Font Size: 

Probabilistic projection of Total Fertility Rate (TFR) for Pakistan and its regions was done using Bayesian Hierarchical modeling approach. The model for projection is a random walk model depending on current TFR level, a decrement function and a random error term, whereas the decrement function is the sum of two logistic functions.The estimates were obtained based on data collected from demographic surveys of Pakistan and from UN data base using a Markov chain Monte Carlo algorithm. Results were assessed based on coverage of countries and for different values of μ . It was revealed that as the number of countries increase the behavior of regions approaches to national level in the long run, and for smaller values of μ (less than 2)the trajectories were approaching fast to μ than for larger values.

Bayesian hierarchical model, Double logistic function, Fertility projection, Total Fertility Rate.