Search

"This article describes the main ingredients of a methodology currently under development for projecting families [in Canada]--these groups of individuals linked by conjugal ties or by blood--according to their principal characteristics: family type and number/ages of children. The crux of this methodology is a multidimensional model which can be viewed as an extension of the traditional cohort-components model that incorporates the conjugal and parity statuses of women. With an appropriate generalization of the conjugal status variable and the addition of an appendage for dealing with the special case of step families, this model eventually allows one to shift from the number of women according to demographic status (stemming from cross-classifying conjugal status and numbers of children present) to those families according to their main characteristics." (SUMMARY IN ENG AND SPA), (excerpt)

"Users of multistate life tables and projections have recognized that the Markovian assumptions underlying such models are unduly restrictive and should be relaxed whenever data permit. Efforts to include the influences of previous occupancies have included the incorporation of place-of-birth dependence. This paper addresses the stable growth properties of such generalized multistate models. It shows how place-of-birth-specific stable growth measures can be calculated without projection simply by solving the characteristic equation. An example using Canadian data illustrates the argument." (SUMMARY IN FRE), (excerpt)