The long-term objectives of this project are to develop and extend methods for the projection and dynamic analysis of human populations.
The specific aims of the proposed work involve three main areas of inquiry. First, the dynamics of a parity-progression model of birth projection will be analyzed. This continuous-time model is built upon distributions of the time of birth for women who are classified by parity, i.e., by the number of children already born. A demographic sensitivity analysis of this model will be done to examine changes in growth rate, transient properties, and age-composition given arbitrary changes in mean and variance of timing of first and later children, and changes in parity progression ratios. Second, several essential aspects of stochastic demographic models will be analyzed. The theory of stochastic reproductive value will be developed and the mean and variances in the stochastic model used to characterize the effects of stochasticity on future generations. A demographic sensitivity analysis will be performed on growth rate and age distribution of stochastic models given arbitrary changes in birth and death schedules. Finally some exactly soluble and approximately soluble analytical models will be explored to develop further insight into the properties of stochastic models. Third, the robustness of an internally driven dynamic analysis of the Easterlin effect will be examined. The main question to be considered is the robustness of cyclical dynamics given a blurring of the precise age composition of a cohort. The proposed work will treat the effective birth cohort affecting population fertility as a sum over actual births in a finite time interval around the time of birth.
The aim i s to see how the resulting averaging will affect intrinsic cycles.
