Content area
Abstract
The analytical framework proposed by Davis and Blake (1956) divides the process of reproduction into three segments; (i) exposure to the risk of pregnancy, (ii) the ability to conceive and (iii) successful gestation. This dissertation is concerned with the first segment.
In this thesis, data from the 1973 National Demographic Survey of Tanzania are used to investigate the role of marriage behavior in determining fertility levels. The level of the exposure variable is apportioned between the prevalence and age patterns of first marriages, marital dissolutions and remarriages. Indices are developed to represent each of the three phenomena. Techniques for obtaining detailed information on the process of marital dissolution and subsequent remarriage are presented. The proportions ever dissolving a union, proportions ever remarrying and the intermarriage durations for those who remarry, are obtained from classifications of women according to their marital statuses. The potential effects of changes in nuptiality patterns on fertility are discussed. Characteristics of polygyny and their effects on maternal fertility are also investigated.
The evidence of these data is that the effects on fertility of high incidence rates of marital dissolution are largely compensated for by equally high remarriage rates and short intermarriage durations. Further, because unmarried women exhibit fairly high fertility rates, the potential reductions in fertility through changes in marriage behavior (e.g. later marriage) are minimal. Women entering polygamous unions are self-selected on the basis of their fertility experience and are linked to older men in such unions.
