Model Inputs \(\rightarrow\) Demographic Data \(\rightarrow\) Population Projections
We obtained country-specific population projections from the 2022 UN World Population Prospects.[1] We used annual, age-specific estimates of the female population to model each country’s population from 1985 to 2021. To project forward to 2050 we used the probabilistic population projections (PPP) to estimate the number of women in the population from 2022 onwards. Annual age-specific projections from the deterministic Medium Fertility scenario were used to construct the projected age structure (by single ages).
For computational efficiency we do not simulate the entire population of women in each country. Instead we specify a scalar and simulate a proportion of the female population (sampled at random) and re-weight our modeled estimates back to the total population size. To improve the stability of modeled estimates for small countries we specify a minimum population size of 10,000 women to simulate in a country and adjust the country-specific population scalar accordingly.
We obtained projections of the urban proportion of each country’s population from the UN Urbanization Prospects 2018.[2] Due to lack of data we assumed that these projections were representative of males and females of all ages. This approach may thus overestimate the urban proportion at older ages in countries where urbanization is driven by younger people. However, given that the population of interest in our model is women of reproductive age this assumption is unlikely to substantially impact our estimates.
GMatH (Global Maternal Health) Model - Last updated: 28 November 2022
© Copyright 2020-2022 Zachary J. Ward
zward@hsph.harvard.edu