, and fertility. Such records also contain highly sensitive information. Data privacy protections, secure access arrangements, and statistical disclosure controls are integral to their use, particularly when reproductive histories can be connected to identifiable individuals or small communities.

Biological and behavioral pathways to births

Social and economic conditions affect fertility through mechanisms that change exposure to conception, the probability of conception, or the outcome of pregnancy. Kingsley Davis and Judith Blake formalized this distinction in 1956 by identifying intermediate variables connecting social organization to fertility. Bongaarts subsequently developed a widely used proximate determinants framework, notably in a 1978 formulation emphasizing exposure to sexual intercourse, contraception, induced abortion, and postpartum infecundability.

In a simplified version of the model,

TFR=CmCcCaCi TF,\mathrm{TFR}=C_m C_c C_a C_i\,\mathrm{TF},

where the CC terms are indices representing fertility-reducing effects associated with marriage or sexual exposure, contraception, induced abortion, and postpartum infecundability, while TF\mathrm{TF} denotes theoretical total fecundity. These indices summarize population processes; they do not imply that every individual experiences the same sequence of events. Later applications modify the framework to reflect sexual activity outside marriage and more diverse reproductive circumstances.

Historically, the timing and prevalence of marriage strongly influenced fertility where childbearing occurred predominantly within marital unions. Delayed marriage reduced the years spent exposed to pregnancy, while permanent nonmarriage excluded some people from socially recognized childbearing. In settings where births outside marriage are common, marriage alone is a poor measure of exposure. Cohabitation, nonresident partnerships, periods without partners, and the frequency of sexual intercourse become important to interpretation.

Contraception changes the probability that sexual exposure results in conception. Its aggregate effect depends not only on the proportion using a method but also on method effectiveness, consistency of use, discontinuation, and the distribution of users by reproductive intentions and biological risk. Induced abortion affects the proportion of pregnancies resulting in live births, while spontaneous pregnancy loss also separates conception from recorded fertility. Standard fertility indicators therefore cannot, by themselves, reveal how frequently conception occurs.

After a birth, breastfeeding can suppress ovulation and extend the interval before fecundity returns. Postpartum sexual abstinence can independently lengthen birth intervals. The duration of these processes varies across populations and historical periods. Biological fecundity also changes with age, and infertility can prevent intended births even when exposure and intentions favor childbearing.

Assisted reproductive technology, including in vitro fertilization, enables some births that would otherwise not occur. Its demographic contribution depends on access, eligibility rules, treatment uptake, and outcomes at different ages. These technologies do not remove the distinction between biological capacity and observed fertility: population birth rates still reflect the interaction of reproductive biology with partnership patterns, preferences, institutions, and access to services.

Historical transitions and demographic theory

Before sustained fertility decline, many populations combined relatively high birth rates with high mortality. This did not mean that fertility was biologically unconstrained or identical across societies. Differences in marriage age, breastfeeding, postpartum abstinence, infertility, and separation between spouses produced substantial variation. Historical demographers use natural fertility to describe fertility without deliberate limitation according to the number of children already born, not reproduction unaffected by social practices.

Northwestern Europe provides an important example of fertility regulation through marriage. John Hajnal's 1965 account of the European marriage pattern emphasized relatively late marriage and appreciable permanent nonmarriage in parts of Europe west of an approximate line from Saint Petersburg to Trieste. Such arrangements could restrain births without widespread parity-specific contraception. Historical reconstructions also reveal considerable variation within this broad region and changes over time.

The demographic transition describes the long-term movement from high mortality and high fertility toward lower levels of both. Warren Thompson's 1929 classification and Frank Notestein's work in the 1940s helped establish the modern framework. Mortality decline often preceded sustained fertility decline, creating a period of rapid population growth. The sequence was not universal or mechanically determined: France experienced comparatively early fertility decline from the late eighteenth century, while many other European populations underwent sustained declines mainly during the nineteenth and early twentieth centuries.

