4834628877_73b0920702_b.jpg
VoxEU Column Health Economics

The impacts over time of smoke-free air ordinances in Texas

Progress in adopting smoking bans across the US has been slow, despite a majority of Americans supporting a ban in public places. This column uses aggregate and establishment-level data from Texas to examine the economic effects of smoking bans on bars and restaurants. The results suggest that bars and restaurants are not adversely affected by the adoption of a ban. 

Many communities in the US are not smoke free. According to a recent report from the Center for Disease Control and Prevention, as of December 2015, only half of the US population is protected by comprehensive smoking bans in workplaces, restaurants, and bars (Tynan et al. 2016).1 The same report highlights that most of the gains in protections were made in the period 2000-2010, where the share of the US population protected increases from 2.72% in 2000 to 47.8% in 2010. Nonetheless, little progress has been made in protecting the population since 2010. Indeed, by December of 2015 the share of the US population protected by comprehensive bans had increased to 49.6%, an increase of only 1.8% from the 2010 level. The report highlights important disparities in coverage. For example, no state in the southeast has comprehensive legislation, and in 8 of the 24 states that lack legislation, states pre-empt local municipalities from enacting smoking bans.  

What could explain the slowing progress in creating smoke-free environments in recent years?  The lack of progress is not motivated by a lack of public support for smoking bans. Since 2011 a majority of Americans have supported bans of smoking in public places (Gallup 2015). As a response to changing public attitudes many businesses have responded by banning smoking on company premises.2 The vast majority of literature that estimates the impact of smoking bans on the profitability of bars and restaurants across the US finds no significant impact (Eriksen and Chaloupka 2007, IARC 2009, Scollo et al. 2003). However, several studies find that smoking bans may adversely affect bar employment, and alcohol sales among bars and restaurants (Adams and Cotti 2007, Clower and Weinstein 2004).  A previously untested hypothesis in this literature posits that null results in early studies examining the economic impacts of smoking bans were driven by sample selection. Early adopters could better absorb the shock of bans, but among worse performing late adopters, bans would adversely impact bars and restaurants.

In a new paper, we utilise aggregate and establishment level data from the state of Texas for the period 2002 through 2011 to examine the economic effects of smoking bans on bars and restaurants (Nikaj et al. 2016). The state of Texas lacks a comprehensive state-wide policy, and only 80 out of 1,209 Texas municipal governments have adopted smoke free ordinances in bars and restaurant as of 2016 (ANRF 2016). We exploit the variation in timing of policy adoption at the local level (municipality) and utilise a difference-in-differences methodology to identify the causal impact of smoking bans on bar and restaurant sales and alcohol tax expenditures. We compare outcomes between early and late adopters, and track the adjustment trajectories that sales and establishments experience after policy implementation.

There are several methodological concerns that can lead to spurious results in differences-in-differences estimates. While we detail these threats in our working paper, the overarching concern is that timing of ban implementation is chosen so as to minimise the impact of the policy.  For example, municipal governments may choose to institute such policies when bar and restaurant revenues are robust. Fleck and Hanssen (2008) show that negative trends in restaurant sales prior to the implementation of the state-wide ban in California could account for a share of the negative impact that was erroneously attributed to the ban. Unfortunately, failure to control for trends in the outcome prior to policy adoption may lead to wrong inference of causality, where such inference may not be possible.

In our analysis we utilise the variation in timing of policy adoption, but directly test the exogeneity of policy adoption by tracking trends in our outcome variables right before policy implementation. We find that municipalities that adopt smoking bans exhibit higher than average sales prior to policy adoption, suggesting that studies that do not account for policy selection likely produce biased estimates of the policy impact.

Our aggregate results that track restaurant and bar sales suggest that bars and restaurants are unaffected by bans of smoking on premises. We conduct several robustness checks to make sure our results are not driven by sample selection. Furthermore, we check to see if alcohol sales in liquor store are affected by such bans. Theoretically, if smoking bans in bars and restaurants reduce clientele among smokers, but do not affect clientele among non-smokers, then we should observe that smokers are pushed to substitutes for drinking in bars or restaurants – such as purchasing alcohol from liquor stores. In our case, this would suggest that demand for alcohol in liquor stores increases after ban implementation. We find smoking bans in bars and restaurants do not affect sales of liquor stores.

