Loading...
HomeMy WebLinkAboutCOM 0013.083 2002-2004 ~ Testimony on Bill 260, Draft 2: Relating to Smoking in Certain Places May 5, 2003 ; =Ltivtt9 Dr. Fred Holschuh ~ , ~~6~03..___ ~~~~~C.f:fi Chair, Human Services and Ecatomic Development Committee ~ ~ Gxunty (,;pencil - Q~s Hawai`iCountyCouncil ~ 25 Aupuni Street `r"°^""°"rree'•_'n"e~"' HiIo.HI96730 F/nnalnlu, ff/ 96x1,' . Trt BOR) 432-9[ l ~ r'ac: ~xob'~ sz.r-9n~z Dear Chair Holschuh and members of the Human Services and Economic Development Governing Hoard COn11T1lltee: ./ay Adaddnc'k. Ph. D..Clruu' ('mxer Xrsenrch Center nJNa,er„ % 'The Coalition for a Tobacco Free Hawaii appreciates this opportunity to testify in strong Stacy evensc,t rice Cba,r nauarI A9ed,ea/ support of BiII 260, Draft 2 Relating to Smoking in Certain Places. We thank the .1'erese dr~sncfanon or~aniza[ion. Hawaii County Council f' or taking up the important health and safety issue of, and for ,[,nerican cancer sae,em the very strong support of the Council for extending smokefree protection to all Nanp('i Pacific. Inc An:,-clean Neorr ~tssoclanun workplaces, including restaurants, bars, nightclubs and bowliing alleys. The Coalition ~f su,ra(~l strongly supports protection trom second hand smoke for all workers in all workplaces, A,ner(can Lung Association °f/laivoi~' and supports the passage of Bill 260 Draft 2 as a tremendous step in accomplishing this. Cancer 2erearch Censer ojNamrd r As you know. if passed as currently written, Hawai"i County will have the most Hawal'i Island ToAacca Free l'a,.mersh;p ens, Na,varr comprehensive workplace protection from exposure to second hand smoke in the state, Nawa; r 1,~/mid Tobacco Free and will have one of the best laws in the country. l'ar[ner.chip- Wesr Hmvm'i Ha maf'I Srnrc Deparrmenr of Education Na,~~a~ ~ scam Deparrn,enl ~ am Clifford Chang Director of the Coalition for a Tobacco Free Hawaii, a statewide of /(eallh association of over 140 member organizations, businesses and individuals committed to 1/uu¢,i7 M1ledicn[Ase'ociarinn Norvarl Medtca~ working together to reduce the negative impact tobacco use has on the health and Services Associarion wellbeing of I Iawai`i's resideuis and visitors. The seven fomidim* members of the Wa,'nnne Coa.vr Comprehensive o Nealrn r~en,er Coalition are the American Cancer Society Hawai` i Pacific, Inc., American Heart Tobacco-hiec Kaum ~tmn~ robocco hree Porrnersh!/, Association of Hawai'i_ American Lung Association of Hawaii, Cancer Research Center individuals of Hawaii, Hawaii State Department of Education, Hawaii State Department of Health, KIm h1, Sete, Birn(e npe-r)~ean tee, t'1, a and the tlawai'i Medical Association. Ken Miller 13e11, Pnrema,,, N.CD I would like to touch on two issues: :I nU,orty ,Sogu,bu, [ ) sryf~ ~ I I Iighlight recent information that continue to strengthen the science base regarding ClffJord C'Irang, h1PN. CHIiS otr~~ror the health impact of second hand smoke, and Herh Kuch ('annn,ndcn[ions Coordfnorar 2. Address the concern that restaurants and bars wi II lose business should they go Ton, don .1yurons Smoke-frec9. ass Boars ddencac}~ Canrdmamr ' Comm. No. 1'undcrl m parr hT the RBf• TO: 'nherr N'oad Juhnsnn Fwmdar°°~ RCf. DC1Q MAY f1 7~~~ Dc Fred Holschuh; Human Services and F,conomic Development Committee May 6, 2003; page 2 Regarding the negative health effects of exposure to second hand smoke, you have heard previous testimony indicating that over 53,000 Americans die each year as a direct result of exposure to second hand smoke- over 50,000 from heart disease and over 3000 from lung cancer. A recent study conducted in New Zealand compared the blood levels of cotinine of 40 non-smoking hospitality workers working in restawaunts and bars that allow smoking with 50 people who work in anon-smoking work environment cotinine is a derivative of nicotine indicates the level of exposure a person has to tobacco smoke. ]n this study saliva samples were taken and a questionnaire answered before and after each workshift. Many of the non-smoking hospitality workers had cotinine values as high as 5. I ng/ml (smokers generally have cotinine levels of 7.0 ng/ml and higher). Previous studies have demonstrated Neat a cotinine level of .4 ng/ml, is associated with a lifetime increased risk dying of lung cancer to 1 in 1000, and for heart disease to I in 100. "fhe study also found the hospitality workers reported more reparatory infections than those working in anon-smoking environment, and reaffirmed that hospitality workers are exposed [o levels of second hand smoke of 1 S to 4.4 times greater than Neat of someone living with a smoker. A recent study conducted in Japan, reported in the July 25, 2001 issue of the Journal of the American Medical Association (JAMA), examined the coronary flow velocity reserve (CFVR) in 30 healthy Japanese men, I S smokers and 15 nonsmokers exposed to second hand smoke for 30 minutes in a smoking room in the hospital. This study found that after only a thirty mimic exposure, the carbon monoxide levels in the nonsmokers increased dramatically, and that exposure to second hand ,smoke reduced the CFVR in the nonsmokers, sufficient to cause endothelial dysfunction of the coronary circulation of nonsmokers. Another study examining 