Research associated with the Princeton European Fertility Project demonstrated that the timing of decline could not be explained by industrialization or income alone. Linguistic, religious, and cultural boundaries often structured its spread. Ansley Coale's 1973 formulation identified three broad preconditions: fertility had to enter the realm of conscious choice, reduced fertility had to appear advantageous, and effective means of control had to be available. These conditions linked economic incentives to knowledge, legitimacy, and practical ability.

Economic theories subsequently emphasized the costs and benefits of children, including the trade-off between their number and investment per child. Gary Becker's work made this quantity–quality relationship central to household models. John Caldwell's wealth-flows theory focused on whether resources moved predominantly from children to parents or from parents to children. Diffusion approaches stressed the transmission of new practices and ideas through social networks rather than treating each household as an isolated decision-maker.

The second demographic transition, associated with Ron Lesthaeghe and Dirk van de Kaa from the 1980s, connected delayed parenthood, low fertility, and changing partnership behavior with broader transformations in values and autonomy. Its applicability varies across regions. Contemporary explanations commonly combine material conditions, institutions, social norms, and life-course constraints rather than assigning fertility change to a single universal cause.

Education, work, partnerships, and social differences

Education influences fertility through several pathways. School enrollment often postpones union formation and first births, while qualifications can alter employment opportunities, access to information, and expectations about family life. Educational attainment is consequently associated with differences in both birth timing and completed fertility. The relationship is not identical across countries or cohorts: highly educated groups may delay births substantially but recover part of that delay later, while institutional conditions can change the association between education and family size.

The economic costs of children include direct expenditure and the time devoted to care. In many contemporary economies, time away from paid work can affect earnings, career progression, and pension accumulation. Housing availability, insecure employment, and the cost of childcare can also shape the sequence of leaving the parental home, forming a partnership, and becoming a parent. These mechanisms connect fertility to the labor market, but an association between income and fertility does not establish a simple causal relationship: income may affect childbearing, while childbearing can itself change earnings and employment.

The distribution of paid and unpaid work within households is another major line of research. Gender equality in schooling and employment can coexist with unequal responsibility for domestic labor and care. Peter McDonald argued around 2000 that a mismatch between relatively egalitarian individual-oriented institutions and less egalitarian family institutions could help explain very low fertility. Comparative studies examine whether childcare provision, working-time arrangements, and men's participation in care alter the compatibility of employment and parenthood; estimated relationships depend on context and measurement.

Partnership patterns directly affect opportunities for childbearing. Later union formation, periods without a partner, separation, divorce, and repartnering can change both first-birth rates and progression to additional births. A rise in births outside marriage does not necessarily indicate a rise in unpartnered parenthood, because many occur within cohabiting couples. Conversely, marital status at birth provides only a snapshot and does not describe the child's subsequent household circumstances.

Fertility also differs by residence, socioeconomic position, ethnicity, and migration history. Urbanization may change housing costs, schooling, child employment, and access to services, but urban–rural differences also reflect selective migration. Migrants' fertility patterns can reflect conditions before migration, disruption during movement, family reunification, and adaptation after arrival. National averages conceal these distinct histories.

Reproductive intentions are informative but not fixed predictions. People may revise their desired family size as relationships, health, employment, or existing children's needs change. Surveys distinguish ideal family size, desired additional births, and intentions within a specified period. Gaps between intentions and subsequent births can therefore reflect both constrained opportunities and genuine changes in preferences.

Global distribution and recent change

The global decline in fertility has been large but geographically uneven. The United Nations World Population Prospects 2024 estimated global total fertility at about 2.25 births per woman in 2024, compared with approximately 3.31 in 1990. More than half of countries and areas had fertility below 2.1 births per woman. These figures are period estimates assembled from registration, censuses, surveys, and demographic models, rather than direct measurements of completed family size.

Europe, eastern Asia, and many parts of the Americas have experienced sustained below-replacement fertility. Some eastern Asian populations have recorded particularly low period rates: the Republic of Korea's total fertility rate was 0.72 in 2023, according to its national statistical office. Such a value describes the year's age-specific birth rates; it does not mean that every woman, or the average completed cohort, has precisely that number of children.

