The effect of outdoor smoking ban: Evidence from Korea

Abstract

To address exposure to secondhand smoke, which is highly prevalent in Korea, local governments have implemented smoking bans at open public places (parks, bus stops, and school zones) since 2011. Exploiting temporal and spatial variation in the implementation dates of these bans, this study estimates their causal effects on individual smoking behavior. The individual‐level longitudinal data from the 2009–2017 Korean Labor and Income Panel Study are linked to the smoking ban legislation information from the National Law Information Center. I find robust evidence that outdoor smoking bans increased the probability of making a quit attempt by 16%. This effect appears immediately after a ban goes into effect and lasts for three or more years. People who spend more time outdoors are more likely to change smoking behavior. I also find heterogeneity in effects across the amount of monetary penalty. Whereas the policy change did not affect the prevalence of smoking overall, higher penalties had stronger impacts on reducing the intensity of smoking and increasing the propensity to try to quit.

Full text

RESEARCH ARTICLE
The effect of outdoor smoking ban: Evidence from Korea
Hansoo Ko
1,2
1
Wagner Graduate School of Public
Service, New York University,
New YorkNew York
2
Division of Health Policy &
Administration, University of Illinois at
Chicago School of Public Health, Chicago,
Illinois
Correspondence
Hansoo Ko, Wagner Graduate School of
Public Service, New York University, 295
Lafayette Street, Room 3034, New York,
NY 10012.
Email: hansooko@nyu.edu
Abstract
To address exposure to secondhand smoke, which is highly prevalent in Korea,
local governments have implemented smoking bans at open public places
(parks, bus stops, and school zones) since 2011. Exploiting temporal and spa-
tial variation in the implementation dates of these bans, this study estimates
their causal effects on individual smoking behavior. The individual-level longi-
tudinal data from the 2009–2017 Korean Labor and Income Panel Study are
linked to the smoking ban legislation information from the National Law
Information Center. I find robust evidence that outdoor smoking bans
increased the probability of making a quit attempt by 16%. This effect appears
immediately after a ban goes into effect and lasts for three or more years. Peo-
ple who spend more time outdoors are more likely to change smoking behav-
ior. I also find heterogeneity in effects across the amount of monetary penalty.
Whereas the policy change did not affect the prevalence of smoking overall,
higher penalties had stronger impacts on reducing the intensity of smoking
and increasing the propensity to try to quit.
KEYWORDS
outdoor smoking ban, secondhand smoke, smoking behavior
JEL CLASSIFICATION
I10; I12; I18
1|INTRODUCTION
The World Health Organization (WHO) reported that cigarette smoking is responsible for the death of six million peo-
ple and half a trillion dollars of economic damage annually (WHO, 2013). The 2014 Surgeon General's Report linked
smoking to numerous cancers and chronic conditions and concludes that smoking affects nearly every organ of the
body (U.S. Department of Health and Human Services, 2014). International Agency for Research on Cancer (IARC) des-
ignated both active smoking and passive smoking (secondhand smoke or involuntary smoking) as carcinogenic (Group
1) agents to humans (IARC, 2004).
1
In 2004, over 30% of adult nonsmokers and 40% of children were regularly exposed
to secondhand smoke, and 603,000 premature deaths (approximately 1% of worldwide mortality rates) occurred due to
passive smoking (Öberg, Jaakkola, Woodward, Peruga, & Prüss-Ustün, 2011).
2
1
The IARC Working Group concluded that there is sufficient evidence in both humans and in experimental animals that tobacco smoking and
secondhand smoke cause cancer.
2
These estimates (throughout this article), based on smoking-attributable fraction and epidemiological estimation techniques, can be subject to errors
due to unobservable confounding factors.
Received: 31 December 2018 Revised: 20 October 2019 Accepted: 4 November 2019
DOI: 10.1002/hec.3979
Health Economics. 2019;1–16. wileyonlinelibrary.com/journal/hec © 2019 John Wiley & Sons, Ltd. 1

Over 180 parties have joined the World Health Organization Framework Convention on Tobacco Control
(WHO FCTC), which is the first international treaty negotiated under WHO, to advance the implementation of
evidence-based tobacco control policies (WHO, 2018).
3
Since the WHO FCTC entered into force on February 2005, the
past decade has witnessed the vast growth of smoking restrictions across the world to reduce tobacco consumption
among smokers and to protect nonsmokers from the harmful health effects of secondhand smoke.
To date, smoking restriction policies aimed at reducing exposure of nonsmokers to secondhand smoke have been
primarily implemented in indoor places (Hahn, 2010). WHO reported that 92% of parties participating in WHO FCTC
have implemented any kind of smoke-free legislation in 2016 and the most common places designated as smoke free
are public transport, educational and health-care facilities, government buildings, private workplaces, restaurants, pubs,
and bars (WHO, 2016).
4
A 2016 Cochrane systematic review of 77 studies found consistent evidence that national
indoor smoking restrictions reduce secondhand smoke rates and mortality for smoking-related illnesses (Frazer et al.,
2016). Also, Hahn's (2010) systematic review found that indoor smoke-free legislations improve indoor air quality.
Smoking restrictions in outdoor places have emerged only recently. Using an interrupted time-series approach,
Johns, Farley, Rajulu, Kansagra, and Juster (2014) reported that the 2011 New York City smoking ban in public parks
and beaches was negatively associated with the probability of exposure to smoke. Okoli, Johnson, Pederson, Adkins,
and Rice (2013) also reported that observed smoking rates decreased after the implementation of an outdoor smoke-free
law in the city of Vancouver. However, due to the observational nature of the studies, no causal inference could
be made.
Policymakers and scholars have debated the effectiveness of banning smoking in outdoor public places and raised
ethical questions in relation to individual liberty. Opponents of these bans assert that the negative health effects of
outdoor smoking have not been demonstrated and that such policies go too far (“paternalism”; Chapman, 2008).
Supporters argue that there are sufficient ethical and practical justifications for the policy, because it is expected to
reduce secondhand smoke, reduce the likelihood of children to follow unhealthy behavior, and help smokers to quit
(Thomson, Edwards, & Woodward, 2008). One proponent argues that the “incredible success of the indoor ban”makes
it unnecessary to provide an evidence base for banning outdoor smoking (Barber, 2015).
Studies of the impact of smoking bans, however, have provided inconclusive evidence of their effects on smoking
rates. Switzerland's natural experiment—progressive implementation of smoking bans in public venues at a state
level—was found to reduce the prevalence of smoking by 1% a year after the implementation (Boes, Marti, & Maclean,
2015). Anger, Kvasnicka, and Siedler (2011) reported that, after the implementation of state-level public smoking ban
in bars, restaurants, and dance clubs in Germany, the prevalence of smoking among people who often go out to such
places fell significantly. In the United States, comprehensive indoor smoking bans effectively reduced smoking rates
(Carton, Darden, Levendis, Lee, & Ricket, 2016) and improved the health of infants and children (McGeary, Dave,
Lipton, & Roeper, 2017).
On the other hand, Jones, Laporte, Rice, and Zucchelli (2015) reported that the introduction of smoking ban in
enclosed public places did not have short-term effects on the prevalence and intensity of smoking. Adda and Cornaglia
(2010) showed that U.S. state-level smoking ban on workplaces, restaurants, and bars did not affect the proportion of
people who smoked or attempt to quit but led smokers to spend more time at home smoking. They found, consistent
with this hypothesis, that smoking bans targeting restaurants and bars led to an increase in cotinine concentration
among children exposed to smoke at home. Cooper and Pesko (2017) also found that U.S. county-level electronic ciga-
rette indoor vaping restrictions led to an increase in consumption of traditional tobacco products among pregnant
women.
5
Given the mixed evidence of the behavioral impacts of indoor smoking ban and the very limited evidence of the
impacts of outdoor smoking bans, it is useful to examine how individuals respond to smoking bans in outdoor places. A
unique research opportunity arose in South Korea, where a policy change banning smoking in outdoor open places was
recently implemented to address prevalent secondhand smoke exposure. Exploiting a gradual rollout of outdoor
3
Such as price/tax measures, protection from exposure to secondhand tobacco smoke, regulation of the content of tobacco products, regulation of
labeling, public awareness raising program, regulation of tobacco advertising and promotion, regulation of illicit trade, and regulation of sales to
minors.
4
However, in 2016, only 20% of the global population (1.5 billion people in 55 countries) were covered by comprehensive smoke-free legislation
(WHO, 2017).
5
In addition, Shetty, DeLeire, White, and Bhattacharya (2010) analyzed the impact on mortality rates and hospitalization of local smoking ban and
did not find significant short-term effects. County-level indoor (workplaces, restaurants, and bars) smoking ban was also found to have null impacts
on neonatal health in the United States (Hankins & Tarasenko, 2016).
2KO

