Does a Rising Median Income Lift All Birth Weights? County Median Income Changes and Low Birth Weight Incidence Rates for Black and White Americans
Differences in low birth weight incidence (LBW) by race and place are long-standing, often embedded in enduring social ecologies where insufficient health resources are paired with an array of risk factors. Local or group-specific economic resources are known to be a fundamental contributor to these social ecologies, yet few studies have investigated how within-area changing economic conditions are linked to birth outcomes. This study examines county-level change in median income and black-white income differences as predictors of LBW incidence and LBW racial disparities. Time-varying county prevalence and black-white differences in maternal sociodemographic and health risk factors (e.g., non-marital childbearing, smoking during pregnancy) are considered as explanations for income estimates. Data come from U.S. birth records for approximately 24.8 million non-Hispanic black and white mothers with a singleton live birth (1992-2014). Records were aggregated in three-year county-period measurements for the 732 counties meeting eligibility requirements. Based on county by period fixed effects models, a $10,000 increase in county median income was associated with a reduction in LBW incidence of 2.7 per 1000 live births, and in the black-white LBW gap by 5.6 per 1000. Time-varying county maternal sociodemographic and health risks attenuated the link between median income and LBW by 72% and 31%, respectively, but not the association between median income and the racial LBW gap. Contrary to our hypothesis, conditioning on median income changes, a widening racial income difference was associated with a smaller black-white LBW gap (a finding explored in post hoc analyses). Our results suggest that, if successful in raising median income, local government efforts to stimulate economic growth and employment opportunities are likely to reduce both population incidence and black-white differences in LBW. [This draft paper is intended for review and comments only. It is not intended for citation, quotation, or other use in any form]