Date of Award

Summer 8-22-2019

Level of Access

Open-Access Thesis

Degree Name

Master of Science (MS)




Angela Daley

Second Committee Member

Mario Teisl

Third Committee Member

Caroline Noblet


There is an established body of literature on the effect of public-place smoking bans on smoking behavior. However, these studies focus on the effect of the public-place bans on the overall population and ignore racial/ethnic heterogeneity, therein. Given that there is ample evidence of racial/ethnic differences in risky health behaviors (such as smoking and drinking) in the United States, research that sheds light on this differential impact is necessary. Thus, the two studies presented in this thesis work estimates the heterogeneous racial/ethnic effects of public-place smoking bans on smoking and drinking-related behavior of U.S. adults.

In the first chapter we estimate the effect of public-place smoking bans on drinking behavior in the United States (i.e. probability of drinking, number of drinks, binge drinking occasions and instances of drunk driving) by testing for heterogeneity across those who identify as American Indian/Alaska Native, Native Hawaiian/Pacific Islander, Asian, Black/African American and Hispanic/Latinx, respectively. There is a small body of literature on the effect of public-place smoking bans on alcohol consumption, given the complementarity between smoking and drinking. We extend this literature by considering differences across ethnic subgroups. The data for this study comes from the Behavioral Risk Factor Surveillance System (2002-2014), which is a nationally representative survey of adults in the United States. Our sample consists of 453,898 individuals living in the 36 states (including Washington DC) covered by state-wide public-place bans. We use a difference-in-differences model to exploit variation in the timing of the bans across states. Results are presented separately by gender. Our findings suggest that bans have a differential impact across racial/ethnic subgroups, with Black/African American and Hispanic/Latinx men and women experiencing most of the impact.

In the second chapter, we again use data from the Behavioral Risk Factor Surveillance System (2002 to 2012) to access the racial/ethnic heterogeneous effect of the public-place smoking bans on smoking behavior. Our sample consists of 8,210 pregnant women across all 50 states (as well as Washington DC) who self-identify as White, Black/African American, Asian and Hispanic/Latinx. We do so by applying the difference-in-differences model and testing for heterogeneity across these races/ethnicities for five outcome variables: current versus former smokers; daily versus occasional smokers; occasional versus former smokers; daily versus former smokers and whether the respondent attempted to quit smoking. The results are presented separately by educational attainment level (low of high educational attainment). The results show that, while there is no significant effect of the bans when looking at the population as a whole, differences exist across racial/ethnic subgroups. The estimates suggest that the bans have led to worsening smoking behavior among Hispanic/Latinx women from the low education group while there were improvements (or no change) across women in the high education group. Both of these studies show the importance of disaggregating the impact of public-place smoking bans and similar policies by race/ethnicity to better understand the heterogeneity therein. The results also suggest that further research explaining the reason behind these differences is required.