smoking ban at the local government level in Korea, this quasi-experimental study aims to fill the knowledge gap.
Theory predicts ambiguous effects of smoking ban on smoking behavior. Restrictions on smoking may lower the
demand for cigarettes by reducing smokers' opportunities to consume tobacco products, requiring smokers to make
additional investment of time to smoke at nonregulated places, and changing social norms regarding acceptability of
smoking. However, displacement or compensatory behavior (smoking each cigarette more intensely) may offset these
effects.
This study is the first, to the best of author's knowledge, to estimate causal effects of outdoor smoking bans on indi-
vidual smoking behavior. My results indicate that outdoor smoking ban increased the probability of making a quit
attempt by 16%. This effect lasted for three or more years after the implementation of ban. I also find heterogeneity in
effects across the amount of monetary penalty. Whereas the policy change did not reduce the prevalence of smoking,
higher penalties had stronger impacts on reducing the intensity of smoking and on increasing the propensity to try
to quit.
This study contributes to the literature by revealing mechanisms through which outdoor smoke-free policy affects
individual smoking behavior. Outdoor smoking bans raise awareness about harmful effects of smoking, and this leads
to an increase in quit attempts—though it is not strong enough to reduce the demand for cigarettes. This effect is stron-
ger among persons who spend more time outdoors, indicating that socially active persons are more likely to be exposed
to changes in social norms regarding tobacco use in public places. In addition, as the amount of penalty has differential
impacts on quit attempts and the intensity of smoking, outdoor smoking bans raise the monetary costs of smoking.
2|BACKGROUND
Cigarette smoking is highly prevalent in Korea. Figure 1 shows that the proportion of smokers in the general popula-
tion 15 years or older remains stable at 20% after a sharp decrease in the 1990s. The proportion of men who smoke daily
has rapidly declined from 66.7% in 1995 to 31.4% in 2015, whereas the share of female daily smokers has fluctuated
between 4% and 7% in the past two decades. Smoking cost more than 1.3 million disability adjusted life years in Korea
in 2013 (Zahra, Cheong, & Park, 2017). Also, using nationally representative claims data taken from the Korean public
insurance scheme, Oh et al. (2012) estimated the total economic costs of smoking-related cancers
6
as $3 billion in 2008.
According to WHO (2015), the retail price of the most popular brand in Korea was $2.43 in 2014, which was the
cheapest among Organization for Economic Cooperation and Development (OECD) member states. Cigarette taxes
comprise 72% of the retail prices and consist of consumption tax, sales tax, health promotion and education charges,
and value added tax. The amount of taxes is set by the central government and is uniform across the nation—thus,
there is no spatial variation in cigarette prices. The central government raised the retail prices of the most popular
brand (by increasing cigarette taxes) from 2,000 won (approximately $2) to 4,500 won in 2015. However, cigarettes
remain an accessible good because the price of a pack of cigarettes is still cheaper than the average prices in most
OECD member states.
7
Although Korea ratified the WHO FCTC in 2005, tobacco control policies have been weak (Cho, 2014). Tobacco
advertising, promotion, and sponsorship are not comprehensively restricted, the sale of tobacco products to minors is
FIGURE 1 The smoking prevalence in Korea, 1995–2015. The
prevalence represents the share of the general population aged
15 years or older who reported to smoke daily. Source: OECD
(2017))
KO 3

poorly enforced, though it is prohibited, and smoking cessation services are not covered by the public insurance
scheme.
The rapid expansion of smoking restrictions in the country, which ranks at the top of the list of OECD countries in
the proportion of adult male smokers (OECD, 2017), reflects the will of the public to reduce harmful health effects of
exposure to secondhand smoke. According to Statistics Korea (2017), the prevalence of secondhand smoke among adult
nonsmokers was 46% at workplaces and 14.7% at home in 2007. Total burden of disease due to secondhand smoke was
over 44,000 disability adjusted life years in 2013 (Zahra, Cheong, Lee, & Park, 2016).
In 1995, with the enactment of the Health Promotion Act, the central government banned indoor smoking in some
public places and selling cigarettes to minors. Smoking in government buildings, hospitals, nurseries, schools, bars, and
restaurants larger than 150 m
2
was banned nationwide in December 2012. A nationwide smoking ban in all restaurants
was instituted in January 2015.
In addition, under the National Health Promotion Plan, the central government gave local authorities the power to
enact bylaws banning smoking in outdoor public places (Cho, 2014). The main purpose of the act was to address preva-
lent secondhand smoke in 2010. For instance, a bill “Seoul metropolitan city passive smoking prevention ordinance”
that passed the Seoul Metropolitan Council in 2011 specified the purpose of it in Article 1 as “protecting all citizens
from harmful secondhand smoke and improving the health of citizens.”
This has led to a rapid spread of outdoor smoking bans across the country. Beginning with Gwanak-gu, Seoul, in
June 2011, 219 out of 226 cities have implemented bans as of the end of 2016 (Figure 2). The rollout of the bans was
over 5 years (2011–2015), but most localities implemented bans during the 2011–2014 period: The share of cities with
outdoor smoking bans increased from zero in 2010 to 19.5% in 2011, 54.9% in 2012, 88.1% in 2013, 91.2% in 2014, and
94.2% in 2015. Urban and metropolitan cities tended to implement bans earlier than rural cities did. Lee, Park, Kim,
and Jung-Choi (2014) also found that outdoor smoking bans were implemented earlier in cities with higher average
educational attainment and fiscal independence ratio. Bus stops, public parks, school zones, and outdoor parking lots
are the most commonly protected places under local outdoor smoking bans (Cho, 2014).
3|PREVIOUS RESEARCH
Restrictions on smoking are primarily aimed at reducing secondhand smoke exposure of nonsmokers (Chaloupka &
Warner, 2000). Secondhand smoke involves inhaling toxic agents and carcinogens including benzene, 1,3-butadiene,
benzo[a]pyrene, and 4-(methyl-nitrosamino)-1-(3-pyridyl)-1-butanone (IARC, 2004), and there is no “safe”level of sec-
ondhand smoke exposure (WHO, 2017). Acute exposure to involuntary smoking causes increases in blood pressure,
heart rates at rest, and levels of carbon monoxide in blood and causes endothelial cell damage and platelet aggregation,
thus elevating the risk of atherosclerosis (He et al., 1999; Jefferis et al., 2010). A meta-analysis of epidemiological studies
of the effect of secondhand smoke found evidence that exposure to involuntary smoking increases the risk of coronary
heart disease among nonsmokers by 25% (He et al., 1999). Pan, Wang, Talaei, Hu, and Wu (2015) also found the associ-
ation of passive smoking with an increased incidence of a chronic condition (type 2 diabetes).
Numerous studies have shown that smoking bans are associated with decreases in morbidity and mortality from
smoking-related illnesses. A Scottish smoking ban in pubs was found to lead to reductions in PM
2.5
(particulate matter
<2.5 μm) compared with the period before the ban (Semple, Creely, Naji, Miller, & Ayres, 2007). Azagba (2015), analyz-
ing the impact of Canadian smoking ban in restaurant and bar patios, reported that smoke-free legislation reduced the
probability of secondhand smoke exposure by 20%. Meyers, Neuberger, and He (2009) systematic review and meta-
analysis found that smoking bans in enclosed public places reduced the risk of acute myocardial infarction by 17% with
the strongest effect found among nonsmokers and young populations. Smoke-free legislation was also found to reduce
hospitalization and deaths for coronary, cerebrovascular, and respiratory diseases (Tan & Glantz, 2012) and reduce
7
Using yearly time-series data, Kim and Seldon (2004) estimated the short-run price elasticity for cigarette demand as −0.28, which was slightly lower
than the conventional range of −0.3 to −0.5 (Chaloupka & Warner, 2000). Authors explained that this low price elasticity is attributed to (a) low
prices of cigarettes and (b) the fact that no substitute tobacco products for cigarettes are available (until recently, the use of cigars, pipe tobacco, and
smokeless tobacco has been rare in Korea).
7
Using yearly time-series data, Kim and Seldon (2004) estimated the short-run price elasticity for cigarette demand as −0.28, which was slightly lower
than the conventional range of −0.3 to −0.5 (Chaloupka & Warner, 2000). Authors explained that this low price elasticity is attributed to (a) low
prices of cigarettes and (b) the fact that no substitute tobacco products for cigarettes are available (until recently, the use of cigars, pipe tobacco, and
smokeless tobacco has been rare in Korea).
4KO

