Understanding Recent Fertility Trends in the UK and Improving Methodologies for Fertility Forecasting
The project ‘Understanding Recent Fertility Trends in the UK and Improving Methodologies for Fertility Forecasting’ (FertilityTrends, 2019-2023) examined the significant fluctuations in fertility levels in the United Kingdom in the last three decades, investigated their causes, and developed improved methodologies for fertility forecasting. Fertility rates declined in the 1990s, increased in the first decade of this century and have declined thereafter. Fertility dynamics have been similar in the four constituent countries of the UK, but the levels in Northern Ireland are higher than in England or Wales and lower in Scotland.
The project used census-linked administrative data from England and Wales and from Scotland. The analysis of fertility trends by birth order showed that declining first-, second-, and third-birth rates contributed to declining fertility in the 1990s. Changes in first-birth rates have explained most fluctuations in total fertility since the turn of the century. The timing of parenthood, i.e. the postponement of childbearing (in the 1990s and 2010s) or the stopping of it (in the 2000s) has been mostly responsible for the changes in first-birth rates. Second- and third-birth rates declined in the 1990s, but remained relatively stable in the first decades of this century. Finally, we observed relatively similar fertility trends and timing in England and Wales and in Scotland, but significantly lower second- and third-birth rates in Scotland. Our analysis of cohort fertility supported that Scotland had smaller families than England and Wales, whereas Northern Ireland had larger families.
The gradual postponement of family formation can be explained by the long-term economic changes that have led to increased uncertainty. Recent declines in first-birth rates have been fastest among those with the least education. Family formation among women in their thirties has become a norm among the younger generations. It is possible that the stability in second- and third-birth patterns has been supported by welfare state provisions and policies promoting the reconciliation of work and employment and the involvement of both parents in childrearing. Parenthood is postponed due to economic uncertainty experienced in young adulthood or due to normative factors, but when a couple is confident they can have a child they have a second and a third child.
Our research showed that fertility preferences are relatively similar in England and Wales and in Scotland, but women in Scotland are less likely to have a second and third child. The challenge for the policy-makers, especially in Scotland is to reduce the gap between desired and actual family size for parents. This may involve supporting young adults in their transition to the labour market after finishing education and providing opportunities for affordable housing. For example, private renting should provide long-term security for those who cannot afford homeownership and/or wish to rent.
The project developed methods for fertility projections. We developed a Bayesian parity-specific fertility projection model within a generalized additive model framework, estimating smooth age-cohort rate surfaces for each birth order simultaneously. Our key innovation was the incorporation of constraints on parity progression ratios which increase forecast precision and accuracy. The application of the model in England and Wales and in Scotland shows a gradual fertility decline for the cohorts born in the 1980s and later with an increased share of zero- and one-child families. These developed methodologies could be applied to project fertility in other industrialised countries.