Despite widespread adoption across the United States, there is little empirical evidence on whether Medicaid managed care (MMC) reduces costs without compromising health outcomes. This paper exploits an arbitrary component of MMC enrollment in New York to examine the causal effects of MMC on hospital responses and newborn health. During the study period, infants with birth weight below 1,200 grams were excluded from MMC and were instead served through the traditional fee-for-service (FFS) system. Using a regression discontinuity design framework, I find that infants enrolled in MMC stay fewer days in the hospital following birth and thus have less expensive visits. I show that this is driven by birth hospitals retaining more infants enrolled in FFS while transferring away those enrolled in MMC to another hospital. The effects are stronger when the birth hospital is spatially constrained and for infants with high predicted list prices. I find little effect on short-term health measured by mortality during hospitalization and the incidence of readmissions. I show that hospitals engage in these behaviors only when they have a high-quality hospital in close proximity, suggesting that hospitals do not sacrifice quality of care for cost reductions.