In the developing countries that account for 80 percent of the world population, the mortality of infants and children is a great social, psychological, and economic burden. All of the resources invested in infants from conception to their deaths, including the participation of mothers in economic production, such as subsistence farming and/or paid work, is lost. Any governmental resources devoted to health care or education are also lost. The mortality of children under age 1 is 10 times higher in developing countries than in developed ones. It is 50 times higher in Liberia (15 percent) than in Finland (0.3 percent). Similarly, the total mortality of children under age 5 is 10 times higher in developing countries. It is close to 25 percent in some countries of sub-Saharan Africa but around five percent in many European and some Asian countries.
The economic and medical aspects of this situation are fairly clear. Unsanitary conditions attributable to poverty, ignorance, and lack of government investment in health or infrastructure result in gastrointestinal and malarial disease, which are particularly deadly to infants and children. But while we know something about broad correlations between mortality rates and economic and epidemiological circumstances on average at the population level, we know almost nothing in such countries about the circumstances of individual families; thus we cannot account for variation within populations. The necessary data are very difficult to obtain. There is excellent detailed information from ethnographic observation, but the sample sizes are limited, and the observations are often anecdotal. In particular, we know almost nothing about the way in which the composition of families and extended families affects the survival of children. That kind of information is almost never recorded in health surveys.
Therefore we turn to history and particularly parish records of baptism, marriage, and burial, which allow the reconstitution of families and extended families. They give information almost impossible to obtain from modern surveys. This project analyzes a particularly good set of such data for the peasants of part of Croatia in the 18th and 19th centuries. These peasants lived under economic and political conditions similar to those of developing countries today. Previous research on maternal mortality showed that economic and political conditions had strong effects on the chance of death in childbirth, by drawing male labor away from the family farm, putting labor burdens on the women, and affecting the level of care that could be devoted to new mothers. Previous research also showed that composition of the extended family was an important factor. The larger was this support group, the lower was the risk of death. However, the family system was one in which the wives of brothers were in strong competition, and the data show that the more sisters-in-law a new mother had, the higher was her risk of death, given the overall size of the family network. Similarly, the more senior a brother was, the better were his wife's survival chances. Thus internal family politics affected survival of mothers as much as overall economic power and productivity.
In this research on infant and child mortality, the researchers will apply the lessons from the analysis of maternal mortality, using the same population database and the same statistical techniques. They will take account of changing economic and political conditions (such as the abolition of serfdom, warfare and military recruitment, increasing intrusion of cash into the peasant economy), as well as family size and structure. They will also be able to take into account the prior child mortality history of individual mothers, and of other mothers in their kinship network. They expect to find similar results of broader conditions and of details of family network composition.
From a scientific perspective, this analysis should encourage better collection of more detailed data and attention to ethnographic relevance in contemporary health surveys. Adding a few questions on the composition of households and extended families would be simple, but potentially very productive.
The broader impacts of the research are clear: the lessons learned from the historical analysis should be useful in attacking problems of infant and child mortality in the modern world. In developing countries, men are increasingly drawn away from home as wage labor, and women live in an extended family context, often in a polygamous context where rivalry is to be expected. These conditions are similar to those in this historical database. Current efforts by organizations such as UNICEF, USAID, and NGOs to cut rates of infant and child mortality would be aided by closer understanding of factors often ignored, such as the politics of survival within kinship groups.
