Author: Ingrid van Dijk (Lund University)
Ingress:
In historical the Netherlands, up to half of children did not make it to their first birthday. But not all families were similarly likely to lose a child. Young children are highly dependent on their families to survive. The infants and young children of fathers and mothers from families with high mortality – who lost many siblings in childhood – had an increased risk of early mortality themselves.
Intergenerational inequality in child mortality
Social inequalities in child mortality are commonly understood from a social background perspective, studying families’ levels of income, their social class, educational background or neighborhood characteristics. Studying such inequalities is highly relevant, as inequalities in child loss captures inequalities between families in adverse living conditions and the ability to keep children alive. But these characteristics are often measured imperfectly – we have limited knowledge about the exact living conditions of historical families. Moreover, adverse conditions accumulate in families – they are not just poor, but also live in neighborhoods with unhealthy living conditions, and in crammed quarters, for example. How can we get a better grasp of inequalities in infant and child mortality?
A highly relevant marker of inequality in mortality, child loss, and living conditions is whether children escape the conditions of their parents. In other words, children have better outcomes than their parents. Do children from parents who lost many children in childhood have a higher risk of mortality among their own children, or do they escape the fate of their own parents? If adverse living conditions are shared between generations, the level of intergenerational similarities in high child mortality should be high – resulting in larger inequalities in child loss between families.
Child mortality concentrates in families
Using linked civil certificates from the Netherlands and analyzing mortality patterns among more than 200.000 children born between 1835 and 1914, I show that there is a clustering of mortality in families, which exists both for upper class and working class families. There are more families without any child deaths or with low mortality, and a small number of families experiencing a large part of child mortality.
Figure 1. There are more families with no child deaths, fewer families with moderate child mortality, and more families with a large number of child deaths observed in the population (filled line) than expected (dotted line).
In the period 1835-1919 infant and child mortality is initially high, then falls rapidly from the 1880ies onwards. Despite these changes, there is still a significant inequality in the distribution of mortality. Social class inequalities in child mortality persist. Moreover, intergenerational inequality persists. Particularly children of mothers who originated from a high mortality family – and among whose siblings at least two children died before the age of five – had a higher risk of mortality before their fifth birthday. The risk is also higher for children of fathers from a high mortality family, but the effect is not as strong, and attenuates over time.
Inequalities accumulate in families
The highest mortality is found for lower-class, high-mortality families. The inclusion of the social class of families – and of the grandparental families – does not explain the intergenerational correlation in child mortality. The contribution of whether or not parents are from high-mortality families is even stronger than that of the social class of their family in childhood (based on their father’s occupation) and adulthood (based on husbands’ occupations). They are also not explained by place: mothers and fathers who moved away from their own childhood environments also experience intergenerational persistence in high mortality.
Figure 2.
Contribution to the literature
High infant and child mortality in historical contexts has been explained by the low living standard at the time. Inequalities between social groups in their resources, access to nutrition, housing, and information was related to social inequalities in child mortality by occupational status. Although Zeeland in the studied period is no exception to this, and I show evidence for a social gradient in child mortality, intergenerational persistence in child mortality is not driven by social class of parents or grandparents alone. Only a small fraction of intergenerationally high risk of child death is explained by socioeconomic characteristics. These findings underline that intergenerational similarities in mortality are an important indicator of inequality of opportunity already at the start of life. The paper thus contributes to the literature on inequality in childhood mortality in historical contexts with relevancy also for contemporary inequalities in health and survival over the life course.