THE IMPACT OF COVID-19 ON U.S. BIRTH OUTCOMES ABSTRACT The COVID-19 pandemic will shape most domains of life in the US. Among the most enduring will be effects on the health of the next generation of Americans. Over 4 million pregnancies are currently in progress in the US. Between 1-3 million pregnancies will be initiated over the next six months. Prenatal exposure to large- scale health and economic events have serious implications for birth outcomes, including delivery conditions, pregnancy duration, and birthweight. These outcomes, in turn, predict welfare throughout life—education, earnings, and even lifespan. Pandemic effects on pregnancy will have implications that last decades. Currently we have little systematic population-level evidence on newborn health with which to make policy decisions about COVID management. Evidence is urgently needed because COVID-19 has shaped well-established predictors of pregnancy health, including infection exposure, stress, economic precarity, and health care access. These individual-level factors are shaped by community-level dynamics like infection spread, distancing policies, and economic decline. Because these factors differ across the U.S., the impact of COVID-19 on birth outcomes is likely stronger among vulnerable groups defined by location, race, ethnicity, and socioeconomic disadvantage, potentially exacerbating inequality in infant health. This research estimates the effects of COVID-19 on birth outcomes at the population level, over time, and across groups defined by different sources of disadvantage. To provide the earliest possible evidence, we use birth records obtained at the state level with early release. We focus on six states that provide large and diverse samples in areas in which the pandemic unfolded with significant variation: Alabama, Arizona, Florida, Hawaii, New Jersey, and Wisconsin. The analysis uses causal inference techniques and explicitly incorporates changes to fertility and the composition of pregnancies. Aim 1 estimates the effect of local-level exposure to COVID-19 on birth outcomes, including intrauterine growth restriction, birthweight, and delivery complications. Aim 2 tests for differences in COVID-19 effects on birth outcomes by maternal age, race, nativity, economic disadvantage, and pre-pandemic, local-level economic conditions and health care infrastructure. Establishing across-group differences is critical for understanding how the pandemic will shape early-life inequality. Aim 3 addresses a serious form of bias in prenatal exposure estimates—selection into birth—by modeling the change in the number and composition of women giving birth as a result of COVID-19 exposure. We use this information to adjust estimates of COVID-19 effects on infant health, while also providing the earliest evidence of COVID effects on population fertility. The research provides a rapid, comprehensive assessment of the impact on birth outcomes across localities in which the pandemic u...