Child mortality in the Spanish Civil War. Geographical impact estimated from 1940 census retrospective fertility data

Albert Garcia Soler, Universitat Autònoma de Barcelona
Fernando Gil-Alonso, Universitat de Barcelona

Following an indirect estimation method proposed by W. Brass, this paper uses 1940 Spanish census data on retrospective fertility (questions done to ever married women on the number of children ever born and their survival at the time of the census) to estimate women’s descendants survival levels at different ages, and therefore child mortality up to the age of 15. As data was classified by province and by mothers’ marriage duration, we have been able to reconstruct child survival territorial evolution of at different ages and at diverse moments, more specifically before and during the Spanish Civil War (1936-1939). Although child survival levels had already significantly increased between 1918 (Spanish flu) and 1936, important differences among provinces still existed. The initial hypothesis would be that more developed, industrialised and urban provinces, with better health infrastructure and lower child mortality, would be better prepared to cope with the shock that the conflict meant for child survival. Preliminary results however show that even though the war worsened child survival in practically all the provinces, the least affected were Catalan and Northern Meseta ones, which showed opposite child mortality levels before the war -i.e. low in the former and very high ones in the latter. This would mean that, although the regional development level should be inversely related to the vulnerability of its population to unfavourable situations, child mortality during the war did not seem to be significantly affected by the previous demographic situation. Other factors such as access to food, crop availability, and logistic aspects related to its storage, transport and distribution, that is to say, those regarding rearguard organisation in general, seem to be more significant in explaining differences in child mortality.

  See extended abstract

Presented in Poster Session 2