5.196 nonsmoking police officers in Hong Kong exposed to second hand smoke, reported in the August 9, 2000 issue of JAMA. "Chas study found a clear association and a dose response relationship between second hand exposure at work and respiratory symptoms in both men and women. And for a bit of diversity, a recent issue of the American Journal of Epidimiology reported on a study examnng the effect of second hand smoke on pet cats. Cats living with smokers have more than twice Nie risk of developing malignant lymphoma than those lip ing with non-smokers, with the risk 3 times greater for those cats living with a smoker for 5 years or ma~e Some people have Cestitied questioning whether second hand smoke is a Class A Carcinogen, pointing to Judge Osteen's ruling overturning the chapters related to second hand smoke as a Class A carcinogen in the 1992 FDA report. On the Philip Morris's website is the following statement regarding this ruling: The court - in a decision that EPA is now appealing -rejected the EPA assessment, largely based on findings that the agency did not adhere to applicable legal requirements and failed to adequately explain its results. Dr. Fred Holschuh; Human Services and I~;conomic Development Committee May 6, 2003; page 3 Note that Philip Morris itself did not say the ruling was invalidated because the conclusions were wrong, only that the FDA did not adhere to legal requirements and did not adequately explain the results. In addition, the ruling did not overturn any of the FDA conclusions regarding ETS exposure causing other diseases, including heart disease and asthma, in nonsmokers. We are pleased that the Hawaii Island and the Kona-Kohala Chambers of Commerce, which represent approximately 1300 businesses on the Big Island and whose letters of support have been submitted separately, are in toll support of this law. 'hhe Chambers understand business will not suffer. 1 remind you of the USA'foday national survey that found that 7U% ofthose surveyed indicated they do not patronize bars specifically because of exposure to secondhand smoke, ~9% do not patronize nightclubs, and 48% do not patronize restaurants that allow smoking. As for our Asian visitors, a survey conducted in Hong Kong in September 2002 regarding smoke-free restaurants found that 772 % of the 1320 people randomly surveyed indicated that their dining out habits would not change with smoke-free restaurants, 19.7% indicated they would dine out more often, and only 32% indicated they would dine out less. In the past several months New York City, followed by the entire state of New York, and the entire state of Delaware have passed laws for completely smoke-free workplaces, including restaurants, bars and nightclubs. As of yesterday, Monday, May 5°', all workplaces in Boston became totally smoke-free-including restaurants, bars and nightclubs. Connecticut is capected to follow with the passage of a similar comprehensive law this week. Maine, which has had smoke-tree workplaces, including restaurants, since 1999, is expecting to pass a state law designating bars and nightclubs smoke-flee as well. Passed by a voter initiative, Florida now has a constitutional amendment that makes workplaces including restaurants, smoke-free. As of March 3, 2003 over 300 municipalities across the country have enacted smokefree workplace ordinances including restaurants, over 160 have ordinances that include bar areas within restaurants, and over 90 with ordinances that include all bars. Not one of these ordinances have been repealed since their passage. One would think, had restaurants and bars in these areas lost up to 38% of their business as a result of such laws, as the National Restaurant Association, supported by the tobacco industry continues to claim, that some of these laws would have been repealed. Instead, more and more locations are going completely smoke-free to protect the health of all workers, without financial hardship to businesses. It is simply the right thing to do to provide worker protection from exposure to the harmful substances that make up second hand tobacco smoke. It is time, based on what is known about the harmful effects of exposure to second hand smoke and the lack of any negative economic repercussions, to pass a bill that protects all workers on the Big Island Dr. Fred Holschuh; Human Services and Economic Development Committee May 6, 2003; page 4 Also attached as part of my testimony is a press release from California after 8 years of smoke-free workplaces including restaurants, and five years of smoke-free bars. I would like to share a quote from Jim Abrams, the executive vice president of the California Hotel and Lodging Association. "We in California's lodging industry were understandably nervous ab~mt the potential detrimental effects of a smoking ban would have on our businesses. But none of the doomsday predictions happened. In fact, tourism in California increased each year following passage of AB 13. We can say with certainty to the rest of the country that a workplace smoking ban is good for the lodging industry." Michael Casey, President of the Hotel and Restaurant Employees, Local 2, AFL-C[O in San Francisco stated "The tobacco industry tried to scare our members into