Declining Response Rates and Emerging Challenges for Studying Sexual and Reproductive Health Using the National Survey of Family Growth
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
The National Survey of Family Growth (NSFG) has been a cornerstone of U.S. sexual and reproductive health research for five decades, providing nationally representative data on fertility, contraception, and sexual behaviors. However, recent changes to NSFG data collection raise new challenges for data quality, generalizability, and interpretation. In the 2022–2023 cycle, the NSFG implemented a multimodal design with the majority of respondents completing the survey online and experienced a dramatic decline in response rates, falling to 26.8%—nearly forty percentage points lower than in 2017–2019. In response, the National Center for Health Statistics (NCHS) advises against comparisons between the most recent data and earlier NSFG waves.Using a synthetic cohort approach, we examine whether the 2022–2023 NSFG exhibits patterns suggestive of selection bias that extend beyond challenges to comparisons over time. We assess consistency across survey waves for demographic characteristics that should remain stable within cohorts and for lifetime measures that should only increase with age. While estimates from 2011–2019 display expected stability, the 2022–2023 data show implausible shifts, including increases in socioeconomic advantage and declines in lifetime contraceptive use, even after application of NCHS-provided survey weights. These patterns suggest that weighting adjustments may be insufficient to address nonresponse bias in the most recent NSFG.Our findings reinforce existing guidance against temporal comparisons and raise additional concerns about the interpretation of cross-sectional estimates from the 2022–2023 NSFG. We discuss implications for researchers and policymakers relying on NSFG data and offer guidance for cautious use of recent estimates. More broadly, we highlight the need for renewed investment and methodological innovation to sustain high-quality national surveillance of sexual and reproductive health in an era of declining survey participation.