Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

Bayesian Fertility Rate Projections

About This Project

Fertility rates in the United States and other developed countries have been declining for several decades. Understanding these trends and projecting future fertility patterns is crucial for policy planning in areas such as education, healthcare, and social security. This project uses Bayesian statistical methods to model historical fertility patterns and project future cohort fertility rates using data from the US Census Current Population Survey (CPS).

Motivation

Cohort fertility rates measure the average number of children born to women in a specific birth cohort over their reproductive lifetime. Unlike period fertility rates (which measure births in a given year), cohort fertility rates better capture long-term trends and can reveal generational shifts in reproductive behavior.

Recent trends suggest that cohorts born after 1980 may experience substantially lower lifetime fertility than previous generations. This has important implications for:

Key Questions

This project addresses several key questions:

  1. What are the long-term fertility trends for recent birth cohorts?

  2. How low might fertility rates go for cohorts currently in their reproductive years?

  3. Can we reliably project future fertility using historical patterns?

  4. How well do model predictions validate against historical Census data?

Contents

Methodology Overview

Our approach uses a Bayesian hierarchical log-linear model to capture age-specific fertility patterns across birth cohorts:

The model accounts for survey weights through resampling, ensuring that results reflect the actual US population rather than just the survey respondents.

Key Findings

Preliminary results suggest:

For detailed findings, see the Technical Report.

Data Sources

About the Model

The Bayesian approach provides several advantages:


This project is a work in progress. Results are preliminary and subject to revision.