thinking their employers would lose business due to the smoking ban, and that they would lose their jobs. I am proud to say our union members and many employers saw through those tactics and stood up for the right to a safe and healthy workplace. "Fhe restaurants, bars and hotels where our members work did not suffer any adverse effects from the smoking ban. If anything, it's been good for business and customers as well as for our members." Both the Hilo l'ribune-Herald and the Honolulu Star Bulletin have reported that the Executive Director of the Hawaii Restaurant Association, Ms. Michelle Van Huessen, has called attention to two studies conducted by Michael Marlow at the California Polytechnic State University indicating the loss of business in restaurants and bars as a result of smoke-free ordinances. I have attached one of the articles to my testimony for your reference. You should note that both of these studies were paid for entirely by Philip Morris, and the co-author of the second study, Mr. John Dunham, is an an employee of Philip Morris. You should also note that both studies are not based on comparing revenue figures before and after the passage of smoke-free ordinances, but on predictions of restaurant and bar owners of what the impact of such ordinances will be. Since 1987 the tobacco industry has been predicting that smoke-free restaurant ordinaces will cause a 30% loss of business to restaurants -this predictiai has never come true. Yet still the tobacco industry continues to scare restaurant and bar owners with "studies" that predict such loss of business. It simply is not true-not in California, New York, Florida, Delaware, Boston, Oahu, Maui or Kaua` i, and it will not be true if Bill 260 is passed in Hawaii County. Each time a municipality has addressed this issue, the claim that restaurants and bars will lose business is raised. Over 57 studies globally have been conducted examining whether smoke-free restaurant and bar ordinances hurt business. Only those funded by the tobacco industry have found any negative economic impact, and all based on predicted or anecdotal evidence, not tas or ,ales receipts. As for the concern for Asian tourists, one indication that things are changing even in Asia iv that Hong Kong is currently considering the passage of legislation making all their restaurants smoke-tree. Dr. Fred Holschuh; Human Services and f:conomic Development Committee May 6, 2003; page 5 Since the Oahu smoke-free restaurant ordinance has gone into effect on July 1 preliminary indications are that restaurants are doingjust tine. While it is too early to have good data based on sales revenue, anecdotal information seems to indicate restaurants have not lost any business, and tourism has not fallen off. Reporter Grace Lee, KGMB 9 News, reported that one month after O`ahu's ordinance took effect, the Hawaii Restaurant Association, which had opposed the Oahu ordinance, called 40 of their member restaurants. None reported any negative effect on business. with a number reporting in increase in business. KGMB 9 called 14 restaurants on their own, and again, none reported any loss of business and a number reported an increase in business. (I did call the Hawaii Restaurant Association to see if 1 could get the names of the restaurants that they called, but the Association would not release this information to me.) The Coalition believes all workers, including those working in bars and nightclubs, deserve a workplace free from second hand smoke. Second hand smoke exposure in restaurants is often double what is found in a typical workplace, and it is 4 to 6 times higher in bars than in offices. However, the Coalition also recognizes that these laws work best when they have strong public support. While approximately 70% of fiawai`i's residents support smoke- free workplaces including restaurants, the number is less for the inclusion of bars. Should bars and nightclubs be exempted from the Bill 260 at this time, the Coalition is committed to increasing public support for all workplaces, including bars and nightclubs, to be smoke-free. Going smoke-free costs nothing fur businesses to implement; in fact, businesses will save money in reduced maintenance costs. hnplementatiun will not result in any loss of revenue or a decrease in tourism, in fact, it may result in increased revenue. It protects workcrs' health and their working conditions, which does much to improve their morale. Passage of Bill 260 will demonstrate that all counties of Hawaii are committed Hawaii being the "Health State." Thank you for this opportunity to testify. 