Fertility Differences by Housing Type: The Effect of Housing Conditions or of Selective Moves?
Kulu, Hill, Vikat, Andres, Demographic Research
This study examines fertility variation across housing types and childbearing patterns following housing changes. While the effect of family changes on housing choices has been studied in detail, little is known about childbearing patterns within various housing types, and this despite the fact that many studies suggest housing to be an important determinant of fertility. We use longitudinal register data from Finland and apply hazard regression. First, we observe a significant variation in fertility levels across housing types - fertility is highest among couples living in single-family houses and lowest among those residing in apartments, with the variation remaining significant even after controlling for the demographic and socio-economic characteristics of women. Second, our results show elevated fertility levels after couples have changed dwellings, suggesting that much of the fertility variation across housing types is attributed to selective moves. Third, the study reveals a relatively high risk of third birth for couples in single-family houses several years after the move. This suggests that living in spacious housing and in a family-friendly environment for a relatively long time leads to higher fertility.
There is a long research tradition that looks at the effects of family changes on spatial mobility and housing choices in Europe and North America. Previous studies showed that an increase in family size leads to a reduction in the desire and chances to make long-distance moves, particularly to urban destinations (Sandefur and Scott 1981, Courgeau 1989, White et al. 1995, Kulu 2005, 2007). The birth of a child significantly increases the propensity of couples to move short distances because they wish to adjust their dwelling size to their family size (Clark et al. 1984, Courgeau 1985, Deurloo et al. 1994, Davies Withers 1998, Clark and Huang 2003, cf. Murphy 1984). Recent studies on the timing of moves relative to childbearing revealed that many couples move when waiting for their child to be born (Mulder and Wagner 1998, Michielin and Mulder 2005, Kulu 2007). Some researchers argued that couples increasingly move in anticipation of childbearing, particularly those that move to home-ownership and to single-family houses (Feijten and Mulder 2002).
While it is not surprising that family events are important triggers of housing transition, it is less clear to what extent a change of housing or of housing conditions shapes the childbearing patterns of couples. Naturally, the question is not new, and it challenged researchers as early as in the 1930s when below-replacement fertility emerged in several European countries (Chesnais 1992). For example, Goodsell (1937) examined the causes of low fertility in Sweden and argued that home overcrowding was partly responsible for low fertility in the urban areas of Sweden. Swedish architects and builders, in their zeal to re-house urban workers in modern flats, produced a standardised tenement of one room and kitchen, and this might have forced couples to consider limiting their family size, particularly as more spacious, convenient, and inexpensive housing remained unattainable for many couples (Goodsell 1937: 855).
Thompson (1938) suggested that similar conditions might have existed in the U.S. The author argued that the availability of adequate housing at a desired standard was an important factor in determining the number of children reared in many families: 'There can be little doubt that housing which costs so much that a family cannot afford the space it considers proper for its position, if it has several children, will tend to discourage the rearing of more than one or two children, or, indeed, any children at all. Under present conditions, where many families must live in one or two or three rooms in order to keep their housing expenditures within bounds, it is not surprising that they feel they can afford at most only one or two children' (Thompson 1938: 363). …