American Journal of Public Health | August 2007, Vol 97, No. 81400 | Research and Practice | Peer Reviewed | Alamar and Glantz
RESEARCH AND PRACTICE
Effect of Smoke-Free
Laws on Bar Value and
| Benjamin Alamar, PhD, and Stanton A.
The tobacco industry has claimed
that smoke-free bar laws caused bar
revenues to decline by 30%. After
we controlled for economic vari-
ables, we found that bars located in
areas with smoke-free laws sold for
prices that were comparable to
prices for similar bars in areas with
no smoking restrictions. Other stud-
ies have reported that sales did not
decline, and we also found that
neither price nor sales declined.
of their bars are unfounded. (Am J
Public Health. 2007;97:1400–1402.
Many communities and several states and
provinces have enacted policies prohibiting
smoking in bars. The tobacco industry, work-
ing through the hospitality industry, opposes
August 2007, Vol 97, No. 8 | American Journal of Public Health Alamar and Glantz | Peer Reviewed | Research and Practice | 1401
RESEARCH AND PRACTICE
TABLE 1—Summary of Bar Transaction Data, by Smoke-Free Status: United States,
Smoke-Free BarsBars in Which Smoking Was Allowed
Median sales price,$ (interquartile range)
Gross sales,$ (interquartile range)
Median seller’s discretionary cash flow/sales
Median price-to-sales ratio (interquartile range)
0.40 (0.25–0.58)0.38 (0.21–0.57)
these laws on the basis of the claim that
smoke-free bar laws will reduce profitability
of bars.1,2Scollo et al.3reported that all high-
quality studies have found that smoke-free
laws have no effect or a positive effect on res-
taurant and bar revenues, tourism, and em-
ployment. Another study4found that restau-
rant profits increased by 16% when smoke-
free legislation was enacted.
In our study, we followed the methodology
of a previous study4we conducted to analyze
the effects of smoke-free bar laws on the prof-
itability of bars. We used a database that rec-
ords the purchase price of bars that are sold
and found that bars located in US jurisdictions
with smoke-free ordinances sold for prices that
were comparable to the prices of similar bars
in areas with no restrictions on smoking. Thus,
the existence of smoke-free bar laws had no
detectable effect on the profitability of bars.
We obtained data on sales of bars from the
BIZCOMPS database5for transactions by
Standard Industrial Classification codes (de-
fined by the Statistical Policy Division of the
US Office of Management and Budget to clas-
sify all industries in the US economy)—code
5813 (Drinking Places). BIZCOMPS is a pro-
prietary database that contains information
about businesses that are sold and is used ex-
tensively by business valuation professionals.5
BIZCOMPS provides the details of each
transaction, including sale price, seller’s dis-
cretionary cash flow (defined as the reported
pretax cash received by the owner from the
operation of the restaurant, not including all
noncash costs such as depreciation from the
year of the sale), annual gross revenues, geo-
graphic location, and date of transaction.
Given the broad set of uses of this database,
there is no reason to believe that the samples
were biased in any direction. Should an un-
recognized bias exist, however, it would be
consistent across all data points and not cor-
related to smoke-free policies, because the
database was established for other purposes.
We searched the online version of the data-
base on May 15, 2006.5This search pro-
duced 197 sales of bars between January 31,
1993, and December 31, 2005 (Table 1).
We used the American Nonsmokers’ Rights
Foundation Local Ordinance Database (as of
May 15, 2006) to determine which businesses
were covered by local or state 100% smoke-
free bar laws. An ordinance was deemed to be
100% smoke free only if it did not allow
smoking in separately ventilated rooms and
had no exemptions based on the size of the
bar. We used the date the laws were enacted,
as opposed to the date of implementation, be-
cause any prospective buyer would have
known that the law was to come into effect.
This information would have been factored
into the sales price agreed to by the buyer and
seller of the bar. We defined a dummy variable
to indicate the presence of a smoke-free bar
law (1=a smoke-free law had been enacted
before the transaction date; 0=otherwise).
To control for the economic differences
across time and geographic region, we ob-
tained the per capita gross state products and
annual percentage gross state product growth
for all states for all years in which transac-
tions occurred from the Bureau of Economic
Analysis and unemployment rates by state
and by year from the Bureau of Labor Statis-
tics.6Gross state product is reported in real
terms with a base year of 2005 according to
the standard inflators of the Bureau of Eco-
nomic Analysis. All variables reported in dol-
lars were adjusted to real terms according to
the standard inflators of the Bureau of Eco-
nomic Analysis and a base year of 2005.7To
further control for any secular trend, we used
a year variable, with 1992 counted as year 0.
The ratio of the transaction price to gross
revenue is a standard valuation measure to
compare transactions. Dividing the price of
the bar by its annual total gross sales allowed
for a standardization across bars of various
sizes. This is a particularly good measure of
comparative value in service industries that
are not capital intensive.8The price-to-sales
ratio was regressed on the smoke-free dummy
variable, the ratio of the seller’s discretionary
cash flow to sales, which represents the profit
margin of the business; per capita gross state
product in dollars; gross state product per-
centage growth rate; statewide unemploy-
ment percentage; and time.
White’s test9on the residuals from ordinary
least squares regression indicated the presence
of heteroscedasticity. To correct for this prob-
lem, a weighted least squares regression was
used with seller’s discretionary cash flow as
the weight. This procedure assumes that the
variance in the error terms is inversely propor-
tional to the seller’s discretionary cash flow;
thus, the larger (higher seller’s discretionary
cash flow) bars were weighted more heavily.