'Thank you again for the opportunity to testify. We look forward to the next steps in the passage of the most comprehensive smoke-free workplace bill in the state. Since el Cli rd Chang, MPH, CH S Dire for NATIONAL CONFERENCE ON TOBACCO OR HEALTH www.tobacwcontrolconference.org FOR IMMEDIATE RELEASE CONTACT: Darlene Colvin November 20, 2002 (310) 552-6922 CALIFORNIA'S HEALTH, LABOR and BUSINESS LEADERS CONFIRM 8-YEAR SUCCESS OF SMOKE-FREE WORKPLACES SAN FRANCISCO (Nov. 2Q 2002) -While other cities, states and countries debate smoke-free workplace laws and listen to unsubstantiated threats of economic doom, California's health, labor and business leaders confirmed the success of California's Law for aSmoke-free Workplace (Assembly Bill 13). State Health Director Diana Bonta, along with leaders of the California Labor Federation AFL-CIO, Hotel & Lodging Association, small businesses and others celebrated the eight-yeaz success of AB 13 -California's most controversial employee protection law -and the five-yeaz anniversary of smoke-free bazs. "The Califomia Labor Federation is proud to have sponsored one of the most sweeping worker health protection measures ever enacted in California or the United States," said Tom Rankin, president, California Labor Federation, AFL-CIO. "We know cigazette smoke contains over 4,000 chemicals, including dozens of known human carcinogens. Workers should not have to choose between their jobs and their health." "We in California's lodging industry were understandably nervous about the potential detrimental effects a smoking ban would have on our businesses," said Jim Abrams, executive vice president of the Califomia Hotel and Lodging Association. "But none of the doomsday predictions happened. In fact, tourism in California increased each year following passage of AB 13. We can say with certainty to the rest of the country that a workplace smoking ban is good for the lodging industry." -more- Smoke-free Workplaces Page 2 "California's Smoke-Free Workplace Law has benefited millions of workers over the past several years by protecting them from the deadly effects of secondhand smoke," added Rankin. "We hope California serves as an example to other cities, states and countries that a smoking ban is good for workers, employers and the public." "More and more of our hotel and motel members have gone fiuther than the law mandated," added Abrams. "While the law required only 35 percent of hotel and motel rooms be non-smoking, most members set aside a much higher percentage due to customer demand. The fact is that the vast majority of customers prefer asmoke-free environment." Echoing statements from eight years ago, John F. Henning, former executive secretary- treasurer, California Labor Federation, AFL-CIO, added, "What was true then is still true today: death by cigazette smoke should not be a condition of employment." "The tobacco industry tried to scare our members into thinking their employers would lose business due to the smoking ban, and that they would lose their jobs," said Michael Casey, President, Hotel & Restaurant Employees, Local 2, AFL-CIO, of San Francisco. "I'm proud to say our union members and many employers saw through these tactics and stood up for the right to a safe and healthy workplace. "The restaurants, bazs and hotels where our members work did not suffer any adverse effects from the smoking ban. If anything, it's been good for business and customers as well as for our members," concluded Casey. "As the author of AB 13, it's immensely gratifying to see these reports that reflect how successful the legislation has been," said Hon. Terry Friedman, Los Angeles County Superior Court Judge and former state Assemblyman. "We fought an uphill battle against the tobacco industry's massive financial and political resources when we passed the smoke-free workplace law. We were successful because we stuck to the facts. Our coalition of labor, business, local government and health groups knew that smoke-free workplaces would benefit everyone. The law's success and the continued enormous, broad-based support prove that." %~~1~7//VI ~ir l'I/ ~7p ~ • ioJ'/'I/I//oIGP~'y Helping you braenre a IIWe seder PATRON SURVEYS AND CONSUMER BEHAVIOR January 6, 2003 Although sales tax data provides the on/y objective measure of the economic impact of smokefree ordinances, it can be difficult to extrapolate from such data internal shifts in consumer behavior and spending. Restaurateurs attempting to prepare for and adapt to smokefree legislation may find it useful to know how many patrons actively seek out smokefree environments, how many actively avoid such environments, and the relative purchasing patterns of both groups. A small number of statistically reliable and veriSable studies have been conducted on patron behaviors and consumer preferences in response to smokefree policies. It is important to distinguish such studies (conducted by independent researchers and academics) from those studies sponsored by the tobacco industry, which rely solely on anecdotal information and restaurateur predictions of economic impact. S Hong Kong: The first comprehensive survey of public opinion on smokefree policies in Asia found strong community support for smokefree dining. Researchers at the University of Hong Kong Medical Center randomly selected 1,320 subjects who were questioned regarding thew tobacco use, demographic characteristics, and dining behaviors, and asked to predict how a smokefree policy would affect their patronage of restaurants. The majority of respondents (77.2%) anticipated no change in their dining behavior. Of those who did predict a change, 19.7% predicted that they would dine out more often and only 3.2% predicted that they would eat out less often. (Lam, T.H.; Janghorbani, M.; Hadley, AJ.; Ho, S.Y.; McGhee, S.M.; Chan, B., 'PubNcc opinbn on smoke-hee poldes in restaurants and