Socio-Economic and Cultural Differentials in the Mortality of Sub-Saharan Africa
Citation: Tabutin, Akoto (1992) Socio-Economic and Cultural Differentials in the Mortality of Sub-Saharan Africa.
Introduction: This paper reviews evidence on socioeconomic differentials in child mortality in sub-Saharan Africa and conducts an analysis of World Fertility Survey data.
(1) Education: "Everywhere, mortality falls as expected, more or less strongly and more or less rapidly, with an increase in the mother's education" (p. 33). However, the effects of education also vary by country and age of child-educational differentials in mortality are greatest between the ages of 1-5. Father's education also has an effect, but it is typically not as pronounced as mother's education. Interestingly, in Nigeria mother's education can mitigate the deleterious effects of low income, while illiterate mothers with high income suffer relatively high child mortality.
(2) Occupation: In Rwanda, "The important factor in explaining the level of mortality is whether or not the father is in the agricultural category" (p. 40). This basically holds true for the other countries as well, with agricultural workers exhibiting relatively high child mortality, and administrative/professional workers exhibiting relatively low child mortality.
(3) Region: Even after controlling for education, large regional differences exist in all countries. This may point to resource endowments or cultural differences.
(4) Place of Residence: "Overall, in sub-Saharan Africa, large cities enjoy the best situation, with the exceptions of Kenya and Ghana, where child mortality in small cities is appreciably less than in metropolitan areas" (p. 45). Rural areas generally have very high excess mortality, presumably due to a lack of resources and medical care. Although cities enjoy an mortality advantage, high rates of inequalities exist within cities. Importantly, "Education does not really play a role except in privileged and well-provided zones" (p. 48).
(5) Effect of Type of Economic Activity of the Wife: Except for Ghana, "female family workers showed a clearly excess mortality" (p. 49). However, when women worked in clerical positions, their children benefitted from lower mortality than non-working women.
(6) Religion: Catholics and Protestants exhibit lower mortality rates than Muslims, who in turn exhibit lower rates than adherents to traditional faiths. However, the introduction of control variables eliminates much of these differences.
(7) Ethnicity: Even when control variables are introduced, the effects of ethnicity are strong in all nations.
Analysis: A Multiple Classification Analysis of WFS data for Kenya and Cameroon is performed in order to determine the explanatory force of the aforementioned variables. Overall, ethnicity has the most explanatory power (33% in Cameroon and 39% in Kenya), followed by mother's education and father's occupation. Religion is the least significant variable. But some differences were observed between Kenya and Cameroon. For instance, education remained significant in Cameroon after the introduction of control variables, but in Kenya education was determined to be largely a function of place of residence.
Conclusions: "Of the variables that we used, ethnic group, maternal education, occupation and, to a lesser degree, the place of residence, turned out to be the most important. . . The ethnic group is clearly a variable that must be controlled in studies of mortality as well as in studies of fertility in Africa" (p. 62).