preterm births and child asthma admissions (Been, Nurmatov, Nawrot, van Schayck, & Cheikh, 2014). A recent study
of smoking ban in bars and restaurants from Germany showed that state-level smoking bans resulted in short-run
reductions in cardiovascular admissions (−2.1%) and asthma admissions (−6.5%; Kvasnicka, Siedler, & Ziebarth, 2018).
FIGURE 2 Spatial and temporal variations in outdoor smoking ban at city
level (N= 226)
KO 5

In addition, theory predicts that smoking bans would have inhibitory effects on tobacco consumption. Restrictions
on smoking can lower the demand for cigarettes by reducing smokers' opportunities to consume tobacco products
(Chaloupka & Warner, 2000). Smoking restrictions also can change individual's smoking behavior by increasing
vdisutility from consuming tobacco products because of changes in social norms regarding the acceptability of smoking
(Jones et al., 2015). In addition, smoking ban requires smokers to make additional investments of time to smoke at
nonregulated places (Chaloupka & Warner, 2000; Cooper & Pesko, 2017). Smokers may also change smoking
intensity/frequency or attempt to quit when the perceived marginal costs of smoking exceed the marginal benefits.
According to the theory of marginal smokers, there are certain groups of smokers who regret their addictive habits,
want to quit, and fail to do so because of limited willpower (Odermatt & Stutzer, 2015). These smokers tend to demand
self-control devices, which are believed to help them successfully quit. A smoking ban might serve as a trigger to these
motivated smokers and could result in an increase in their making a quit attempt.
A systematic review of the literature found that nicotine dependence, cigarettes consumed per day, educational
attainment, and wealth were negatively associated with making a quit attempt, whereas age, past attempts to quit, per-
sonal motivation and intention to quit, and home smoking bans were positively related with quit attempts (Vangeli,
Stapleton, Smit, Borland, & West, 2011). The review also revealed that age, personal motivation, intention to quit, and
(a lower level of) cigarette dependence were predictors of successful quitting, whereas the level of education and
income were not significantly related to quit attempt success.
Smoking cessation guidelines published in the United Kingdom suggested that relapse is a normal process of quit-
ting and smokers on average make three or four quit attempts before finally quitting (Raw, McNeill, & West, 1998).
However, numerous studies reported that smokers who made quit attempts in the past were less likely than those who
never tried to quit to succeed because of experienced withdrawal and fear of failure (Murray et al., 2000; Nakamura,
Oshima, Ohkura, Artega, & Suwa, 2014; Vangeli et al., 2011).
A Cochrane review concluded that the effects of indoor smoking bans on tobacco consumption are not clear (Frazer
et al., 2016). This is particularly true when a ban is not comprehensive or not enforced appropriately. These findings
indicate that smoking bans alone would likely not be strong enough to change the demand for cigarettes at measurable
levels. As well, a ban's impacts would not be substantial unless the open public places designated as smoke free are the
places at which smokers used to smoke.
Even if outdoor smoking ban reduces smokers' opportunities to smoke, it can be offset by individuals' compensatory
behavioral changes. For instance, smokers may increase the number of cigarettes they consume at once to compensate
the additional inconvenience and to maintain their desired level of nicotine.
In addition, by making smokers consume cigarettes at private places that are separated from others, smoking bans
may decrease the chance for smokers to face peer pressure (Odermatt & Stutzer, 2015). If smokers consume tobacco
products at their shared havens, they would spend more time with other smokers. As Lee and Kahende's review (2007)
suggested, having daily contact with other smokers may reduce the probability of successful quitting.
4|METHODS
I combine two data sources to estimate the effects of outdoor smoking ban: annual individual smoking behavior data
from the 2009–2017 waves of the Korean Labor and Income Panel Study and information on the implementation of
bans and monetary penalties from the National Law Information Center (http://www.law.go.kr/eng/engMain.do).
The Korean Labor and Income Panel Study is an annual nationally representative panel survey of individuals and
has collected a wide range of individual-level information regarding demographic and socioeconomic characteristics,
job status, and health-related questionnaires.
8
My study sample, an unbalanced panel, includes 13,095 unique persons
aged 18–80—minors younger than 18 years are not allowed to buy tobacco products in Korea—at baseline and 71,414
person-years of data. Persons who ever changed their locations of residence during the study periods are excluded from
the study sample.
9
8
The original panel of 5,000 households was constructed in 1998. To address an attrition issue (follow-up loss ratio of 25.8% in 2008), around 1,400
new households were newly added to the panel in the 2009 survey. Thus, this study utilized the panel constructed in 2009.
9
It also helped me to cluster standard errors at city level. In addition, as individual-level fixed effects were included in my main specification, city-
level fixed effects were canceled out due to perfect multicollinearity.
6KO

