Modeling Global Historical Trends in Conditional Unintended Birth Rates by Moving towards a More Specific Definition of Fertility Intention

Jessica Rosenberg, Guttmacher Institute
Jewel Gausman, Guttmacher Institute
Jonathan Bearak, Guttmacher Insitute

Estimating unintended fertility is critical to quantifying women’s success in achieving their fertility goals. Thus far, studies estimating unintended births have used the number of reproductive-age women as the denominator for their rates. However, recent research estimating unintended pregnancy rates argues that “conditional” rates, that use for their denominator the number of women at risk of unintended fertility, better proxy women’s success in implementing their preferences. In this study, we estimate conditional unintended birth rates (CUBRs) in LMICs for the first time, producing estimates for every five-year period from 1985 to 2019 using a Bayesian hierarchical time series model. We report estimates using a new denominator we developed that addresses limitations with the contraceptive need construct previously used as a proxy for women wanting to avoid pregnancy. We find substantial geographic variation, with rates largest in Sub-Saharan Africa regardless of time period. We find declines in CUBRs across time across all geographic regions, with the largest declines in Southeast and South Asia. For our preliminary results, we estimated our model using 222 DHS datasets from 75 countries. Before IPC, we will reestimate our model using MICS datasets and recently released DHS datasets to expand coverage and produce estimates for 2020-24.

Keywords: Bayesian methods , Fertility, Sexual and Reproductive Health and Rights, Family Planning and Contraception

See extended abstract.