Mortality in twentieth-century Malawi

This dissertation is a composite of three related papers that focus on mortality in Malawi. The first paper utilizes census data to examine changes in adult mortality between 1987 and 1998. In the second paper, I use the same data to examine a cohort of women aged 45 and older at the time of the 1987 census and follow them in 1998 to assess the influence of children ever born and school attainment on their survival. The third paper employs information on household characteristics that was gathered in the two censuses to examine household socioeconomic status and its relationship to childhood mortality in Malawi. In the first paper, I use a demographic technique to infer mortality conditions directly from census age distributions and age-specific growth rates. The results show that adult mortality has increased rapidly between 1987 and 1998 with huge declines in the estimated life expectancy between ages 15 and 45 for males, and between ages 15 and 35 for females. In the second paper, I use a statistical technique to examine the survival of women in 1987 to the 1998 census. The results show that women with few children and higher schooling are more likely to survive than women with more children and low schooling. The third paper explores the extent to which housing and household characteristics, including asset possessions, could be used as a proxy for income and, consequently, socioeconomic status of households in Malawi where direct measures of income are not usually available. The results are interesting. In 1987, child mortality is low among “rich” households and high among “poor” households. However, analysis of the 1998 data shows that child mortality is higher among “rich” households and also among middle-aged women. These results are consistent with parallel analysis from the 1992 and 2000 Demographic and Health Survey data suggesting the impact of HIV/AIDS which is taking a toll on the country as evidenced by the current high HIV prevalence. This analysis underscores the importance of information on household characteristics collected in censuses for studying the relationship between socioeconomic status and other demographic outcomes