UC Davis

Environmental and epidemiological research has linked exposure to air pollution and extreme temperatures during the prenatal period to the incidence of fetal and infant death, adverse birth outcomes, and worse outcomes in adulthood and across generations. To minimize these adverse effects, policymakers can reduce ambient exposures (i.e., mitigation) or intervene with some program that lessens the impacts (i.e., adaptation). The first two chapters focus on the effects of air pollution shocks and the efficacy of free prenatal care to lessen its adverse effects on health at birth. The last chapter focuses on the effects of extreme temperatures and the efficacy of free prenatal care and air conditioning to lessen its adverse effects on birth outcomes (e.g., birthweight, weeks of gestation, low birthweight rate).

The first chapter shows that the estimates of sulfur dioxide (SO2) effects on birth outcomes are susceptible to the window used to measure exposure during the prenatal period. Measuring exposure from conception to birth, I find a negative impact of SO2 on birthweight. In contrast, the estimate is positive when exposure is measured from conception to 39 weeks. Using each county’s 52-week lagged SO2 concentrations as a placebo, I find that using a fixed 39-week window from the date of conception is the most reliable methodology. However, this methodology's estimates indicate that higher SO2 concentrations increase birth weight. I present evidence suggesting that this counterintuitive result is caused by livebirth bias (i.e., the infants that survive pollution shocks are positively selected). I overcome this problem by using the number of infants born with non-adverse outcomes per woman of reproductive age as the dependent variable instead of traditional outcomes (e.g., birthweight, low birthweight, or preterm birth rate). Applying this transformation, I find that SO2 worsens health at birth, and its effects increase with the pollutant’s concentration (i.e., the SO2-birth outcome damage function is convex). Furthermore, the effects are more prominent for blacks than whites.

The second chapter tests whether access to free prenatal care lessens the adverse health effects of exposure to air pollution in utero. I study how the expansion of Medicaid (publicly-provided health insurance for low-income households) changed the effect of prenatal exposure to SO2 on fetal death and birth outcomes. Theoretically, the effect is ambiguous: Even if free prenatal supplementation (i.e., vitamins, iron, calcium) lessens the biological impact of air pollution, there could be a substitution between access to prenatal care and pollution avoidance. High SO2 concentrations increased fetal deaths, and Medicaid’s expansion attenuated this effect. Estimating the impact of Medicaid on the SO2-birth outcome relationship is empirically challenging because the infants marginally saved by Medicaid could be positively or negatively selected. The analysis of traditional outcome variables (e.g., birthweight, low birthweight rate) suggests that Medicaid had no impact or even intensified the damage of SO2 on health at birth. To account for the possibility of livebirth (i.e, sample selection) bias, I instead analyzed the number of non-low birth weight (i.e., healthy) infants per woman of reproductive age (nlbw/w). Using this dependent variable, I find that Medicaid mitigated the effect of SO2 on nlbw/w in low-pollution areas and at the national level. Furthermore, the reduction was larger for blacks than whites; thus, Medicaid improved environmental justice in the US by shrinking the gap in the health effects of in-utero air pollution between races.

The third chapter tests whether access to free prenatal care and air conditioning lessens the adverse health effects of extreme temperatures in a non-rural setting. I study how the expansion of Medicaid changed the effect of extreme in-utero temperatures on birth outcomes in the US. In developed countries, physiological stress is the primary mechanism through which temperature affects pregnancy outcomes. In rural areas of the developing world, it can also do so indirectly through changes in real income, increased incidence of maternal disease, or increased conflicts. The results suggest that access to prenatal care did not lessen the impacts of extreme temperatures on birth outcomes. However, the diffusion of air conditioning reduced the effects of extremely hot days.

Overall, the results of these chapters suggest that providing low-income women with free prenatal care is a promising intervention to lessen the health impacts of in-utero air pollution but not those of extreme temperatures. On the other hand, air conditioning is a promising intervention to lessen the health effects of extreme heat on birth outcomes.