Exploiting temporal and spatial variation in exposure to outdoor smoking bans (explained in the Background sec-
tion), by using the information on the dates of the survey, I construct the policy indicator (Ban
ct
), which takes 1 if out-
door smoking ban was in effect at localities
10
where respondents were residing. My main estimation model uses a
difference-in-differences approach as below:
Yicpt =αicpt +βBanct +πXit +θZpt +Yt+γi+εicpt:ð1Þ
The unit of analysis is the person-year, and standard errors are clustered at the city level. Y
icpt
represents self-
reported smoking behavior: (a) an indicator whether currently smoking cigarettes, (b) an indicator whether having tried
to quit, (c) and indicators for smoking intensity (fraction of current smokers)
11
of individual iin city cand province pat
time t.X
it
is annual individual earned income that may related to smoking behavior. Gross regional domestic product
per capita is also included to control for time-variant economic conditions (Z
pt
)
12
.
Year dummies (Y
t
) are expected to absorb the common economic/policy shocks affecting smoking behavior system-
atically across all subnational regions. Such a shock includes the 2012 indoor smoke-free policy implemented by the
central government banning smoking in indoor public places including government offices, medical facilities, and large
restaurants/bars (Cho, 2014). Note that cigarette prices are strictly controlled by the central government and identical
across the nation. I do not include cigarettes prices/taxes because these are set at the national level and are absorbed by
year fixed effects. Also, to the best of my knowledge, no other city-level indoor smoking restriction policies such as a
public health education campaign were enacted in the same period as the outdoor smoking ban.
13
Including individual fixed effects (γ
i
) is beneficial since depreciated consumption activities occurring in the past,
which might affect the present decision, can be canceled out (Chaloupka & Warner, 2000). In particular, Cooper and
Pesko (2017) document, omitting individual-level time-invariant unobservable factors can result in biased estimates of
the impact of a smoke-free policy.
For the above difference-in-difference approach to be valid, a parallel trends assumption, where both intervention
and control groups would have followed same secular trends without the policy change, should be satisfied. For
instance, it is plausible that local governments select into the smoke-free policy because the social issue (secondhand
smoke) has been serious in their jurisdictions, so trends in smoking behavior would have moved differently in treated
cities even without bans. If this is the case, the estimated impacts would be upwardly biased.
To address this issue, first, I include province-specific linear time trends given the possibility that unobservable fac-
tors affecting the smoking/quitting trends vary within administrative regions (Kurtulus, 2016). Specifically, I check if
adding province-specific linear trends changes results from the main specification.
14
In addition, even with individual fixed effects, year fixed effects, and region-specific trends controlled for, there
remains a possibility of reverse causality that changes in anti-smoking laws are affected by changes in smoking trends.
Such a scenario can arise if proban governments were more sensitive to the anti-smoking sentiment of their citizens. If
this is the case, I would expect to observe preexisting decreasing (increasing) trends in smoking rates (making quit
attempts) before the implementation of the ban. To check this dynamics around the timing of the law enforcement, I
estimate a dynamic event study specification that includes leads and lags of the implementation of the outdoor smoking
ban as below:
Yicpt =αicpt +Xt+3
j=t−4βjBancj +πXit +θZpt +Yt+γi+εicpt,ð2Þ
10
Both upper level (provinces) and lower level (cities) localities are independently able to implement bylaws. My policy indicator equals 1 if there was
any outdoor smoking ban in place at the time of survey.
11
The questionnaire item has four mutually exclusive options: (a) two packs or more daily, (b) 20–39 cigarettes, (c) 10–19 cigarettes, and (d) fewer
than 10 cigarettes. For the ease of interpretation, I classify the intensity of smoking into three indicators: 40 or more, 20–39, and 19 or fewer cigarettes
per daily.
12
This varies across provinces. City-level information is not publicly available.
13
Health warnings on cigarettes packs (1976), banning sales to minors younger than 18 (1995), tobacco tax increase (2005), and smoking cessation aid
programs at public health centers (2005) were all implemented nationwide at the same time.
14
To avoid a significant loss of degrees of freedom (Rocha & Soares, 2010), city-specific trends are not included in the model. There are 17 provinces
and 226 cities in the nation as of 2016.
KO 7

where indicator variables Ban
ct −4
–Ban
ct+2
are equal to 1 only in the relevant year (for instance, Ban
ct0
equals 1 only in
the year of implementation and Ban
ct+1
indicates one year after implementation). An indicator variable Ban
ct+3
equals
1 in every year beginning with the third year after implementation. In this model, with “1 year before the implementa-
tion”as the omitted category, I expect to find no evidence suggesting anticipatory behavioral changes or reverse causal-
ity, allowing me to provide robust evidence on the causal effects of outdoor smoking ban. Specifically, coefficients on
the policy leads (^
βj−4,^
βj−3, and ^
βj−2) should be statistically not different from zero.
In additional analyses, I replace the policy indicator (Ban
ct
) with indicators for different penalty levels (<50,000,
<100,000, and 100,000 won) to see if the strictness of ban is associated with individual smoking behavior. As of the end
of 2015, violators are subject to fines of 20,000 (in 44 cities), 30,000 (in 42 cities), 50,000 (in 75 cities), 70,000 (in two cit-
ies), or 100,000 won (in 43 cities).
15
5|RESULTS
Table 1 shows summary statistics of my study sample. At baseline (survey year 2009), the sample consists of 52%
women, and mean age is 46. Smoking rate is 24.6% (49.0% among men and 1.67% among women). Among current
smokers, just below 30% tried to quit in the previous month. Percentages of current smokers who smoke 40 or more,
20–39, and 19 or fewer cigarettes daily are 2.3%, 31.8%, and 65.9%, respectively. Due to attrition and moves across cities,
the sample size decreased to 5,755 by the endpoint (survey 2017). Also note that, compared with statistics at baseline,
persons remaining in the sample until the endpoint were more likely to be older and low educated. These composi-
tional changes in the study sample would bias the estimates of the impacts of smoking ban. To check this, I additionally
estimate the main specification only with the balanced panel (shown in Table 2).
Table 2 reports the results from the main specification (Equation (1)). The outdoor smoking ban did not affect the
prevalence of smoking. However, the implementation of the ban increased cessation attempts by 4.8 percentage points
(16.3% of the mean) among current smokers. The policy change was also associated with a shift from consuming more
than a pack a day (1.1 percentage points) to consuming less than a pack a day (+1.4 percentage points), though none of
these coefficient estimates are statistically significant at the traditional level of significance.
21
Adding province-specific
15
As shown in Section 5, I use both categorical variables and continuous variable and find that results are qualitatively similar.
21
A failure to find a significant association might be due to limited statistical power (1,036 subjects of the balanced panel were currently smoking), but
I am not able to test this with the data set I used for this study.
TABLE 1 Summary statistics
Variable
Baseline
(survey year 2009)
Endpoint
(survey year 2017) Whole period (2009–2017)
Current smoker 0.246 0.180 0.222
Attempt to quit (among current smokers) 0.294 0.232 0.264
Smoking intensity (among current smokers)
≥40 cigarettes per day 0.023 0.013 0.017
20–39 cigarettes per day 0.318 0.231 0.288
<20 cigarettes per day 0.659 0.757 0.695
Female 0.518 0.559 0.538
Age 46.112 (15.978) 59.338 (14.608) 51.766 (15.171)
Wage workers 0.383 0.376 0.397
Educational attainment
Primary completion or less 0.184 0.257 0.212
Secondary completion or less 0.450 0.494 0.480
More than college attendance 0.366 0.250 0.309
Annual earned income (10,000 Korean won) 1,369.639 (2,101.797) 1,597.075 (2,302.497) 1,458.855 (2,164.690)
Married 0.670 0.751 0.714
Observations 12,277 5,755 71,414
Note. Mean and standard deviation (in parentheses) are presented. Due to attrition and moves across cities, the number of survey participants reduced to 10,377
(2010), 8,856 (2011), 8,044 (2012), 7,214 (2013), 6,695 (2014), 6,208 (2015), 5,988 (2016), and 5,755 in 2017. US$1 ≈1,100 won. Monetary values are adjusted to
the 2015 US$ value by consumer price index.
8KO

