Beyond Boom and Bust: Heterogeneous Fertility Effects of the COVID-19 Pandemic

PROJECT SUMMARY The COVID-19 pandemic arrived at a time when U.S. birth rates had been declining for several years. Contrary to initial predictions of a "COVID baby boom," early evidence shows that the pandemic coincided with a steeper decline in birth rates. But fertility patterns vary dramatically across the U.S., and the mortality and health effects of COVID-19 have also varied by location, race/ethnicity, and socioeconomic status. Therefore, we expect significant variation in the pandemic's effect on fertility. However, currently available data make this difficult to study - fertility rates by household income are not available in the U.S., even at the national level, nor are rate schedules by combinations of race/ethnicity and income or parity. This project seeks to leverage a new restricted data source, which our team has produced under the NICHD-funded "Increased access to contraception: an opportunity dividend?" project (R01-HD101480-01, 2020-2024), to assess the effect of the COVID-19 pandemic on fertility. In partnership with the U.S. Census Bureau, we will use a dataset we created, Reproduction in People's Lives (RIPL), which provides the full count individual-level longitudinal data needed to study how changes in fertility during the COVID-19 pandemic have varied by age, state, race and ethnicity, household income, and birth order. Using this dataset, we will link multiple U.S. administrative data sources, including individual tax filings microdata, the 2010 decennial Census, and social security data, to (Aim 1) calculate age-specific fertility rates for all U.S. women by state and demographic subgroup for the years 2015 through 2021. These rates - calculated by race/ethnicity, household income, parity, and most combinations of these characteristics - represent a significant improvement in level of detail over current publicly available rates. Using the rates calculated for the pre-pandemic period (2015-2019) we will (Aim 2) use demographic forecasting to generate counterfactual rates that represent estimated 2020-2021 fertility levels by state and demographic subgroup in the absence of the COVID-19 pandemic. We use a forecasting method that is demonstrated to perform well over the short- to medium-term in fertility contexts like the contemporary U.S. and which is based on the Lee-Carter method, a cornerstone of demographic forecasting. We will (Aim 3) assess heterogeneity in the COVID-19 pandemic's effect on fertility by comparing these counterfactuals to the observed rates constructed under Aim 1 during the years 2020-2021. The forecast counterfactuals allow us to estimate the portion of fertility change over the pandemic period that is the result of the pandemic, and to compare this portion across different age groups, sociodemographic subgroups, parities, and states. We anticipate that we will also be able to publicly release a subset of the rate schedules and forecasts we generate, allowing other researchers access to this important new data source.