predicted effect on pabonage in Hong Kong,' Tobacco Control 11(3): 195.200, September 2002.) v Denver, CO: Forty percent of regular bar patrons would visit bars more frequently if smoking was prohibited, according to a survey conducted by Aspen Media and Market Research. The survey polled 455 Denver residents over the age of 21. Even amongst smokers, who comprise approximately 19% of the Denver adult population, there was a majority of support for smokefree bars and restaurants. ([n.a.], 'Denver health bar patron survey final," (Aspen ReseamhJ, January 15, 2002.) EI Paso, TX: The Institute for Policy and Economic Development conducted a random survey of 845 El Paso residents to assess the probable economic impact of a proposed smokefree ordinance. Of those surveyed, 26.8% responded that they would eat out mare often if restaurants were smokefree, 62.9% responded that they would eat out the same amount, and 10.4% responded that they would eat out less often. Researchers concluded: "Our findings suggest that a smoking ban would increase the dining frequency of nonsmokers by more than would be reduced by smokers - thus, a positive economic impact of smoking restriction could occur." (Johnson, S.h; Soden, D.L.; Ulh, C.; ~brales, J,~PartTS~rt3: po~ana-lysls aTeconomk unpa s u o s e-Gee fndaor as mantes in a-Hi -Paso del tTrie - - - - region. IPED Technical Report 2001-3],' EI Paso, TX: Instlfue for Polfcy and Economic Development, University of Texas, EI Paso, May 2001.) California: A survey conducted for the Department of Health Services measured bar patrons' response to California's smokefree bar law (AB 13). Random digit dial methodology was used to interview 1,000 Californian adults who had recently visited a bar. The survey found that ninety-one percent of bar patrons go to bars more often or have not changed their bar-going behavior since the law was fast enacted in 1998. On average, the study found that bar patrons stay longer in bats and enjoy their visits as much or more than they did prior to the law. (California Department of Health Services, "Support for snake-free bare grows stronger in California: statewide poll of California's bar patrons shows drema0c increase in support for state's 2112-year-0Id law prohibNng snaking in bars' Business tMre, October 16, 2000.) Victoria, Australia: Restaurants would receive more business, not less, if they went smokefree, according [o a study conducted by Auspoll in Victoria, Australia. Using a survey of 816 randomly selected telephone numbers, researchers found that up to 52.5% of respondents felt they would be more likely to attend smokefree venues, while only 29% of respondents felt they would be less likely to attend smokefree venues. The survey concluded that "overall smoke free venues would appear to make people more likely to attend" venues. (Auspoll, January 2000.) Tempe, AZ: A month-long survey of more than five hundred Tempe residents was undertaken to assess residents' response to smokefree ordinances. More than half of those polled (57%) reported that they were more likely to go to a smokefree restaurant; only 9% reported that they would be less likely to go to a smokefree 2530 San Pablo Avenue, Suite J • Berkeley, California 94702 • (510) 841-3032 /FAX (510) 841-3071 7~~~CR4~ ~D`' /Y~~~y Helping you breaMe a little easier restaurant. When asked whether a smokefree baz ordinance would make them more or less likely to go to a baz, 34% of those polled responded that they would be more likely to go and 11 % responded that they would be less likely to go. (O'Neil Associates, Inc., "AtMudes toward smoking in Tempe: report of a survey conducted for American Heart Associaton [sic),' American Heart Association (AHA), 1999.) i• Gilbert, AZ: A survey of 500 Gilbert residents was conducted by O'Neil Associates to assess public attitudes toward smoking. Reseazchets found that 62% of respondents felt they would be more likely to go to a restaurant if they knew it was smokefree, 31% said it made no difference, and 8% felt they would be less likely to go to a restaurant if they knew it was smokefree. When asked if they would be more or less likely to go to a bar if they knew it was smokefree, 44% of respondents felt they would be more likely, 39% said it made no difference, and 11 % felt they would be less likely. (O'Neil Associates, Inc., 'Attitudes toward smoking in Gilbert report of a survey conducted for American Heart Associaton Isis],' American Heart Association (AHA), 1999.) •S Oregon: A survey undertaken by Oregon Smokeless States reported that 8 out of ] 0 Oregon residents did not feel that a restaurant smoking ban would affect how often they eat out. A majority of those remaining felt that they would eat out more often if restaurants were smokefree. ([n.a.j, "State of public opinbn in Oregon: 1998 Smokeless States Tobacco Control Survey -Oregon," (Smokeless States Oregon), [1998].) Massachusetts: Researchers at the University of Massachusetts and Boston University studied the potential impact of smokefree policies on baz and restaurant revenues. Using random-digit dialing techniques, 2,356 individuals were questioned regarding their attitudes towards smoking, secondhand smoke, and tobacco control policies. Approximately two-thirds