linear trends to the specification does not change the results much (panel B). In addition, estimating with the balanced
panel (5,755 persons at baseline) provides qualitatively similar results (panel C).
22
The results from panel event study are presented in Table 3 and Figure 3. Compared with the reference period (the
year before the implementation of ban), the impact of a ban on the probability of making a quit attempt appeared as
soon as the ban went into effect (year 0) and lasted for three or more years with similar marginal effects. In particular,
coefficient estimates of the effects on cessation attempts are not statistically different from zero any time before the pol-
icy took effect. Together with the findings that including/excluding region-specific time trends does not qualitatively
change the main findings (Table 2), these results reaffirm that a parallel trends assumption holds. There were no dis-
tinct upward trends in quitting attempts before the implementation of bans in general, but I found an (statistically
insignificant) increase in quit attempts between periods −2 and −1. This could reflect anticipatory behavior given the
possibility that individuals might have knowledge about upcoming regulatory changes through media coverage of pol-
icy debates. I check this possibility in part by leveraging the intervals between the enactment dates and the actual
implementation dates of bans.
23
Results in Table A1 show that there were no systematic behavioral responses to the
enactment (no enforcement) of outdoor smoking bans, suggesting that increases in the future nonmoney price of
smoking did not affect current consumption of cigarettes.
Table 4 reports results from subgroup analyses where outcome measures were regressed by policy indicator sepa-
rately for subpopulation groups across sociodemographic dimensions. Results show that outdoor smoking ban resulted
in statistically significant increases in the probability of quit attempts among persons who are young, employed, or in
good health status. In addition, outdoor smoking bans reduced the intensity of smoking among unmarried smokers
(panel E). There is also evidence suggesting that a ban was positively associated with cessation attempts among edu-
cated or unmarried individuals. Results show similar policy impacts on both men and women (panel F).
22
At baseline, persons in the balanced panel were less likely to smoke (p= .009), less likely to smoke fewer than a pack a day (p< .001), but more
likely to smoke 1–2 packs of cigarettes a day (p< .001) than persons not in the balanced panel (who ever changed city of residence or who were
unfollowed due to attrition). There is no difference in quitting attempt and the share of smokers consuming two or more packs a day between two
groups.
23
Of all the provinces (N= 17) and cities (N= 226), 12 provinces and 153 cities passed bills without immediate effects (mean of grace period =
5.5 months; median = 6 months).
TABLE 2 The effect of outdoor smoking ban on individual smoking behavior
Variable
Current
smoking
Quit attempt (among
current smokers)
Smoking intensity (fraction among current smokers)
40 or more
cigarettes per day
20–39
cigarettes per
day
19 or fewer
cigarettes per day
Panel A. Without province-specific linear trends
Outdoor smoking ban 0.005 (0.005) 0.048
**
(0.022) −0.003 (0.004) −0.011 (0.020) 0.014 (0.020)
Observations 71,414 15,842 15,842 15,842 15,842
Baseline outcome
means
0.246 0.294 0.023 0.318 0.659
Panel B. With province-specific linear trends
Outdoor smoking ban 0.006 (0.005) 0.064
***
(0.022) −0.002 (0.004) −0.015 (0.020) 0.017 (0.020)
Observations 71,414 15,842 15,842 15,842 15,842
Baseline outcome
means
0.246 0.294 0.023 0.318 0.659
Panel C. Balanced panel (N= 5,755 at baseline) with province-specific linear trends
Outdoor smoking ban 0.006 (0.005) 0.054
**
(0.024) −0.005 (0.005) −0.027 (0.023) 0.031 (0.023)
Observations 51,795 11,107 11,107 11,107 11,107
Baseline outcome
means
0.235 0.301 0.025 0.358 0.617
Note. Standard errors, in parentheses, are clustered at city level. Each cell represents separate regression results. Regressions also include individual fixed
effects, year dummies, and annual earned income amounts.
*
Significant at.1.
**
Significant at.5.
***
Significant at.01.
KO 9

Disparities in the impacts of ban on quitting attempt along age groups, employment status, and health status suggest
that persons who actively spend more time outdoors (so more likely to be exposed to the restrictions) are more likely to
change their smoking behavior. It implies that one of channels through which outdoor smoking affects individual
smoking behavior is more exposure to smoke-free policy (and higher social pressure).
Table 5 presents the association of penalty level with self-reported smoking outcome measures. Smoking restrictions
imposing different amounts of fines (<50,000, <100,000, and 100,000 won) all had positive associations with the pro-
pensity for making a quit attempt, though only the highest level was statistically associated with the outcome. There is
TABLE 3 Panel event study: Outdoor smoking ban and individual smoking behavior
Variable
Current
smoking
Quit attempt (among
current smokers)
Smoking intensity (fraction among current smokers)
40 or more
cigarettes per day
20–39 cigarettes
per day
19 or fewer
cigarettes per day
Years since implementation
4 years before 0.014 (0.021) 0.027 (0.066) 0.0001 (0.014) −0.075 (0.074) −0.026 (0.138)
3 years before 0.012 (0.011) 0.013 (0.045) 0.004 (0.009) −0.072 (0.047) −0.009 (0.086)
2 years before 0.002 (0.006) −0.027 (0.025) 0.0003 (0.006) −0.021 (0.028) −0.012 (0.042)
1 year before
(reference)
—— ———
Implementation 0.005 (0.005) 0.050
*
(0.027) −0.003 (0.006) −0.001 (0.024) 0.033 (0.037)
1 year after 0.003 (0.008) 0.069
*
(0.041) −0.002 (0.010) 0.036 (0.040) 0.023 (0.067)
2 years after 0.010 (0.010) 0.128
**
(0.052) 0.003 (0.014) 0.058 (0.056) 0.027 (0.095)
3 and more
years after
0.014 (0.012) 0.156
**
(0.069) 0.006 (0.018) 0.080 (0.077) 0.026 (0.121)
Observations 57,691 12,622 12,622 12,622 12,622
Baseline
outcome
means
0.246 0.294 0.023 0.318 0.659
Note. Standard errors, in parentheses, are clustered at city level. Each column represents separate regression. Estimates of coefficients β
j
are taken from
Equation (2) with the year before implementation (“−1”) as the omitted category. Regressions include the same covariates as Table 2.
*
Significant at.1.
**
Significant at.5.
***
Significant at.01.
FIGURE 3 Panel event study:
effects of outdoor smoking ban on
individual smoking behavior. Each
plot shows estimates of coefficients
β
j
from Equation (2) with the year
before implementation (“−1”) as the
omitted category. Coefficient
estimates are reported in Table 3.
Regressions include the same
covariates as the main specification
in Table 3
10 KO

