Natural gas flaring, respiratory health, and distributional effects

Although there is strong evidence that oil and natural gas development lead to decreases in local ambient air quality, there is less evidence of a causal link between these activities and human health. This paper explores the environmental health costs of burning natural gas by-products during crude oil extraction – flaring. We estimate the impact of flared natural gas on respiratory health by using quasi-random variation in upwind flaring generated by the interaction of wind patterns and natural gas processing capacity. Specifically, we construct a novel dataset to estimate the causal effect of increased upwind flaring on the monthly respiratory-related hospital visitation rate by using the number of upwind wells that are connected to a capacity-constrained natural gas processing facility as an instrument for monthly upwind flaring. We find that a 1% increase in the amount of flared natural gas in North Dakota would increase the respiratory-related hospital visitation rate by 0.73%. Furthermore, zip codes that were exposed to more than half of all flared natural gas extracted less than 20% of all resource wealth during the sample time period, and the zip codes exposed to a disproportionate amount of flaring tend to be economically-disadvantaged and communities of color. Our estimates indicate that the health costs constitute a material portion of the external cost of flaring, and therefore ought to be considered in global initiatives to reduce flaring.