01 Section B Option 1, Key question 1 - Thirty years of family and demographic change (part 2)

Reading family trends carefully

A family trend describes a pattern across a population, rather than the experience of every individual family.

Demography is the study of populations, including their size, age structure, births, deaths and migration. Demographic change affects who is available to form relationships, become a parent or provide care. It interacts with people's choices and the conditions in which they make them.

Distinguish a family from a household. A family involves relationships such as those between partners or parents and children; a household consists of people sharing living arrangements. Grandparents can belong to a wider family while living elsewhere. A person living alone forms a household, although they may have close family relationships outside it. In official statistics, the precise categories also matter: an ONS cohabiting-couple family can have children or no children.

Three distinctions make trend evidence useful:

For example, a decrease in the number of weddings does not reveal whether couples are living together without marrying, delaying marriage, remaining single or living separately. It needs evidence about living arrangements as well. A short-term change also needs context: postponed pandemic weddings could increase the following year's marriages without reversing a longer trend.

Survey estimates have sampling uncertainty because they describe a population using a sample. ONS family estimates use the Labour Force Survey; recent smaller achieved samples have increased uncertainty. A small difference between years may therefore be insufficient evidence of a real population change. Registration statistics record legal events, but their coverage still depends on what is registered and where. ONS families and households, 2025.

Marriage and cohabitation change together

Marriage has become less frequent over the long term, while living together without marriage has become a more common part of family formation.

Cohabitation means living together as a couple without being married to each other. It can be a stage before marriage, a lasting alternative to it, or a relationship following separation. It therefore does not always indicate rejection of marriage.

ONS recorded about 312,000 marriages in England and Wales in 1992 and 247,000 in 2022, a decline of about one-fifth across thirty years. The 2022 total nevertheless exceeded the pre-pandemic 2019 total, illustrating why a long trend need not fall every year. Among opposite-sex couples marrying, around 60% had previously cohabited in 1994, compared with about 90% in 2022. These percentages concern people who married, rather than all couples. ONS marriages, 2021 and 2022.

For the UK, ONS's 2017 release estimated 1.5 million cohabiting-couple families in 1996 and 3.3 million in 2017. Its 2025 release estimated 3.5 million, or 17.6% of families. However, ONS did not find the smaller increase from 2015 to 2025 statistically significant: long-term growth is not evidence of an equally rapid rise in every decade. ONS families, 2017, ONS families, 2025.

Marriage has also moved later in the life course, the sequence of stages and transitions in a person's life. ONS's England-and-Wales cohort analysis shows that half of women born in 1973 had married by age 31, whereas women born in 1983 reached that milestone by 35. The equivalent ages for men were 35 and 40. These are comparisons between birth cohorts, rather than the average age of everyone marrying in a particular year. ONS marriages, 2023.

Why can these patterns develop? First, reduced social disapproval can make cohabitation an acceptable way to form a family. NatCen's British Social Attitudes report, published in 2023 using 2022 survey evidence from adults in Great Britain, found that agreement that couples may live together without intending to marry had risen from 64% in 1994 to 81%. This repeated survey measures expressed attitudes, not the motives of each cohabiting couple. It supports a change in norms without proving that attitudes alone caused living arrangements to change. NatCen, Moral issues.

Second, education, paid employment and financial independence can give people more ability to choose when and whether to marry. Conversely, insecure employment and housing costs can make establishing a separate home difficult. Couples may share costs while postponing a wedding, while others remain with parents. Secularisation, a reduction in religion's social influence, can weaken religious pressure to marry, but religious people also differ in their attitudes and choices. These mechanisms can combine; they do not apply equally to every class, community or individual.

Divorce and changing relationship histories

Divorce changes a legal marriage, but relationship breakdown is wider than the number of divorces recorded.

Divorce legally ends a marriage; separation means partners cease their relationship or living together without necessarily ending the marriage legally. Cohabiting couples can separate without divorcing. Consequently, divorce figures cannot measure all relationship breakdown.

The familiar claim that divorce has continually increased is misleading for the last thirty years. Divorce numbers in England and Wales peaked in 1993 and have generally declined since. ONS recorded 102,678 divorces in 2023, including opposite-sex and same-sex marriages. The rise from the unusually low 2022 total partly reflects changes in processing and waiting periods, rather than straightforward evidence that relationships suddenly became less stable. ONS divorces, 2023.

If fewer people marry, the number of marriages available to end also changes. Assessing marital stability therefore needs a rate relating divorces to the married population, alongside information about marriage duration and timing, rather than an annual count alone.

To explain divorce, distinguish the desire to leave a relationship from the ability to leave it. Greater financial independence can make leaving economically possible. Reduced stigma can lessen social penalties, while expectations of affection, equality and personal fulfilment can make an unhappy relationship less acceptable. These are possible mechanisms, not evidence that every divorce has the same cause. Housing costs, childcare responsibilities or limited income can still constrain separation.

Law also shapes how an existing breakdown becomes a legal event. In England and Wales, the Divorce, Dissolution and Separation Act 2020 reforms took effect on 6 April 2022. The new procedure removed the need to establish fault or a specified period of separation and allowed joint applications, but introduced a minimum twenty-week period before the conditional order. It therefore changed both access to divorce and the timing of recorded final orders. It cannot explain changes that occurred before its introduction. Ministry of Justice, April 2022.

A divorce may lead to a one-person household, a lone-parent family or later a new partnership. A reconstituted family brings together a couple where one or both have children from an earlier relationship. Such families can also follow the breakdown of cohabitation or bereavement. A child's family relationships may extend across two households, so an address-based count does not describe the whole network of care.

