|
|
Intergenerational clustering of infant and child mortality
| In the European past, infant and children’s deaths were unevenly distributed among families. The extent to which this played a role is unknown. Neither do we have a clear explanation for this mortality clustering. However, the recent finding of this clustering has far-reaching theoretical and methodological consequences. Instead of focusing on differences between individuals, it seems clear that it is equally important to look at differences between families. It has been argued that families appear to be a better unit for analysing infant and child mortality than the individual child. A focus on the family as a primary unit for analysis will refine our ideas about the preconditions of infant and child mortality.
Among the possible explanations for mortality clustering on a family level, biological (genetic) and social (socialization, social learning) causes seem to be the most logical ones. Some families may have ‘weaker’ genes, leading to an excess in infant and child mortality. In other families a higher infant and child mortality can be explained by specific nursing practices (e.g. breastfeeding), passed on from mother to daughter. If these explanations are to some extent important to the death clustering phenomenon, then this clustering might not only be found within families, but between generations as well.
In this paper I will focus on the intergenerational aspects of both infant and child mortality clustering during the second half of the nineteenth century. Is mortality clustered between generations as well? Do the surviving children of families with an excess infant and child mortality, so called ‘high-risk families’, have a higher chance to become parent of a high-risk family themselves? Or does the excess of mortality results in stronger surviving children, a biological selection? Special attention will be paid to the control of the effect of the decline of infant and child mortality during the research period.
Data for this research can be found in the Antwerp COR-database. This database, built by the Leuven research group since 2003, spans nearly eight decades (1846-1920) and covers three successive generations (birth cohorts 1820-1870). It consists of information on individuals with surnames beginning with cor* and their relatives, collected from the population registers and vital registration records of the district of Antwerp. The Antwerp district includes the port city of Antwerp, the biggest and fastest growing city in nineteenth-century industrializing Belgium, and the surrounding towns and villages (a total of 62 municipalities). This database thus consists of a unique collection of longitudinal data, containing almost 30,000 linked individual life courses, spread over a large area with both fast urbanizing and rural communities. Because of the sampling technique, the Antwerp cor-database contains an extensive amount of family ties. This feature grants the Antwerp COR-database with almost unique possibilities for intergenerational research.
|
|
|
|