no clear evidence that stricter smoke-free policies led to changes in the prevalence of smoking or heavy smoking
(consuming more than two packs a day). However, I find evidence that imposing the highest amount of monetary pen-
alty led smokers consuming light to moderate amount of cigarettes (1–2 packs a day) to decrease their intensity of
smoking by around 5 percentage points.
Results from several robustness checks are presented in Table 6. As the nationwide indoor smoking ban was
implemented in December 2012, it is plausible that localities that implemented outdoor bans earlier than the national
indoor smoking ban might have enforced the indoor ban more strictly. To test whether there were such differential
impacts of outdoor smoking bans, I re-estimate the main specification with the periods 2009–2012 (just before the
implementation of the national indoor smoking ban) in panel A. Results are similar to findings in Table 2 in terms of
the direction and magnitude of coefficient estimates.
In panel B of Table 6, I replicate the main specification with only those living in Seoul Metropolitan City and other
localities (where outdoor smoking ban was not implemented until the end of 2012) to address the possibility of hetero-
geneity in the enforcement of restrictions. In Seoul, after the city council passed a bill, the outdoor smoking ban took
TABLE 4 The effect of outdoor smoking ban: Subgroup analysis
Variable
Current
smoking
Quit attempt (among
current smokers)
Smoking intensity (fraction among current smokers)
40 or more
cigarettes per day
20–39
cigarettes per
day
19 or fewer
cigarettes per day
Panel A. Age
Aged 18–39 0.010
(0.009)
0.102
***
(0.037) −0.007 (0.009) −0.033 (0.034) 0.040 (0.034)
Aged 40–64 0.008
(0.007)
0.053
**
(0.025) −0.002 (0.006) −0.026 (0.024) 0.028 (0.025)
Aged 65–80 0.002
(0.007)
−0.037 (0.039) −0.011 (0.010) 0.044 (0.038) −0.033 (0.037)
Panel B. Education
Secondary schooling or
less
0.001
(0.007)
−0.034 (0.043) −0.003 (0.011) 0.053 (0.042) −0.049 (0.041)
At least college
attendance
0.006
(0.006)
0.059
**
(0.023) −0.003 (0.005) −0.020 (0.021) 0.023 (0.022)
Panel C. Employment
Not working for wages 0.003
(0.005)
0.022 (0.025) −0.005 (0.008) −0.001 (0.028) 0.006 (0.028)
Working for wages 0.009
(0.008)
0.072
***
(0.027) −0.003 (0.006) −0.015 (0.024) 0.018 (0.024)
Panel D. Self-rated health status
Moderate/bad/very bad 0.003
(0.006)
0.009 (0.027) −0.007 (0.007) −0.012 (0.025) 0.019 (0.024)
Good/very good 0.006
(0.007)
0.068
**
(0.030) −0.004 (0.005) −0.028 (0.026) 0.032 (0.027)
Panel E. Marital status
Single/divorced/widowed −0.0009
(0.009)
0.055 (0.034) −0.012 (0.007) −0.039 (0.033) 0.051 (0.034)
Married 0.006
(0.005)
0.035 (0.023) −0.001 (0.005) −0.0009 (0.021) 0.0005 (0.021)
Panel F. Gender
Men 0.011
(0.010)
0.046
**
(0.022) −0.003 (0.004) −0.012 (0.020) 0.015 (0.021)
Women 0.0004
(0.002)
0.116
*
(0.061) −0.006 (0.005) 0.041 (0.051) −0.035 (0.048)
Note. Standard errors, in parentheses, are clustered at city level. Each cell represents separate regression. Regressions include the same covariates as Table 2.
*
Significant at.1.
**
Significant at.5.
***
Significant at.01.
KO 11

effect across the city at the same time in July 2011. This setting helps me test the impacts of bans with less heterogeneity
in the enforcement. I find that results are qualitatively similar to the results from the main specification.
In panel C and panel D of Table 6, I do placebo tests to test policy endogeneity.
40
In panel C, for a subset of
preintervention period (2009 and 2010), I randomly assigned 50% of cities a placebo treatment in 2010. In panel D, for a
subset of active rollout period (2011–2014), I randomly assigned 50% of cities that already implemented the ban in 2011
a placebo control. These placebo tests do not provide any evidence of impacts on smoking behavior.
6|DISCUSSION AND CONCLUSIONS
Taken together, these results suggest that outdoor smoke-free policies affect individual smoking behavior through at
least two mechanisms. Outdoor smoking bans raise awareness about harmful effects of smoking among smokers and
lead to an increase in quit attempts—though, on average, not by enough to reduce the demand for cigarettes. This effect
is stronger among persons who spend more time outdoors, indicating that socially active persons are more likely to be
exposed to changes in social norms regarding tobacco use in public places (Hahn, 2010). In addition, the amounts of
penalty have differential impacts on quit attempts and the intensity of smoking, suggesting that outdoor smoking ban
changes individual smoking behavior through increasing the monetary costs of smoking.
This study finds that outdoor smoking bans in Korea increased the probability of making a quit attempt by 16%.
However, the policy change was not strong enough to reduce the prevalence of smoking. In other words, most smokers
triggered by smoking restrictions to try to quit ended up having experienced relapse. Because failed quit attempts and
experienced withdrawal make smokers less likely to successfully quit smoking (Raw et al., 1998), it is safe to say that
40
I thank an anonymous reviewer for comments on this.
TABLE 5 The amount of penalty and individual smoking behavior
Variable
Current
smoking
Quit attempt (among
current smokers)
Smoking intensity (fraction among current smokers)
40 or more
cigarettes per day
20–39
cigarettes per
day
19 or fewer
cigarettes per day
Panel A. Continuous penalty variable (ref = no outdoor smoking ban)
Penalty amount
(10,000 Korean won)
0.001
(0.0006)
0.011
***
(0.003) −0.0001 (0.0006) −0.006
**
(0.003) 0.006
**
(0.003)
Panel B. Categorized penalty variable (ref = no outdoor smoking ban)
Penalty ∈(0, 50,00
won)
0.007
(0.007)
0.021 (0.025) −0.004 (0.007) 0.007 (0.034) −0.003 (0.033)
Penalty ∈(50,000,
100,000)
−0.0003
(0.006)
0.025 (0.033) −0.003 (0.006) −0.023 (0.027) 0.026 (0.027)
Penalty = 100,000 won 0.012
*
(0.007)
0.116
***
(0.033) −0.002 (0.006) −0.052
*
(0.029) 0.054
*
(0.029)
pvalue (<50,000 ≤
100,000)
0.343 0.901 0.894 0.431 0.436
pvalue (<50,000 =
100,000)
0.525 0.011 0.760 0.090 0.102
pvalue (<100,000 =
100,000)
0.097 0.030 0.840 0.308 0.343
Observations 71,709 15,921 15,921 15,921 15,921
Baseline outcome
means
0.246 0.294 0.023 0.318 0.659
Note. Standard errors, in parentheses, are clustered at city level. Each column represents separate regression. Regressions include the same covariates as
Table 2. Samples exclude those who never exposed to penalty (no outdoor smoking ban in their locality of residence). pvalues represent postestimation test if
two coefficient estimates are statistically indifferent. As of the end of 2017, violators are subject to fines of 20,000 (in 44 cities), 30,000 (in 42 cities), 50,000 (in
75 cities), 70,000 (in two cities), or 100,000 won (in 43 cities). US$1 ≈1,100 Korean won.
*
Significant at.1.
**
Significant at.5.
***
Significant at.01.
12 KO

