An Assessment of the Method of Adult Mortality Estimation from Information on Widowhood

Abstract

In less developed countries, both traditional and non-traditional methods of mortality estimation have not been successful in providing complete and accurate information on mortality. To substitute these methods Demographers have tried to develop indirect methods of mortality estimation. For estimating adult mortality two methods based on orphanhood data and widowhood data are being developed. There is a fair literature on the theoretical background and practical application of these methods and by critically reviewing this scattered literature, the suitability and applicability of them can be assessed. The information for the method of widowhood can be collected by asking the simple question “Is your first husband/ wife alive”. The proportion can be converted to life table probabilities using model schedules of mortality, fertility, and nuptiality. The robustness of the method depends largely of the assumptions of the method. Some assumptions are not crucial. However the assumption on the independent of the mortality risks of respondent and spouse should be examined critically. Further development of the method should be carried out using computer simulation rather than depending on plausible estimates achieved by application of the method.