Projecting global educational attainment to 2050 using a hierarchical Bayesian model
Bilal Barakat, Vienna Institute of Demography
There has in recent years been an increased interest in applying hierarchical Bayesian models to the problem of projecting trends in demographic parameters, such as TFR, urbanization, or disease prevalence. Using an inverse probit model for the growth paths of educational expansion, I obtain probabilistic projections of educational attainment for 120 countries for the period up to 2050 that take into account both country-specific, and regional and global trends, using Markov Chain Monte Carlo methods. The study goes in a different direction than past applications of the method by subjecting these projections to expert evaluation, in order to investigate how the results of the statistical projection model differ from qualitative expectations of national education specialists.
Presented in Poster Session 1