the overall impact of the ban on active smoking was minuscule. This finding is consistent with the 2016 Cochrane
review that found inconclusive evidence of the impacts of national indoor smoking bans on the prevalence of smoking
(Frazer et al., 2016).
Studies have reported that the majority of smokers who attempt to quit use the least effective method (willpower
alone), whereas effective individual-level cessation aids (such as counseling, nicotine patch, or drugs) are not free
(Malarcher, Dube, Shaw, Babb, & Kaufmann, 2011; West, McNeill, & Raw, 2000). This has implications for public
policy—outdoor smoking ban might work better if accompanied by another policy change aiding cessation. This is
important because, given the high prevalence of smoking among the general population in Korea, both active and pas-
sive smoking should be targeted to substantially reduce the burden of smoking-related illnesses to society (He et al.,
1999).
Another implication for public policy is that violators of smoke-free policies should be subject to high penalties. This
study finds evidence that impacts of ban on reducing the intensity of smoking only appeared for persons exposed to the
highest amount of penalty, implying potential population health benefits from higher penalties under the ban.
These results should be interpreted with caution. First, smoking bans in outdoor public places have been
implemented for the most part to address exposure to secondhand smoke. Without information on changes in second-
hand smoke attributed to the policy change, I cannot determine whether the outdoor smoking ban had significant
impacts on its goal. Also, additional information on the place of smoking would complement this study by investigating
whether outdoor smoking ban just displace smokers from nonsmoking places or whether the policy change encourages
positive behavioral changes in other settings through norm spreading (Azagba, 2015). Second, it is plausible that ban-
ning smoking cigarettes made smokers switch to smokeless tobacco to circumvent the restrictions, though I am not able
to examine this behavior and its health impacts due to lack of data.
TABLE 6 Robustness check
Variable
Current
smoking
Quit attempt (among
current smokers)
Smoking intensity (fraction among current smokers)
40 or more
cigarettes per day
20–39 cigarettes
per day
19 or fewer
cigarettes per day
Main
specification
0.005 (0.005) 0.048
**
(0.022) −0.003 (0.004) −0.011 (0.020) 0.014 (0.020)
Panel A. Year 2009–2013 (before the implementation of the national indoor smoking ban)
Outdoor
smoking ban
0.008 (0.006) 0.084
***
(0.027) −0.00008 (0.005) −0.029 (0.026) 0.029 (0.026)
Observations 41,249 9,810 9,810 9,810 9,810
Baseline
outcome
means
0.238 0.283 0.018 0.308 0.674
Panel B. Only including the city of Seoul & synthetic control groups (no ban until 2013), 2009–2013
Outdoor
smoking ban
−0.0002
(0.007)
0.100
**
(0.043) −0.004 (0.009) −0.003 (0.037) 0.007 (0.037)
Observations 21,038 4,728 4,728 4,728 4,728
Baseline
outcome
means
0.225 0.307 0.022 0.258 0.720
Panel C. Placebo tests: Pseudotreatment assignment (period 2009–2010)
Outdoor
smoking ban
−0.007
(0.008)
0.014 (0.024) −0.013 (0.010) −0.005 (0.035) 0.017 (0.035)
Observations 22,655 5,526 5,526 5,526 5,526
Panel D. Placebo tests: Pseudocontrol assignment (period 2011–2014)
Outdoor
smoking ban
0.006 (0.007) −0.015 (0.029) −0.004 (0.006) −0.020 (0.024) 0.023 (0.024)
Observations 30,810 6,848 6,848 6,848 6,848
Note. Standard errors, in parentheses, are clustered at city level. Each column represents separate regression. Regressions include individual fixed effects, year
dummies, annual earned income amounts, and gross regional domestic product per capital (province level).
*
Significant at.1.
**
Significant at.5.
***
Significant at.01.
KO 13

ACKNOWLEDGEMENTS
I thank Anthony Lo Sasso, Darren Lubotsky, Lisa Powell, Nicholas Tilipman, Emily Stiehl, Sherry Glied, Hye Myung
Lee, and seminar participants at Korea Labor Institute, Chung-Ang University, and University of California, San Fran-
cisco for their valuable comments.
CONFLICT OF INTEREST
Nothing to declare.
FUNDING INFORMATION
None.
ORCID
Hansoo Ko https://orcid.org/0000-0002-3321-0763
REFERENCES
Adda, J., & Cornaglia, F. (2010). The effect of bans and taxes on passive smoking. American Economic Journal: Applied Economics,2(1),
1–32. https://doi.org/10.1257/app.2.1.1
Anger, S., Kvasnicka, M., & Siedler, T. (2011). One last puff? Public smoking bans and smoking behavior. Journal of Health Economics,30,
591–601. https://doi.org/10.1016/j.jhealeco.2011.03.003
Barber, P. (2015). We don't need an “evidence base”for a smoking ban in outdoor spaces. BMJ,350, h1442. https://doi.org/10.1136/bmj.
h1442
Been, J. V., Nurmatov, U. B., Nawrot, T. S., van Schayck, C. P., & Cheikh, A. (2014). Effect of smoke-free legislation on perinatal and child
health: A systematic review and meta-analysis. Lancet,383, 1549–1560. https://doi.org/10.1016/S0140-6736(14)60082-9
Boes, S., Marti, J., & Maclean, J. C. (2015). The impact of smoking bans on smoking and consumer behavior: Quasi-experimental evidence
from Switzerland. Health Economics,24, 1502–1516. https://doi.org/10.1002/hec.3108
Carton, T. W., Darden, M., Levendis, J., Lee, S. H., & Ricket, I. (2016). Comprehensive indoor smoking bans and smoking prevalence. Ameri-
can Journal of Health Economics,2(4), 535–556. https://doi.org/10.1162/ajhea00061
Chaloupka, F. J., & Warner, K. E. (2000). The economics of smoking. In A. J. Culyer, & J. P. Newhouse (Eds.), Handbook of health economics.
Amsterdam: Elsevier Science B.V.
Chapman, S. (2008). Should smoking in outside public spaces be banned? BMJ,337, a2804. https://doi.org/10.1136/bmj.a2804
Cho, H. J. (2014). The status and future challenges of tobacco control policy in Korea. Journal of Preventive Medicine and Public Health,47,
129–135. https://doi.org/10.3961/jpmph.2014.47.3.129
Cooper, M. T., & Pesko, M. F. (2017). The effect of e-cigarette indoor vaping restrictions on adult prenatal smoking and birth outcomes. Jour-
nal of Health Economics,56, 178–190. https://doi.org/10.1016/j.jhealeco.2017.10.002
Frazer, K., Callinan, J. E., McHugh, J., van Baarsel, S., Clarke, A., Doherty, K., & Kelleher, C. (2016). Legislative smoking bans for reducing
harms from secondhand smoke exposure, smoking prevalence and tobacco consumption. Cochrane Database of Systematic Reviews,2,
1–165. https://doi.org/10.1002/14651858.CD005992.pub3
Hahn, E. J. (2010). Smokefree legislation: A review of health and economic outcomes research. American Journal of Preventive Medicine,39
(6S1), S66–S76. https://doi.org/10.1016/j.amepre.2010.08.013
Hankins, S., & Tarasenko, Y. (2016). Do smoking bans improve neonatal health? Health Services Research,51(5), 1858–1878. https://doi.org/
10.1111/1475-6773.12451
He, J., Vupputuri, S., Allen, K., Prerost, M. R., Hughes, J., & Whelton, P. K. (1999). Passive smoking and the risk of coronary heart disease: A
meta-analysis of epidemiologic studies. New England Journal of Medicine,340, 920–926. https://doi.org/10.1056/NEJM199903253401204
International Agency for Research on Cancer (2004). IARC monographs on the evaluation of carcinogenic risks to humans, volume 83:
Tobacco smoke and involuntary smoking. http://monographs.iarc.fr/ENG/Monographs/vol83/index.php
Jefferis, B. J., Lowe, G. D. O., Rumley, W. A., Lawlor, D. A., Ebrahim, S., Carson, C., …Whincup, P. H. (2010). Secondhand smoke (SHS)
exposure is associated with circulating markers of inflammation and endothelial function in adult men and women. Atherosclerosis,208,
550–556. https://doi.org/10.1016/j.atherosclerosis.2009.07.044
Johns M, Farley SM, Rajulu DT, Kansagra SM, Juster HR. 2014. Smoke-free parks and beaches: An interrupted time-series study of behav-
ioural impact in New York City. Tobacco Control 0: 1-4. https://doi.org/10.1136/tobaccocontrol-2013-051335
Jones, A. M., Laporte, A., Rice, N., & Zucchelli, E. (2015). Do public smoking bans have an impact on active smoking? Evidence from the
UK. Health Economics,24, 175–192. https://doi.org/10.1002/hec.3009
Kim, S. J., & Seldon, B. J. (2004). The demand for cigarettes in the Republic of Korea and implications for government policy to lower ciga-
rette consumption. Contemporary Economic Policy,22, 299–308. https://doi.org/10.1093/cep/byh021
Kurtulus, F. A. (2016). The impact of affirmative action on the employment of minorities and women: A longitudinal analysis using three
decades of EEO-1 filings. Journal of Policy Analysis and Management,35,34–66. https://doi.org/10.1002/pam.21881
14 KO