of respondents reported that smokefree policies would have no effect on their patronage of bars and restaurants. Of those respondents who felt that smokefree policies would have an effect on they restaurant patronage, 31 % stated that [hey would dine out more often. Only 7.9% stated that they would dine out less often. (Biener, L.; Siegel, M., 'Behavior intentions of the public after bans on smoking in restaurants and bars,' American Joumal of Public Health 87(12): 2042-2044, December 1997.) New York, NY: The Smoke Free Air Act went into effect in New York City in April 1995 amid vocal protests that the act would have a devastating impact on the city's restaurant revenues. Four months later, a study was conducted by the prestigious Cornell Center for Hospitality Reseazch to determine the actual impact of the act on consumer behavior. A survey of 389 randomly selected restaurant patrons found that, although there had been some decline in the frequency with which smokers were dining out, nonsmokers were dining out more often. Researchers concluded that nonsmokers, who account for apQroximately 70.4% of New York restaurant revenues, were "more than making up the revenues lost from inconvenienced diners who smoke."(Corsun, D.L.; Young, C.A.; Enz, C.A, "Should NYC's restaureteurs lighten up?: eftects of the city's smoke-freeair act," Comell Hotel and Restaurant Administration Ouartedy:25-33, Apri11996.) w Flagstaff, AZ: In Iune 1993, Flagstaff, Arizona, became the fast city in Arizona to enforce a smokefree restaurant ordinance. A survey conducted by Northern Arizona University assessed the concerns of restaurant representatives prior to the enacttnent of the ordinance and 15 months after the ordinance had taken effect. Initially, restaurant representatives expressed a great deal of concern that the ordinance would have a detrimental effect on sales and would be difficult to enforce. After enactment, however, the majority of respondents reported that the ordinance had no or no known effect on business and 94% found the ordinance easy or very easy to enforce. (Sciacca, J.P., 'A mandatory smolting ban in restaurants: concerns versus experiences,' Joumal o/CommunityHeakh 21(2): 133-150, April 1996.) [FS-15] W:~FILES~MA.TERIAL~FACTSHTS~Public Support Fact Sheet.doc 13 Review of the quality of studies on the economic effects of smoke-free policies on the hospitality industry M Scollo, A Lal, A Hyland, S Glantz Tobacco ConNO12003; 12:13-20 Objective: To compare the qualiy and funding source of studies concluding a negative economic impact of smoke-free policies in the hospitaliy industry to studies concluding no such negative impact. Data sources: Researchers sought all studies produced before 31 August 2002. Articles published in scientific journals were I«ated with Medline, Science Citation Index, S«ial Sciences Citation Index, Current Contenh, Psychlnfo, Econlit, and Healthstar. Unpublished studies were locoted from tobacco company wabsires and through interne[ searches. authorsa affil arionsar Study selection: 97 studies that made statements about economic impact were included. 93% of the studies I«ated met the selection criteria as determined by consensus between multiple reviewers. Data exfradion: Findings and characteristics of studies (apart from funding source) were classified Comespoaden<e to: independently by hvo researchers. A third assessor blind to both the objective of the present study and Mkhelle Scollo, VicH°olth to funding source also classified each study. Centre For Tobacco Data s nthesis: In studies concluding a negative impact, the odds of using a subjective outcome Control, The Cancer Y Council, Vicroria, measure was 4.0 Hmes j95°~ confidence interval (Cl) 1.4 to 9.6; p 0.007) and the odds of not being 100 Drummond Street, peer reviewed was 20 times (95°~ CI 2.6 to 166.7; p ~ 0.004) that of studies concluding no such Carlton, 3053, Aushalia; n alive im act. All of the studies concludin a ne alive im act were su y michelle.scolloQ ~ p 9 9 p pported b the tobacco indus- canc°rvlc.org.au M~ 94% of the tobacco industry supported studies concluded o negative economic impact compared ro none of the non-industry supported studies. Received Conclusion: All of the best designed studies report no impact or a positive impost of smoke-free res- AOeP~ 2002. courant and bar lows on sales or employment. Polirymakers can act to protect workers and patrons 20 txrober 2002 from the roxins in secondhand smoke confident in rejecting industry claims that there will be an adverse economic impact. moke-free workplace policies reduce both exposure and her colleagues have repeatedly called on advocates to to secondhand tobacco smoke and cigarette more forcefully draw to legislators' attention the superior sci- consumption."' Smoke-free restaurants and bars Simi- entific quality of the evidence base relied upon by public larly reduce exposure to tobacco smoke toxins among health groups in calling for clean air legislation."" hospitality workers and patrons but also represent a serious This paper compares the quality of evidence and conclu- business threat to the tobacco industry. lions about the economic impact of smoke-free laws on the In California in 1987, a I00% smoke-free restaurant