Fertility decline has also occurred across much of Latin America and the Caribbean, frequently alongside substantial social and regional inequality. Falling national totals can coexist with differences in adolescent fertility, timing of first births, and access to reproductive services. Earlier first births among some groups and extensive postponement among others can produce a national age schedule that differs from the typical pattern observed in either group.

Sub-Saharan Africa contains many of the world's highest-fertility populations, but the region is not demographically uniform. Countries differ in the onset, speed, and continuity of decline, and urban and educational differences can be considerable. Child survival, desired family size, partnership institutions, contraceptive access, and women's schooling all contribute to these contrasts. A regional average therefore cannot represent a single reproductive regime.

Rapid declines in several Asian and Middle Eastern populations demonstrate that fertility transitions need not take as long as they did in much of Europe. In Iran, for example, a particularly rapid decline occurred from the late 1980s through the 1990s. Explanations examine the expansion of education and health services, changes in marriage and family preferences, and access to family planning. No single national trajectory supplies a universal timetable for other societies.

Annual fertility can also respond to wars, epidemics, economic crises, and disruptions to partnership formation. The COVID-19 pandemic produced different short-term birth patterns across countries rather than one uniform global response. A fall followed by a rebound may reflect shifted timing, changes in registration, or altered birth intentions. Establishing a durable trend requires several years of age- and parity-specific evidence rather than extrapolation from one unusually low or high year.

Public policy and reproductive rights

Governments influence fertility both through explicit population policies and through institutions not primarily designed to change birth numbers. Explicit policies may seek to reduce births, increase births, or enable people to realize their own reproductive preferences. Broader arrangements concerning schooling, housing, employment, taxation, healthcare, and migration also affect the conditions under which childbearing occurs. Policy intent, individual response, and aggregate demographic outcome are separate questions.

Family planning programs expanded in many countries during the second half of the twentieth century. Their activities have included contraceptive supply, counseling, provider training, and integration with maternal and child health services. Their demographic effects depend on existing demand, service quality, method availability, continuity of provision, and broader social change. A decline occurring alongside a program cannot automatically be attributed entirely to that program.

Pronatalist measures include birth grants, child allowances, tax concessions, subsidized childcare, parental leave, and support for housing or fertility treatment. The welfare state shapes these policies through different combinations of cash support, public services, and employment protections. Evaluations distinguish whether a policy changes the timing of births, the likelihood of an additional birth, or completed family size. A payment can bring a planned birth forward without increasing the eventual number of children.

Institutional combinations matter. A leave entitlement operates differently depending on job protection, payment levels, eligibility, and the availability of childcare afterward. Policies may also affect groups differently: employment-linked benefits can exclude people outside stable formal work, while means-tested transfers reach a different population. Cross-country comparisons must therefore consider policy design and coverage, not merely public expenditure totals.

Population policy has also included coercive measures. China's one-child policy, introduced around 1980 and later replaced by two-child and then three-child policies, imposed administrative limits on births with differing rules and implementation across places and groups. Romania's 1966 restrictions on abortion and contraception sought to raise births. Such histories are studied not only for demographic effects but also for their implications for bodily autonomy, family life, and state power.

The 1994 International Conference on Population and Development in Cairo marked a major international emphasis on reproductive health and individual choice. Its Programme of Action recognized the right of couples and individuals to decide freely and responsibly the number, spacing, and timing of their children, with access to relevant information and means. In this framework, human rights and the absence of coercion are distinct from whether a government's preferred numerical fertility target is achieved.

Replacement, population momentum, and age structure

Replacement-level fertility is the fertility level at which a generation of women would, under specified mortality and sex-ratio conditions, be replaced by an equal-sized generation of daughters. In low-mortality populations it is commonly approximated as 2.1 births per woman. The figure is not a universal biological constant: it varies with survival to and through reproductive ages and with the sex ratio at birth. Where mortality is higher, more births are required for generational replacement.

