Death clustering in families in a longitudinal perspective (Antwerp, Belgium, 1846-1905)

In the recent literature on young age mortality, quite some attention has been devoted to the spread of deaths between families. Most infant deaths seem to cluster in a rather limited number of families, an observation which has been named ‘death clustering’. Studying infant mortality from a family perspective relocalizes the focus of explanations from individual characteristics to family traits. Family-level explanations might not only enhance our knowledge of causes of infant mortality, it will also improve our understanding of mortality differentials.
This paper aims at studying death clustering over time. The data stems from a historical, 19th century population where mortality levels are still at a constant high, but the early stages of the fertility transition have already been set off. The changing population structure in terms of family sizes affects how mortality is spread between families. Infant mortality will be studied from both a familial perspective, and a hierarchical (multilevel) structure where infant death risks are dependent from one another. By using the perspective of death clustering, the focus is shifted towards the family, enhancing an in-depth look to the relationship of infant mortality and the fertility decline.