hospitality industry based on the type of data used, how the ordinance m Beverly Hills was rolled back, partly in response studies are designed, analysed and interpreted, and [he fund- toclaims that the ordinance was responsible for reducing res- ing source. courant revenues by 30%, claims which later turned out to be unsubstantiated.' Since then, tobacco companies and allied METHODS groups have routinely predicted that enactment of such legis- Data source lotion would severely impact restaurant and baz sales and Studies included in this analysis are listed in a comprehensive employment."' Health advocates, by contrast, have presented summary produced by the VicHealth Centre [or Tobacco studies indicating that no such adverse effects actually Control." Centre researchers attempted to locate all studies occurred.'° produced in English before 31 August 2002 that purported to Polirymakers are typically presented with a large amount of assess [he economic impact of smoke-free policies in the hos- conflicting material, with evidence ranging from anecdotes pi[ality industry. Peer reviewed articles were located with about individual businesses" to scientific studies analysing Medline, Science Citation Index, Social Sciences Citation objective information collected independently across an entire Index, Current Contents, Psychlnfo, Econlit, and Healths[ar hospitality sector." Such data are often confusing to interpret using the terms smok• and restaurants, bars, hospitality, eco- and it is difficult for polirymakers to reach an evidence based notnic, regulation and law. Unpublished studies were also conclusion. In [heu case study of deliberations by [he included in [he analysis. These studies were located from a Maryland Occupational Sa&ty and Health Advisory Board, compilation by the Alberta Tobacco Control Centre,'° by a Mon[ini et a( demonstrate [ha[ [hose opposing proposed request [o members of [he International Union Against Can- smoke-free workplace regulations lodged twice the number of ceYS International Tobacco Control Network (GLOBALink), submissions as [hose supporting it, but [ha[ evidence from and an examination of hospitality industry websi[es and [he opponents was substantially less scientifically rigorous than websites of tobacco companies based in major English speak- evidence provided by supporters of workplace smoking ing countries, including the Philip Morris "Options' websi[e, regulations." Similar findings were observed in relation to [he www.pmoptions.com .The researchers also conducted an Californian Envvonmen[al Protection Agency's risk assess- Internet search with the Google search engine www.google- ment of secondhand smoke," and in Maryland and Washing- .com ,using [he terms "smok' bans" and "restaurants' or ton heazings on proposed clean indoor air regulations." Bero "bars", limited by [he terms "economic impact" or "study". www. tobaccoconhol.com 14 S<ollo, Lal, Hyland, el al Study selatlion attempts were made to determine whether authors or Studies included measured changes in sales, employment, sponsoring agencies had ever received financial support from numbers oT establishments, bankruptcy data, public reports of a tobacco company or affiliated group. Searches were intentions about or recent changes in patronage, spending or undertaken of previously secret tobacco industry documents time spent dining, proprietors predictions or perceptions of made available as part of settlement agreements between sales changes and costs and estimated numbers of tourists. tobacco companies the US attorneys general" " (accessible Studies were excluded where these made no explicit or through www.tobaccoarchives.com implicit attempt to quantify the economic hnpact of smoking restrictions. Studies assessing opinions about smoke-free Analyses policy were included where the study included a question We used contingency tables and odds ratios to compare stud- asking specifically whether people would attend venues more ies concluding a negative impact with those not concluding a or less frequently were such policies to be introduced. negative impact. Ninety three per cent of the studies located (97/104) me[ [he selection criteria as determined by consensus between multiple reviewers. DATA SYNTHESIS A total of 97 reports were located. Studies covered DATA EXTRACTION numerous local jurisdictions in 31 state or provincial areas, in Findings and characteristics of studies (apart from funding eight countries. Thirty four of the studies examined the source) were classified independently by two researchers in impact of smoke-free policies for drinking establishments, most cases several months before the stazt of the study (MS and 90 the impact of smoke-free policies for restaurants. lWo and AL). Both assessors had tertiary quafifications in studies examined [he impact for recreational venues. One behavioural silence and economics. A thud assessor, (I.H) a examined the impact of smoke-free restaurants on hotels and postgraduate psychology student blind to both the objective of another on overall tourism. the present study and to the funding source, also classified each study. Sfudy qualiy We used Siegel's criteria" to judge study quality: use of Of the 97 studies, 38% (37/97) used