Behavioral Economic Approaches to Understanding Maternal Smoking in Appalachia

Appalachia is a federally recognized region of reduced economic development in the eastern United States. The region is defined as a specific set of contiguous counties lying in the Appalachian mountain range. Residents of Appalachia often experience worse health outcomes when compared to the nation as a whole. Perhaps one of the most alarming statistics for Appalachia is the rate of maternal smoking given the well-known health risks for unborn and newborn babies. Maternal smoking rates in Appalachia (19.5 percent) and in West Virginia (26 percent) are alarming higher than in the rest of the country (8.8 percent).

A large literature examines smoking from an economic perspective; the effect of smoking on infant health; and the effect of policies aimed at reducing maternal smoking on smoking behavior and infant health. Less attention has been paid to the determinants of maternal smoking. We begin to close this gap by documenting the rate of maternal smoking and investigating the determinants of and disparities in maternal smoking. We estimate reduced form regression models to investigate the determinants of maternal smoking using data from the Center for Disease Control and Prevention's (CDC) Natality  files. The models control for observable factors linked to smoking, like the mother's age, race, level of education, and marital status. Even after controlling for these observable characteristics, and state-level tobacco control policies like cigarette taxes and smoking bans, the disparity in maternal smoking in Appalachia persists.

Explaining the high level of maternal smoking in Appalachia is an important research question that, so far, eludes simple answers. Despite excise taxes and bans, widespread informational campaigns about the dangers of cigarette smoking, and pre-natal care aimed at promoting healthy infants, pregnant women continue to smoke. Since educational campaigns, excise taxes, and smoking bans in workplaces, bars, and restaurants do not completely deter maternal smoking, especially in Appalachia, we posit that behavioral economic approaches to understanding why pregnant women smoke represent a potentially fruitful avenue for further research.

We empirically explore three potential behavioral economic concepts that might help to explain the high rates of maternal smoking in Appalachia: present bias; projection bias; and herd behavior. Present bias is the idea that consumers incorrectly discount future costs and benefits back to present value when making decisions. A pregnant woman with present bias would not fully account for the unborn child's future health problems due to smoking while pregnant since the benefits of smoking are immediate and the costs are in the future. When individuals experience projection bias, they recognize that they experience present bias, but fail to properly account for how they will behave when the future becomes the present. Herd behavior refers to the idea that decision making is influenced by what others around the decision maker are doing. If a pregnant women sees others around her smoking while pregnant and sees that the children did not experience severe health problems, then she may conclude that smoking while pregnant is not harmful.

Key words: maternal smoking; natality data; present bias; herd behavior