The net reproduction rate expresses replacement more directly. It combines age-specific female fertility with female survival, usually using a life table, to estimate the average number of daughters a newborn girl would bear under the specified schedules. A net reproduction rate of one indicates replacement under those conditions. The gross reproduction rate also counts daughters but does not make the corresponding adjustment for maternal mortality before and during childbearing ages.

Replacement-level fertility does not imply immediate population stability. A population with many young people may continue growing after fertility falls to replacement because large cohorts are entering reproductive ages. Conversely, a population with a small childbearing-age population and many older people may continue declining after fertility rises to replacement. This dependence on inherited age structure is called population momentum.

Birth numbers depend jointly on fertility rates and the number of people exposed to those rates:

Bt=∑xfx,tEx,t,B_t=\sum_x f_{x,t}E_{x,t},

where Ex,tE_{x,t} is female exposure at age xx during period tt. If the reproductive-age population contracts, births can fall even while age-specific fertility rises. This distinction explains why a recovery in the total fertility rate need not produce an increase in annual births.

Sustained low fertility contributes to population ageing by reducing the size of younger cohorts relative to older ones. Initially, falling fertility can reduce child dependency and increase the working-age share, creating conditions associated with a demographic dividend. Whether this translates into higher incomes depends on employment, health, education, institutions, and economic conditions. Later, smaller entering cohorts can contribute to a higher old-age dependency ratio, although age-based ratios do not directly measure employment or actual economic dependence.

Migration and mortality can modify these outcomes. Immigration can enlarge working-age populations and add future births, while longer lives alter the balance among age groups. Population projections therefore combine fertility, mortality, and migration assumptions. A fertility rate considered in isolation cannot determine the future population size, fiscal balance, environmental impact, or standard of living of a country.

Analytical methods and unresolved questions

Modern fertility research treats childbearing as a sequence of transitions rather than only a yearly aggregate. Event-history analysis estimates the rate at which individuals experience events such as first birth or a subsequent birth, conditional on their observed histories. It can incorporate changing partnership status, enrollment, employment, and time since a previous birth. Censoring is central: a person who has not yet had another child by the end of observation is not necessarily someone who will never do so.

Longitudinal data help establish temporal order, but temporal order alone does not establish causation. Employment, partnership formation, housing, and fertility influence one another. Unobserved preferences or family background may affect all of them. Studies therefore use approaches such as fixed-effects models, policy discontinuities, instrumental variables, and comparisons around institutional reforms. Each identifies an effect only under specific assumptions and often for a particular population.

Selection creates additional difficulties. People entering stable partnerships may differ from those remaining unpartnered, and people using fertility treatment differ from those who do not seek it. Comparing their eventual number of children without accounting for these differences can misrepresent the effect of partnership or treatment. At the aggregate level, correlations between national income, gender equality, and fertility can likewise change when the countries, years, or controls included in the analysis change.

A persistent research problem is forecasting low fertility. Completed fertility for younger cohorts is not yet observable, and future recuperation after postponement remains uncertain. Models differ in how much convergence they assume across countries and whether very low rates are treated as temporary or durable. Probabilistic projections express some of this uncertainty through distributions of possible trajectories, but their results remain conditional on model structure, historical information, and assumptions.

Male fertility and the interaction between partners' reproductive histories remain less comprehensively measured than female fertility in many statistical systems. Further challenges arise from donor conception, cross-border reproductive treatment, and family forms in which genetic, gestational, legal, and social parenthood do not coincide. Recording births remains possible, but attributing fertility and studying pathways to parenthood require carefully defined concepts rather than the assumption that all parenthood follows one model.

Another central question concerns the relationship between stated preferences and realized births. Intentions are revised over the life course, and survey responses depend on question wording, reference period, and partnership context. Repeated surveys linked to subsequent outcomes can distinguish persistent unrealized intentions from changing goals. These approaches require observation over substantial portions of reproductive life, making timely explanation difficult precisely when annual fertility is changing most rapidly.