Anger, S., M. Kvasnicka and T. Siedler (2010) “One Last Puff? Public Smoking Bans and Smoking Behavior,” IZA Discussion Paper No. 4873.Google Scholar
Angrist, J. and J. Pischke (2008) Mostly Harmless Econometrics. An Empiricist’s Companion. Princeton, NJ: Princeton University Press.Google Scholar
Bartecchi, C., R. Alsever, C. Nevin-Woods, W. Thomas, R. Estacio, B. Bartelson and M. Krantz (2006) “Reduction in the Incidence of Acute Myocardial Infarction Associated with a Citywide Smoking Ordinance,” Circulation, 114(14):1490–1496.Google Scholar
Bitler, M., C. Carpenter and M. Zavodny (2010) “Effects of Venue-Specific State Clean Indoor Air Laws on Smoking-Related Outcomes,” Health Economics, 19:1425–1440.PubMedWeb of ScienceCrossrefGoogle Scholar
Carpenter, C. (2009) “The Effects of Local Workplace Smoking Laws on Smoking Restrictions and Exposure to Smoke at Work,” The Journal of Human Resources, 44(4):1023–1046.Google Scholar
Carpenter, C., S. Postolek and C. Warman (2010) “Public-Place Smoking Laws and Exposure to Environmental Tobacco Smoke (ETS),” NBER Working Paper No. 15849.Google Scholar
Cawley, J. and C. Ruhm (2012) “The Economics of Risky Health Behaviors.” In: (Pauly, M., T. McGuire and P. Barros, eds.) Handbook of Health Economics, Volume 2. North Holland: Elsevier, pp. 95–199.Google Scholar
Cesaroni, G., F. Forastiere, N. Agabiti, P. Valente, P. Zuccaro and C. Perucci (2008) “Effect of the Italian Smoking Ban on Population Rates of Acute Coronary Events,” Circulation, 117(9):1183–1188.Web of ScienceGoogle Scholar
Chaloupka, F. and K. Warner (2000) “The Economics of Smoking.” In: (Culyer, A. and J. Newhouse, eds.) Handbook of Health Economics, Volume 1. North Holland: Elsevier, pp. 1539–1627.Google Scholar
European Commission (2007) Green Paper. Towards a Europe Free From Tobacco Smoke: Policy Options at the EU Level. COM(2007) 27 final. Brussels, Belgium: European Communities.Google Scholar
Evans, W., M. Farrelly and E. Montgomery (1999) “Do Workplace Smoking Bans Reduce Smoking?” American Economic Review, 89(4):728–747.Google Scholar
Fichtenberg, C. and S. Glantz (2002) “Effect of Smoke-Free Workplaces on Smoking Behavior: Systematic Review,” British Medical Journal, 325:188–194.Google Scholar
Gallus, S., P. Zuccaro, P. Colombo, G. Apolone, R. Pacifici, S. Garattini, C. Bosetti and C. La Vecchia (2006) “Effects of New Smoking Regulations in Italy,” Annals of Oncology, 17(2):346–347.CrossrefGoogle Scholar
Gallus, S., P. Zuccaro, P. Colombo, G. Apolone, R. Pacifici, S. Garattini, C. Bosetti and C. La Vecchia (2007) “Smoking in Italy 2005–2006: Effects of a Comprehensive National Tobacco Regulation,” Preventive Medicine, 45(2–3):198–201.Web of ScienceCrossrefGoogle Scholar
Gallus, S., I. Tramacere, R. Pacifici, P. Zuccaro, P. Colombo, S. Ghislandi and C. La Vecchia (2011) “Smoking in Italy 2008–2009: A Rise in Prevalence Related to the Economic Crisis?” Preventive Medicine, 52:182–183.Web of ScienceGoogle Scholar
Irvine, I. and H. Nguyen (2011) “Toxic Choices: the Theory and Impact of Smoking Bans,” Forum for Health Economics and Policy, 14(2):Article 6.Google Scholar
Jamrozik, K. (2006) “An Estimate of Deaths Attributable to Passive Smoking in Europe.” In: (Smokefree Partnership ed.) Lifting the Smokescreen: Ten Reasons for a Smoke-Free Europe. Brussels, Belgium: European Respiratory Society. Available at: www.ersnet.org, pp. 17–41.