Kvasnicka, M., Siedler, T., & Ziebarth, N. (2018). The health effects of smoking bans: Evidence from German hospitalization data. Health
Economics,27, 1738–1753. https://doi.org/10.1002/hec.3798
Lee, C. W., & Kahende, J. (2007). Factors associated with successful smoking cessation in the United States, 2000. American Journal of Public
Health,97, 1503–1509. https://doi.org/10.2105/AJPH.2005.083527
Lee, H. A., Park, H., Kim, H., & Jung-Choi, K. (2014). The effect of community-level smoke-free ordinances on smoking rates in men based
on Community Health Surveys. Epidemiology and Health,36,1–11, e2014037. https://doi.org/10.4178/epih/e2014037
Malarcher, A., Dube, S., Shaw, L., Babb, S., & Kaufmann, R. (2011). Quitting smoking among adults—United States, 2001–2010. Morbidity
and Mortality Weekly Report,60, 1513–1519.
McGeary, K. A., Dave, D. M., Lipton, B. J., & Roeper, T. (2017). Impact of comprehensive smoking bans on the health of infants and children.
NBER Working Paper No. 23995,,1–53. https://doi.org/10.3386/w23995
Meyers, D. G., Neuberger, J. S., & He, J. (2009). Cardiovascular effect of bans on smoking in public places: A systematic review and meta-
analysis. Journal of the American College of Cardiology,54, 1249–1255. https://doi.org/10.1016/j.jacc.2009.07.022
Murray, R. P., Gerald, L. B., Lindren, P. G., Connett, J. E., Rand, C. S., & Anthonisen, N. R. (2000). Characteristics of participants who stop
smoking and sustain abstinence for 1 and 5 years in the Lung Health Study. Preventive Medicine,30, 392–400. https://doi.org/10.1006/
pmed.2000.0642
Nakamura, M., Oshima, A., Ohkura, M., Artega, C., & Suwa, K. (2014). Predictors of lapse and relapse to smoking in successful quitters in a
varenicline post hoc analysis in Japanese smokers. Clinical Therapeutics,36, 918–927. https://doi.org/10.1016/j.clinthera.2014.03.013
Öberg, M., Jaakkola, M. S., Woodward, A., Peruga, A., & Prüss-Ustün, A. (2011). Worldwide burden of disease from exposure to second-hand
smoke: A retrospective analysis of data from 192 countries. The Lancet,377, 139–146. https://doi.org/10.1016/S0140-6736(10)61388-8
Odermatt, R., & Stutzer, A. (2015). Smoking bans, cigarette prices and life satisfaction. Journal of Health Economics,44, 176–194. https://doi.
org/10.1016/j.jhealeco.2015.09.010
Oh, I. H., Yoon, S. J., Yoon, T. Y., Choi, J. M., Choe, B. K., Kim, E. J., …Park, Y. H. (2012). Health and economic burden of major cancers
due to smoking in Korea. Asian Pacific Journal of Cancer Prevention,13, 1525–1531. https://doi.org/10.7314/APJCP.2012.13.4.1525
Okoli, C., Johnson, A., Pederson, A., Adkins, S., & Rice, W. (2013). Changes in smoking behaviours following a smokefree legislation in
parks and on beaches: An observational study. BMJ Open,3,1–6, e002916. https://doi.org/10.1136/bmjopen-2013-002916
Organization for Economic Cooperation and Development (2017). Daily smokers (indicator). https://doi.org?0.1787/1ff488c2en(Accessed on
December 31, 2017)
Pan, A., Wang, Y., Talaei, M., Hu, F. B., & Wu, T. (2015). Relation of active, passive, and quitting smoking with incident type 2 diabetes: A
systematic review and meta-analysis. The Lancet Diabetes & Endocrinology,3, 958–967. https://doi.org/10.1016/S2213-8587(15)00316-2
Raw, M., McNeill, A., & West, R. (1998). Smoking cessation guidelines for health professionals—A guide to effective smoking cessation inter-
ventions for the health care system. Thorax,53,S1–S18. https://doi.org/10.1136/thx.53.2008.S1
Rocha, R., & Soares, R. R. (2010). Evaluating the impact of community-based health interventions: Evidence from Brazil's Family Health Pro-
gram. Health Economics,19, 126–158. https://doi.org/10.1002/hec.1607
Semple, S., Creely, K. S., Naji, A., Miller, B. G., & Ayres, J. G. (2007). Secondhand smoke levels in Scottish pubs: The effect of smoke-free leg-
islation. Tobacco Control,16, 127–132. https://doi.org/10.1136/tc.2006.018119
Shetty, K. D., DeLeire, T., White, C., & Bhattacharya, J. (2010). Changes in U.S. hospitalization and mortality rates following smoking bans.
Journal of Policy Analysis and Management,30,6–28. https://doi.org/10.1002/pam.20548
Statistics Korea. 2017. Korean statistical information services. http://kosis.kr/eng/ (Accessed on December 31, 2017)
Tan, C. E., & Glantz, S. A. (2012). Association between smoke-free legislation and hospitalizations for cardiac, cerebrovascular, and respira-
tory diseases: A meta-analysis. Circulation,126, 2177–2183. https://doi.org/10.1161/CIRCULATIONAHA.112.121301
Thomson, G., Edwards, R., & Woodward, A. (2008). Should smoking in outside public spaces be banned? BMJ,337, a2804. https://doi.org/10.
1136/bmj.a2806
U.S. Department of Health and Human Services. 2014. The health consequences of smoking—50 years of progress: A report of the Surgeon
General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for
Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. https://www.cdc.gov/tobacco/data_statistics/sgr/
50th-anniversary/index.htm
Vangeli, E., Stapleton, J., Smit, E. S., Borland, R., & West, R. (2011). Predictors of attempts to stop smoking and their success in adult general
population samples: A systematic review. Addiction,106, 2110–2121. https://doi.org/10.1111/j.1360-0443.2011.03565.x
West, R., McNeill, A., & Raw, M. (2000). Smoking cessation guidelines for health professionals: An update. Thorax,55, 987–999. https://doi.
org/10.1136/thorax.55.12.987
World Health Organization (2013). WHO report on the global tobacco epidemic, 2013: Enforcing bans on tobacco advertising, promotion and
sponsorship. Geneva: WHO. http://www.who.int/tobacco/global_report/2013/en/
World Health Organization (2015). Global health observatory data repository. http://apps.who.int/gho/data/node.main.
TOB1300MOSTSOLD?lang=en
World Health Organization (2017). WHO report on the global tobacco epidemic, 2017: Monitoring tobacco use and prevention policies. Geneva:
WHO. http://www.who.int/tobacco/global_report/en/
World Health Organization (2018). WHO framework convention on tobacco control. http://www.who.int/fctc/en/
KO 15

Zahra, A., Cheong, H. K., Lee, E. W., & Park, J. H. (2016). Burden of disease attributable to secondhand smoking in Korea. Asia-Pacific Jour-
nal of Public Health,28, 737–750. https://doi.org/10.1177/1010539516667779
Zahra, A., Cheong, H. K., & Park, J. H. (2017). Burden of disease attributable to smoking in Korea. Asia-Pacific Journal of Public Health,29,
47–59. https://doi.org/10.1177/1010539516675696
SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of this article.
How to cite this article: Ko H. The effect of outdoor smoking ban: Evidence from Korea. Health Economics.
2019;1–16. https://doi.org/10.1002/hec.3979
16 KO