Mobility, Social Class and Fertility Transition in England and Wales, 1851–1911
Hannaliis Jaadla, University of Cambridge
Alice Reid, University of Cambridge
Eilidh Garrett, University of Essex
Mobility,social class and fertility transition in England and Wales, 18511911
Priorresearch on the fertility transition in England and Wales has demonstratedsubstantial variation in fertility levels and declines by different socialgroups (Woods 1984; Haines 1989; Szreter 1996; Anderson 1999; Barnes andGuinnane 2012). However these findings were generally reported at a broadgeographical level, disguising local variation and complicated by residentialsegregation along social class and occupational lines: the causal pathwaysleading to the changes in behaviour which reduced fertility are far from clear.This paper will focus on the interplay of socioeconomic status, spatialmobility and geographical context determining observed patterns during thefertility transition.
Twobroad frameworks have been developed to explain the European fertilitytransition (Carlsson 1966). The innovation perspective attributes fallingfertility to the spread of new knowledge and attitudes, while the adjustmentperspective conceptualises fertility decline as a response to transformedeconomic and social environment. For both of these explanations, the analysisof the extent of socioeconomic variations in fertility, as well as how thesetransform during the process of fertility decline, has been fundamental tounderstanding the nature of the fertility transition.
Themain data source for this paper is individual level census data for England andWales from 18511911. Our analysis of fertility differentials relies on thenumber of surviving children (aged 04) currently living in the household. Weonly include women (aged 1554) whose spouse was present at the time of census,as socio-economic status is derived from husbands occupational status. In thefirst models, the occupations are grouped by social classes introduced by theRegistrar General in 1911. However, in the next stages, for comparativepurposes, we are planning also to employ HISCLASS classification scheme (vanLeeuwen and Maas 2011). We use distance between the place of birth and theplace of enumeration, measured in km, as an indicator of mobility or netlife-time migration.
Ourmodelling strategy was to estimate the same set of models for all six censusyears. First, the OLS models only measured the relationship between netfertility and the main two variables of interest social class and distancefrom place of birth. These first models only control for the age of woman. Inorder to analyse how much of the relationship between net fertility and SES orlife-time migration is explained by unobserved heterogeneity at the geographicalunit level (Registration Sub-Districts), we estimated fixed effects models.These models include additional individual-level control variables age ofwoman, age difference between spouses, and husbands household status. Finally,the full model includes measures both at individual level (womens employmentstatus, presence of children older than 5) and at RSD-level labour forceparticipation of unmarried women, singular mean age at marriage for women,population density, proportion of agricultural workers and miners.
Theresults of OLS model in Table 1 show quite large fertility differentials acrossdifferent social classes. Women married to upper and middle class menexperienced the lowest net fertility in all the years. Unskilled workers,miners and agricultural labourers had the highest net fertility. Over time thenet fertility of upper and middle classes becomes increasingly distinct fromthose of other social classes.
Whencomparing the OLS model with fixed effects model, considerable part of thesocial class differentials in net fertility are explained by the spatialheterogeneity and structural differences across RSDs and other individualcharacteristics that were included as control variables. Prior to the fertilitydecline, these differences were rather narrow. However, over time the patternsof gross socio-economic differentials prevail to a large extent even in thefixed effects model setting. The differences in net fertility by social classwiden during the first stages of fertility transition. This analysis demonstratesthat the basic patterns of socio-economic differentials in fertility remain thesame in England and Wales even when taking into account the spatialheterogeneity and therefore socio-economic status was an important determinantof net fertility during the fertility transition.
Thelife-time migration variable demonstrates in all the years that women residingwithin 5km to their place of birth have highest net fertility. There seems tobe an expected gradient in the relationship between marital net fertility andthe distance from place of birth, the differences increase with distance. Evenin the fixed effects model setting if some of the observed differences can beexplained by geographical heterogeneity, the main patterns remain unchanged.The temporal pattern suggests that differences were larger earlier in thefertility transition and by 1911 these considerably diminished. In a way thispoints that the meaning of distance was drastically changing over the course ofthe second half of the 19th century and rapidly developingtransport networks were transforming the understanding of short and longdistance migration.
Presented in Session 1233: Posters