Social
Differences in the Relationship Between Voter Registration and Birth Outcomes by Race in South Carolina Michael P. Huynh* Michael Huynh Woo Jung (Amy) Lee Nancy L. Fleischer Abigail Kappelman Yanmei Xie Annie Ro
Background. Political factors are an untapped social determinant of health, and little is known about their influence on birth outcomes, including differences by race. Greater levels of voter registration may be indicative of more agency and health equity. We examined the link between voter registration and birth outcomes, with effect modification by race. Methods. We linked South Carolina (SC) State Election Commission data to SC singleton birth certificate data from 2004-2020 (n=871,709). Our exposure was a non-White to White (NW-W) ratio of county-level voter registration. We divided counts of registered voters by those eligible to vote at the county level. Outcomes were low birthweight (LBW, <2,500g) and preterm birth (PTB, <37 weeks). We used multilevel logistic regression models to examine the county-level relationship between voter registration and birth outcomes, adjusting for infant sex, maternal age, parity, maternal education, Medicaid insurance, presidential term years, polling sites, and urbanicity. We ran effect modification, then stratified analysis by race. Results. Our sample included 57.5% White, 32.9% Black, and 9.6% Hispanic moms. LBW prevalence was 7.9%, and PTB prevalence was 9.6%. On average, more White voters were registered than non-White voters (NW-W ratio=0.93). A higher NW-W ratio (i.e, a larger share of non-White voter registration) was associated with higher odds of LBW for Hispanic moms (OR=1.83, 1.11-3.03), with no association for White and Black moms. A higher NW-W ratio was associated with lower odds of PTB for White moms (OR=0.78, 0.61-0.99) and Black moms (OR=0.75, 0.59-0.96), with no association for Hispanic moms. Conclusion. Racial differences in voter registration had mixed associations with adverse birth outcomes in SC. Policy-level programs targeted at improving civic engagement may help improve agency and birth outcomes, but these mechanisms need closer investigation into what factors explain the possible racial differences.