One concern with analyses utilising aggregate data is that often the estimates are imprecise. To address this concern, we turn our attention to alcohol tax expenditures at the establishment level, and conduct analyses among a sample of 28,000 establishments. Our establishment level analyses account for establishment-level unobservables and trends within them, thus reducing concerns over omitted variable bias. Moreover, the large sample of establishments generates very precise estimates of the impact of smoking bans. Even though alcohol tax expenditures are expected to be affected disproportionally from smoking bans (Adams and Cotti 2007, 2008),3 we find that reductions in tax expenditures from smoking bans are small. More particularly, when we account for threats to identification in difference-in-differences estimation we find the effects of smoking bans have no statistically significant effects on alcohol tax expenditures based on a two-tailed test at conventional significance levels of 5%. Our results are only significant at the 10% significance level, a finding that is surprising given the large sample size. Taken together, our aggregate and establishment level results imply smoking bans do not adversely affect restaurants and bars.

Finally, when we compare outcomes and adjustment trajectories among early and late adopters of smoking bans, we find no evidence that late adopters do worse. On the contrary, we find that late adopters were able to adjust to changes in policy better, with no long-term impacts due to policy adoption. One implication of our over-time analysis is that the many municipalities that are currently considering joining the group of adopters will not experience negative impacts due to adoption. To our knowledge our analysis is the first to look at the long-term impacts of such policies on bar and restaurant sales, and it is only the second study that estimates results at the establishment level.

Authors’ note: The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research, US Department of Housing and Urban Development, the University of Illinois at Chicago, or Texas Christian University.

References

Adams, S, and C D Cotti (2007), "The Effect of Smoking Bans on Bars and Restaurants: An Analysis of Changes in Employment", The BE Journal of Economic Analysis & Policy, 7 (1)

Adams, S, and C Cotti (2008), "Drunk Driving After the Passage of Smoking Bans in Bars", Journal of Public Economics 92 (5), 1288-1305

American Nonsmokers’ Rights Foundation (2016), “States, Commonwealths, and Municipalities with 100% Smokefree Laws in Non-Hospitality Workplaces, Restaurants, or Bars”, July

Clower, T, and B Weinstein (2004), "The Dallas Smoking Ordinance One Year Later. A Report on the Impacts of the City of Dallas Smoking Ban on Alcoholic Beverage Sales March 2003 to March 2004", Report prepared for the Greater Dallas Restaurant Association

Eriksen, M, and F Chaloupka (2007), "The Economic Impact of Clean Indoor Air Laws", CA: A Cancer Journal for Clinicians 57 (6), 367-378

Fleck, R K, and F A Hanssen (2008), “Why Understanding Smoking Bans is Important for Estimating their Effects: California’s Restaurant Smoking Bans and Restaurant Sales”, Economic Inquiry, 46 (1), 60-76

Gallup (2015), “Ban on Smoking in Public Retrains Majority Support,” http://www.gallup.com/poll/184397/ban-smoking-public-retains-majority-su...

International Agency for Research on Cancer (2009), Evaluating the Effectiveness of Smoke-free Policies, Lyon, France

Nikaj, S, J Miller, and J Tauras (2016), “The Over Time Impacts of Smoke Free Air Ordinances in Texas”, National Bureau of Economic Research Working Paper 22352

Scollo, M, A Lal, A Hyland, and S Glantz (2003), "Review of the Quality of Studies on the Economic Effects of Smoke-free Policies on the Hospitality Industry", Tobacco Control 12 (1), 13-20

Tynan M A, C B Holmes, G Promoff, C Hallett, M Hopkins, and B Frick (2016), “State and Local Comprehensive Smoke-Free Laws for Worksites, Restaurants, and Bars — United States, 2015”, MMWR 65, 623–626

Endnotes

[1] A recent policy change in California increases that number to roughly 60% of the US population.  But even then, approximately 40% of the US population is unprotected by comprehensive smoking bans in bars, restaurants, and workplaces.

[2] On May 1st 2016, the world’s largest Honky-Tonk – Billy Bob’s in Fort Worth, Texas – banned smoking on its premises. The ban was undertaken because of overwhelming feedback from customers who wanted the establishment to ban smoking.

[3] Adams and Cotti (2008) find that alcohol related fatal accidents increase following bans of smoking in bars as smokers drive longer distances to bars that allow smoking indoors.

315 Reads