White’s test on the weighted residuals did not
reject homoscedasticity, and the weighted
least squares regression results were reported.
The parameter estimate for the smoke-free
variable did not even approach statistical sig-
nificance (P=.723; Table 2), indicating that
the presence of a smoke-free law had no de-
tectable effect on the sale price of a bar.
Of the control variables, the ratio of the
seller’s discretionary cash flow to sales, gross
state product, gross state product growth rate,
unemployment, and time were significant.
The positive and significant coefficient for
gross state product growth was not surprising,
because high growth rates indicate that busi-
nesses are growing and thus are more prof-
itable. The negative coefficient for the gross
American Journal of Public Health | August 2007, Vol 97, No. 81402 | Research and Practice | Peer Reviewed | Cummins et al.
RESEARCH AND PRACTICE
TABLE 2—Determinates of the
Price-to-Sales Ratio of Bars (Ordinary
Least Squares): United States,
Per capita gross state
Gross state product
3.284 (±1.073) .003
state product suggests that as the gross state
product rises, bar values decline, perhaps be-
cause areas of higher wealth may have more
business competing against each other, which
reduces profitability. The negative coefficient
for time suggests that real bar values have
been decreasing over time, which indicates
that the market has become more competitive
over time. The coefficient for per capita un-
employment was not significant.
We further tested how robust the model
was by adding a quadratic time factor. The
quadratic time coefficient was not significant,
and its inclusion had no effect on the smoke-
Although several studies have reported no
effect or a positive effect of smoke-free bar
laws on aggregate bar revenues as reflected in
tax receipts,3this study was the first to exam-
ine the effect of smoke-free laws on the value
of individual bars. Because the value of a bar
is directly related to the profitability of the bar,
this study also examined the profitability of in-
dividual bars. We can have a high level of cer-
tainty regarding the negative conclusion in this
brief because the available data have 80%
power to detect a 12% change in the price-to-
sales ratio, assuming the other control vari-
ables were orthogonal to the policy variable
and with a significance level of .05, which is
well below the 30% decline in business that
the tobacco industry has often claimed.10,1 1
(The power of this study to detect a 30%
change in the price-to-sales ratio was 99.94%.)
The tobacco industry has argued that bar
owners would be financially burdened by
smoke-free policies.1Despite this rhetoric,
the results from this study that the price-to-
sales ratio was not affected, combined with
the previous results on bar revenues,3,12,13in-
dicate that bar owners’ worries that smoke-
free laws will reduce the value of their bars
About the Authors
The authors are with the Center for Tobacco Control Re-
search and Education, Department of Medicine (Cardiol-
ogy), University of California, San Francisco.
Requests for reprints should be sent to Stanton A.
Glantz, PhD, 530 Parnassus St, Suite 366, University of
California, San Francisco, CA 94143-1390 (e-mail:
This brief was accepted July 10, 2006.
Both authors originated the study and wrote and re-
vised the article. B. Alamar collected the data and ran
the statistical analysis.
This research was funded by the National Cancer Insti-
tute (grant CA-61021).
The funding agency had no involvement in the con-
duct of the research or the preparation and revision of
Human Participant Protection
No protocol approval was needed for this study, be-
cause no human participants were involved.
1.Ritch WA, Begay ME. Strange bedfellows: the his-
tory of collaboration between the Massachusetts Res-
taurant Association and the tobacco industry. Am J
Public Health. 2001;91:598–603.
2. Dearlove JV, Bialous SA, Glantz SA. Tobacco indus-
try manipulation of the hospitality industry to maintain
smoking in public places. Tob Control. 2002;11:94–104.
3.Scollo M, Lal A, Hyland A, Glantz S. Review of
the quality of studies on the economic effects of
smoke-free policies on the hospitality industry. Tob
4. Alamar B, Glantz S. Smokefree ordinances in-
crease restaurant profit and value. Contemp Econ Policy.
5.Sanders J. BizComps business sale statistics; May
2006. Available at: http://www.bvmarketdata.com. Ac-
cessed May 15, 2006.
6.Bureau of Economic Analysis. Regional economic
accounts: state and local area personal income. Avail-
able at: http://www.bea.gov/regional/index.htm. Ac-
cessed May 16, 2006.
ment statistics: January 2006. Available at: http://
www.bls.gov/data/home.htm. Accessed May 16, 2006.
8. Pratt S, Schweihs R. Valuing a Business: The Anal-
ysis and Appraisal of Closely Held Companies. 4th ed.
New York, NY: McGraw-Hill; 2000.
9.White H. A heteroscedasticity-consistent covari-
ance matrix estimator and a direct test for heterosce-
dasticity. Econometrica. 1980;48:817–838.
10. Traynor MP, Begay ME, Glantz SA. The new
tobacco industry strategy to prevent local tobacco con-
trol. JAMA. 1993;270:479–486.
11. Samuels B, Glantz SA. The politics of local to-
bacco control. JAMA. 1991;266:2110–2117.
12. Health Scotland. International review of the health
and economic impact of the regulation of smoking in pub-
lic places. 2005. Available at: http://www.healthscotland.
com/documents/447.aspx. Accessed May 2, 2007.
13. Tang H, Cowling DW, Lloyd JC, et al. Changes of at-
titudes and patronage behaviors in response to a smoke-
free bar law. Am J Public Health. 2003;93:611–617.
Bureau of Labor Statistics. Local area unemploy-