Just back from the 2013 Labour Econometrics Workshop at the Melbourne Institute. Lot of interesting papers, including this one of Cesur and Ulker on the enormous reduction in infant mortality that followed switches at the household level from coal to gas in Turkey. They find that s one percent increase in the uptake of natural gas in households led to a 4.4 percent decrease in infant mortality. Hurray for gas, one then thinks! Their abstract:
One of the consequences of rapid economic growth and industrialization in the developing world has been deterioration in environmental conditions and air quality. While air pollution is a serious threat to health in most developing countries, environmental regulations are rare and the determination to address the problem is weak due to ongoing pressures to sustain robust economic growth. Under these constraints, natural gas, as a clean, abundant, and highly-efficient source of energy, has emerged as an increasingly attractive source of fuel, which could address some of the environmental and health challenges faced by these countries without undermining their economies. In this paper, we examine the impact of air pollution on infant mortality in Turkey using variation across provinces and over time in the adoption of natural gas as a cleaner fuel. Our results indicate that the expansion of natural gas infrastructure has caused a significant decrease in the rate of infant mortality in Turkey. In particular, a one-percentage point increase in the rate of subscriptions to natural gas services would cause the infant mortality rate to decline by about 4 percent, which could result in 383 infant lives saved in 2011 alone. These results are robust to a large number of specifications. Finally, we use supplemental data on total particulate matter and sulfur dioxide to produce direct estimates of the effects of these pollutants on infant mortality using natural gas expansion as an instrument. Our elasticity estimates from the instrumental variable analysis are 1.29 for particulate matter and 0.65 for sulfur dioxide.