Joossens, L. and M. Raw (2007) “Progress in Tobacco Control in 30 European Countries, 2005 to 2007.” Paper presented at the 4th European Conference Tobacco or Health 2007, Basel, Switzerland, 11–13 October 2007.Google Scholar
Juster, H., B. Loomis, T. Hinman, M. Farrely, A. Hyland, U. Bauer and G. Birkhead (2007) “Declines in Hospital Admissions for Acute Myocardial Infarction in New York State after Implementation of a Comprehensive Smoking Ban,” American Journal of Public Health, 97(11):2035–2039.CrossrefWeb of ScienceGoogle Scholar
Karaka-Mandic, P., P. Norton and B. Dowd (2012) “Interaction Terms in Nonlinear Models,” Health Service Research, 47(1):255–274.Google Scholar
Keeler, T., T. Hu, P. Barnett and W. Manning (1993) “Taxation, Regulation and Addiction: a Demand Function for Cigarettes Based on Time-Series Evidence,” Journal of Health Economics, 12:1–18.CrossrefGoogle Scholar
Pell, J., S. Haw, S. Cobbe, D. Newby, A. Pell, C. Fischbacher, A. McConnachie, S. Pringle, D. Murdoc, F. Dunn, K. Oldroyd, P. Macintyre, B. O’Rourke and W. Borland (2008) “Smoke-Free Legislation and Hospitalizations for Acute Coronary Syndrome,” New England Journal of Medicine, 359(5):482–491.Web of ScienceGoogle Scholar
Puhani, P. (2008) “The Treatment Effect, the Cross Difference, and the Interaction Term in Nonlinear ‘Difference-in-Differences’ Models.” IZA Discussion Paper No. 3478.Google Scholar
Sargent, R., R. Shepard and S. Glantz (2004) “Reduced Incidence of Admissions for Myocardial Infarction Associated with Public Smoking Ban: Before and After Study,” British Medical Journal, 328(7446):977–980.Google Scholar
Shetty, K., T. Deleire, C. White and J. Bhattacharya (2011) “Changes in Hospitalization Rates Following Smoking Bans,” Journal of Policy Analysis and Management, 30(1):6–28.Web of ScienceCrossrefGoogle Scholar
Stephens, T., L. Pederson, J. Koval and C. Kim (1997) “The Relationship of Cigarette Prices and No-Smoking Bylaws to the Prevalence of Smoking in Canada,” American Journal of Public Health, 87(9):1519–1521.Google Scholar
Tauras, J. (2005) “The Impact of Smoke Free Air Laws and Cigarette Prices on Adult Cigarette Demand,” Economic Inquiry, 44:333–342.Google Scholar
Wasserman, J., W. Manning, J. Newhouse and J. Winkler (1991) “The Effect of Excise Taxes and Regulations on Cigarette Smoking,” Journal of Health Economics, 10:43–64.Google Scholar
World Bank (2003) Tobacco Control at a Glance. Washington, DC: World Bank.Google Scholar
World Health Organization (2003) WHO Framework Convention on Tobacco Control. Geneva, Switzerland: World Health Organization.Google Scholar
About the article
Published Online: 2013-03-20
Published in Print: 2013-01-01
Other studies have also argued that smoking bans, especially comprehensive ones, should decrease the “social acceptability” of smoking, thus reducing smoking also in private places, particularly at home (Gallus etal. 2007).
Two major exceptions are the UK and Germany where, given their federal structure, these laws have been introduced on different dates in different states.
More specifically, the highest score (22) means that a complete ban (with no exceptions, such as designated smoking rooms) has been adopted and enforced in all workplaces, including cafes and restaurants, on public transport and in other public places. A relatively high score (≥15) is close to the previous case, but allowing for closed, ventilated and designated smoking rooms. A low score (≤5) means that although there are some smoke-free legislation in place, these are not strictly enforced. For additional details on the construction of the scale, see Jossens and Raw (2006).
A major exception is France, which has been implementing its smoke-free legislation in two stages, in 2007 and 2008.
The Eurobarometer survey periodically monitors the attitude of Europeans towards tobacco.
Italy is partly an exception, since the share of regular smokers has been significantly decreasing since 2005 and this reduction has been larger than in the other countries.
