Assessing the Impact of Potential Policies on Fertility in High-Fertility Countries Using Granger Causality and Bayesian Hierarchical Models

Daphne Liu , University of Washington, Seattle
Adrian Raftery, University of Washington, Seattle

The United Nations uses a Bayesian hierarchical model to create probabilistic projections of fertility. Women's education and family planning (WEFP) are two mechanisms identified as having a potentially causal effect on fertility decline. Simultaneously, WEFP can be directly impacted by public policy initiatives. We assess the causal nature of WEFP on fertility decline using a Granger causality framework and propose a conditional Bayesian hierarchical model for fertility projections that conditions on the impact of different WEFP policies on fertility decline. By considering potential policy outcomes within population projection models, we can improve fertility projections and provide an informative tool for policymakers in high-fertility countries.

See extended abstract

 Presented in Session 10. Fertility, Family Planning, Sexual Behavior & Reproductive Health 2