Consider this original illustration: a divorced parent lives with a child and later moves in with a partner who also has a child. The household changes from a lone-parent arrangement to a reconstituted family. The relationships with the children's other parents may continue. Diversity can therefore involve different stages in one person's life, rather than permanently separate kinds of people.

Later childbearing and family size

Changes in when people have children and how many they have affect family life, but different birth measures answer different questions.

A birth rate is not simply a birth count. In an original numerical illustration, 1,000 births in a population of 100,000 gives 10 births per 1,000 people. If births remain 1,000 but the population grows to 125,000, the rate falls to 8 per 1,000. The rate can change because the population's size or age structure changes, as well as because people's childbearing behaviour changes. ONS birth-statistics user guide.

The total fertility rate, or TFR, combines a particular year's age-specific fertility rates. It indicates how many children a woman would have if she experienced those rates across her childbearing life. It is an annual synthetic measure, rather than the actual completed family size of women alive that year.

The broad fertility pattern has fluctuated: ONS reported falling fertility through the 1990s, followed by recovery during the 2000s. Its historical England-and-Wales series put TFR at 1.63 in 2001 and 1.94 in 2010. Later evidence shows renewed long-term decline, reaching 1.41 in 2024. In 2024 the crude birth rate also decreased, although the number of live births increased. This illustrates why counts and rates must be kept separate. ONS historical births, 2012, ONS births, 2024.

Childbearing has also moved later. The standardised mean age of mothers giving birth in England and Wales was 31.0 in 2024; this includes all births, not just first children. ONS's cohort evidence compares women born in 1953 and 1979: they reached an average of one child per woman by ages 27 and 31 respectively. That milestone is not the average age at first birth. ONS births, 2024, ONS cohort fertility, 2024.

Completed family size measures average live-born children across a birth cohort assumed to have finished childbearing. For England and Wales, ONS estimated 2.05 for women born in 1953 and 1.95 for those born in 1979. It excludes adopted children and stepchildren: it is a fertility measure, not a count of everyone in a household or all a person's family ties. Smaller averages can coexist with large families, and an average below two does not mean most families contain a fraction of a child. ONS cohort fertility, 2024.

Several mechanisms can explain later or fewer births. Contraception makes it more possible to plan timing and spacing. Longer education and building a career can encourage postponement. Housing, childcare costs and uncertain income can make desired parenthood difficult to afford. Reduced pressure to become a parent can support chosen childlessness. Some people who postpone births later have children, while others have fewer than intended or face involuntary childlessness. A low annual TFR therefore cannot by itself establish that everyone prefers a permanently smaller family.

These changes create longer periods in couple families without children, single-person households and parental homes before children arrive. Later parenthood can also overlap with caring for older relatives. Family size and household size need not move together: a couple with one child and two grandparents has few children but a relatively large household.

An ageing population and wider family ties

An ageing population contains a growing proportion of older people, which changes the potential relationships between generations.

Other things being equal, lower fertility reduces the size of younger generations, while improved survival allows more people to reach older ages. Large birth cohorts moving into later life can also change the age structure. Migration affects the balance too, depending on migrants' ages. Population ageing is therefore more than a claim that each individual gets older.

The England-and-Wales census found that people aged 65 and over formed 16.4% of the population in 2011 and 18.6% in 2021. These dated results illustrate a continuing ageing pattern; they are neither a UK-wide census percentage nor a prediction. ONS Census 2021.

Longer survival can produce families with three or four living generations. Combined with fewer children in each generation, this may create a beanpole family: a relatively long vertical chain of generations with fewer siblings, cousins, aunts and uncles spreading out sideways. It describes a kinship pattern, not necessarily everyone living in one house. Grandparents might provide childcare or financial help from a separate home, while adult children support older relatives.

The effect on households is not uniform. Older partners may live together for longer. Bereavement can leave a surviving partner living alone. Care needs may encourage living with adult children or moving into supported accommodation. Living alone does not necessarily mean loneliness or the absence of family: frequent contact and practical support can cross household boundaries.

Nor are all older people dependent. Health, income and family circumstances vary. An older relative can be both a recipient of help and a provider of childcare, money or emotional support. Some middle-aged adults face responsibilities towards children and older relatives at the same time. Whether that becomes a serious burden depends partly on employment flexibility, public services and how care is shared within the family. The same demographic trend can therefore sustain family connections and increase pressures on them.

Explaining connected family change

A convincing explanation connects a social change to a mechanism and then to a particular family or household outcome.

Saying “families have changed because society is modern” names no mechanism. A useful explanation identifies what changed, how it affected opportunities or expectations, and what arrangements could follow. It then checks whether the evidence supports that account and considers a limit or alternative.

Consider this original teaching scenario. Two adults want a child but face insecure work and high housing costs. They cohabit while saving for a more suitable home. One person's older mother lives nearby and offers future childcare. They eventually have a child before marrying.

One explanation connects economic constraint to later parenthood: uncertain income makes a larger home and childcare harder to afford, so the couple postpone a desired birth. Another connects changing norms to the child's family form: if marriage is not treated as a necessary condition of parenthood, they can become parents while cohabiting. A third connects longer survival to intergenerational support: the grandmother's childcare can make their plans more feasible. None of these explanations alone accounts for every decision.

This is a reasoned application, not evidence that all couples behave this way. A higher-income couple might face different constraints; a family without nearby relatives might lack that support. Law governs recognition and separation, attitudes affect approval, and resources affect which options are practical. Agency, people's capacity to make choices, operates within these structural conditions, the wider arrangements that enable or constrain action.

The evidence also permits continuity. Marriage remains a major family form, cohabitation can precede marriage, and living apart does not end kinship. More diverse relationship histories and households therefore do not by themselves demonstrate either the disappearance of the family or that every new arrangement is freely chosen. The strongest judgement fits the measured trend and the mechanism being explained.