The comparison between the two figures is not informative in terms of trends over time, since they are based on different sources and different questions. Rather, this comparison may be useful to check the robustness of rankings based on cross-section data. Note also that the Eurobarometer survey is addressed to the entire population (including a sub-sample of workers), while the EWCS survey is focused only on workers. The number of valid observations for work-related statistics is then much larger (and subsequent results more reliable) in the second case. See the next section for further details on the EWCS.
The survey is administered by the European Foundation for the Improvement of Living and Working Conditions. At present, five waves are available (1990, 1995, 2000, 2005 and 2010); for each wave the target number of interviews is at least 1000, except for small countries (such as Cyprus, Estonia, Luxemburg, Malta and Slovenia). The survey also provides sampling weights in order to make reliable comparisons across countries.
Other than the ATT, policy evaluators may be interested in the average effect of the treatment on the entire population (the ATE). Because in our case the treated units are countries and the treatment we consider is not compulsory (EU countries are invited to follow the EU Directives, but they are independent in designing and implementing their national health-related laws), the ATT seems preferable to the ATE.
The 2000 survey was carried out in March–April, whereas the 2005 survey from mid September to the end of November. Some care should be used in the presence of seasonality in smoking behavior; however, it should be noted that indoor smoking behavior is unlikely to be affected and that the Diff-in-Diff strategy is robust to changes that affect both treated and control in a similar way. Another concern is that the treatment period might have been relatively short for Sweden, where comprehensive smoking bans were implemented in June 2005. However, given the features of this policy, some effects, particularly on smoke exposure, should be observables immediately after its implementation.
We decided to exclude the Eastern European countries from the control group because of the lack of perfectly comparable data for the pretreatment period.
Unfortunately this survey does not provide information on individual smoking behavior. This information for a large number of EU countries is usually available in more general surveys (such as the above-mentioned Eurobaromter), but they contain very little information on workers and risk exposure on the job. Furthermore, although the effect of comprehensive smoking bans on smoking prevalence in Europe has been already partly addressed at least at the country level, no evidence is available on their effects within the workplace and on smoke exposure.
The exact wording of the question is “Are you exposed at work to breathing in smoke or fumes?” Workers could answer on the basis of the following 7-point scale: all of the time, almost all of the time, around 75% of the time, around half of the time, around 25% of the time, almost never, never.
More specifically, the workers were asked whether their work affected their health; they had to look at a card showing a list of potential work-related health problems (including respiratory difficulties) and had to mention those affected by their work.
Notice that these changes are quite large in relative terms: The decline registered in the share of workers with work-related respiratory problems in the treated countries corresponds to more than 18% of the initial mean share.
The inclusion of these controls may be a way to better account for unobserved heterogeneity. Note that Angrist and Pischke (2008) point out that the vector of regressors may include individual-level characteristics as well as time-varying variables measured at the state level. Only the latter are likely to be a source of omitted variable bias, but individual-level controls can increase estimate precision by reducing the standard error of the Diff-in-Diff effect. Furthermore, the Diff-in-Diff specification should include only individual controls, which are not expected to be influenced by the treatment. Given that national comprehensive smoke-free laws apply to all the workplaces and public places within a country, we expect our firm and job-related controls to be largely unaffected by the policy considered.
Whereas heterogeneous effects of Equation (2) were estimated also by gradually adding the set of controls, in columns (4) and (8) we only report estimates based on our preferred specification. The complete set of results is available from the authors upon request.
Note that all the EU-15 countries are characterized by a significant reduction in the probability of exposure to smoke (given that the estimated β2 is negative and statistically significant), but this reduction was more pronounced in the treated countries (as measured by the β3coefficient).
Note that in the case of Ireland, this result is mainly due to a relatively high standard error rather than an estimated effect close to zero.
According to official data, only 1.5% of the total inspections carried out by the police and other civil forces resulted in a violation of the current law (Gallus etal. 2006).
As a further sensitivity test, we pooled all the waves available and reestimated our model using the 1995 wave to additionally control for the pretreatment trend. Overall these new estimates confirm our main results and are available upon request.
Between 2000 and 2005, hospital discharges in the case of respiratory diseases increased in Belgium, Denmark, Germany, Spain, the Netherlands and Portugal. These data are downloadable from the public health section of the online Eurostat database.
For each outcome variable, we report results obtained with different model specifications: no controls in columns (1) and (4), the complete vector of controls in columns (2) and (5) and estimates of heterogeneous effects in columns (3) and (6).