|
|
Tommy Andersson , Department of Economic History, Lund University
This paper explores the effect on neonatal, post-neonatal, and child mortality of being a motherless child and the interaction this situation had with socio-economic status and gender in Sweden during the period 1790-1910. By being able to include interaction terms as compared to simply using controls/stratified samples, this paper contributes to existing literature by extending the analysis regarding which mechanisms that impacted the effect of maternal death. Using a longitudinal, individual level, database (SwedPop), covering large parts of the Swedish population, together with a Cox proportional hazards model, the study finds that the effect of maternal death during infancy was largest in the very early stages of life, primarily affecting children under the age of one. For those age 1-4 years, there was a lingering though largely reduced effect. These results give indications of the importance of breastfeeding for a child’s survival chances at the time. But while nursing was difficult to transfer quickly, other childcare practices could be taken over by for example a grandmother. Preferences regarding which child to take care of can however be sensed as the effect of losing one’s mother, during infancy, seems to have been significantly worse for girls than boys, indicating gender differentiated socio-cultural practices in a nation yet to be large-scale industrialized and modernized. Scattered results regarding socio-economic status indicate an effect in favour of children within non-manual occupation families, but the effect, if any, seems to have been small compared to the baseline hazard related to the loss of the mother.
Presented in Session P8. Demographic Trends, History, Data and Methods