Hal Caswell , University of Amsterdam
Any individual is surrounded by a network of kin; that network develops over her lifetime as the result of mortality and fertility of the various relatives. In a famous paper, Goodman, Keyfitz, and Pullum (1974) derived the mean numbers of (female, matrilineal) kin implied by a given mortality and fertility schedule. Here, I present a new formal theory of kinship demography based on the matrix formulation of population projection. The theory provides complete age distributions, total numbers, prevalence, dependency ratios, and the experience of the death of relatives. The model easily computable, and does not require simulation. The approach rests on the observation that the kin of the focal individual form a population, and can be modelled as such. Going beyond the basic model, I also present an extension to multistate kin demography, in which both the focal individual and her kin are jointly classified by some stage variable (e.g., parity, marital status) in addition to age. The matrix formulation permits the analysis of such multistate models with the same calculations as the age-classified model. Using data from the Human Mortality Database and the Human Fertility Database, I will show as examples changes in the prevalence of dementia among kin in Japan, and changes in the parity structure of kin in Slovakia. This formal demographic model of kinship opens new possibilities for the comparative analysis of kinship.
Presented in Session 107. Flash Session Fertility, Family and the Life Course