Housing and First Births in Sweden, 1972–2005

Abstract

The aim of this study is to explore to what extent housing (type, tenure and number of rooms) is a constraint for first births in Sweden 1972–2005. The Swedish Housing and Life Course Cohort Study (HOLK) (n = 2242) is used. The occurrence of childbearing is measured as the event of the birth itself and the time 16 months prior to the birth, i.e. initiation of conception. The main finding is that the size of the dwelling seems to be the housing factor with the strongest association with first-birth intensities. Furthermore, an association between being established on the housing market and the propensity to have a first child is found particularly for the cohort born in 1974. The effect of housing on childbearing seems to be stronger if measured to capture time of any first birth than if measured so as to coincide with the situation 16 months prior to the birth.

Keywords:

• Housing supply
• housing policy
• housing tenure
• fertility

Acknowledgements

The author wishes to thank the editors of this volume, Gunnar Andersson and three anonymous referees for valuable comments, and also Maria Brandén and Ingemar KÅreholt for assisting in the data preparation. Financial support from Byggnads, Folksam, HSB, Hyresgästföreningen Riksförbundet, Riksbyggen and Svenska Byggnadsutvecklingsfonden (data collection), and FAS research grant 2005-1101 is gratefully acknowledged.

Notes

¹ Naturally, low fertility levels in Southern Europe cannot be ascribed to scarce opportunities for access to housing alone. These countries also have family policy models dominated by support from the family rather than state support (Esping-Andersen, 1990).

² A very small number of terraced houses (or ‘rowhouses’ i.e. dwellings situated in a building with three or more houses in a row that share a wall with at least one adjacent neighbour) are accessed through rental tenancy.

³ It can also be argued that the two oldest cohorts should be censored at age 31. First, housing is likely to have different effects on childbearing in young adulthood compared to middle age. Second, the cohort born in 1974 is 31 by the end of the year of data collection (2005) and a substantial proportion of the respondents are thus not likely to have completed their fertility. Analyses for the age group 16–31 have been performed but are not presented here. The results are very similar to those presented in Tables 1 and 2.

⁴ Obviously the realization of pregnancy plans is not always guaranteed.