objective outcome meas- objective data (for example, tax receipts or employment ores and 25% (24/97) were peer reviewed.lWen[y one studies statistics inclusion of all data points after the law was imple- met Siegel's four criteria for methodological quality, represent- mented and several years before; use of regression or other ing 57% (21/37) of the studies that used objective measures. statistical methods that control for secular [rends and random By contrast, 98% (59/60) o[ the studies using only subjective fluctuation in [he data; and appropriate control for overall outcome measures met none of the criteria. The odds of peer economic trend. The more criteria a particular study met, the reviewed articles meeting all four criteria was 5.33 times that more certain one can be about the validity of results. of non-peer reviewed studies (95%confidence interval (CI) 1.9 An outcome measure was deemed "objective" it it was [0 15.1; p = 0.002). based on data collected routinely by an independent agency covering [he periods both before and after [he smoke-free Funding source policy was in form. Objective measures included: sales figures OT [he 97 studies, 32% (31) were funded either by [he provided for [he purposes of taxation assessment; employ- tobacco industry or a group known [o have received funding meat figures provided [o government agencies generally for from a tobacco company or tobacco industry insurance purposes; and numbers of new or existing ally."'°""""'°"°°o-^'°"°"°""Four studies did no[disdose establishments based on business permit applications or reg- funding source, but were conducted by consultants or organi- istrations to the government agency that issues such permits, sations known to have a connection with the tobacco and bankruptcy data. industry.1° For three of these studies, there was strong Unverifiable predictions of future changes or estimates oT evidence of collaboration between tobacco companies and the recent changes in patronage or spending were deemed study authors."'°" "subjective". Subjective measures included anecdotal reports The two studies by consultants where close ties with [he and self report data collected in polls of, or interviews with, tobacco industry had been established and those funded by patrons or owners of restaurants, bars or simflaz businesses, organisations known to have received funding from [he conducted either before or after the policy was put in place. tobacco industry were included with the tobacco industry Another indicator of the quality of a study is whether it has funded studies, yielding 31 tobacco industry supported stud- been subject [o peer review. The scientific quality of original ies. The funding source was treated as missing for six research on secondhand smoke published in peer reviewed s[udies1°"^'0'-106 because no data on funding source could be journals is superior [o [ha[ in non-peer reviewed publications located. in terms of study design, reporting, and evaluation." A study The other 60 studies""""° °-i 140.10"°"" were was deemed to have been peer reviewed if it was an azticle funded either by government, health related organisations or published in an academic journal. independent market research organisations. Studies were classified as indicating or not indicating a All the studies used in this analysis is summarised in tables negative economic effect based on their stated conclusions 1 and 2. about the impact or potential impact on employment or prof- itability of the various sections of the hospitality industry a[ Study quality and funding issue. All three raters agreed on the conclusions in al] but one The methodological quality of the industry sponsored studies of [he 97 studies. The Maso[[i s[udy1° was classified as was significantly lower (p < 0.001) than the non-industry negative, as per the conclusion of two of [he three raters. 7Wo studies ([able 3). Only one of [he 31 tobacco industry s[udies1°S° did no[ draw conclusions; we included [hem in [he supported studies (3%) has been published in a peer reviewed analysis based on [heir face-value findings. journal compared [0 38% (23/60) of [he non-industry funded Funding sources for each paper were noted after completion studies. None (0/31) of [he tobacco industry supported studies of all [he other classification tasks. Funding was determined met all of Siege]'s" four methodological quality criteria. from acknowledgments in the studies (obscured on copies of Indeed, 84% (26/31) of the industry supported studies me[ papers classified by [he third rater). When the source was not none of the criteria. By contrast, 35% (21/60) of the studies no[ clearly disclosed, authors were contacted where possible and supported by the tobacco industry met all o£ Siegel's" vrww.tobaccoconho.cam Economic effects of smoke-free policy in hospitaliy industy 15 Ta61e 1 Sh1dNs wing objecliw msasuros b assess acorwmie impact of smoke6ee policies in the hospitaliy industry ~i.~a' ` r MYBS~XyraRert ssudbs tundNtroas esw[waeeiMlben eso soBwNsdeeey TaaoMe sabe Badosds and fbpo (199 ~y ~py~ fT9i}J00• CoIlFomla Sbss Board d rsaipb Ba9sals add9aea ~ Ond CFONi(T49T)°; E4uo6mHo"(1998p tNY of Or.wr 11999)°; ~ saJ 1}l1tJ°: BouWor 11996 ; FMkhw alsab aall h °~a6sds aaa (s99fp°; Ilyq~r nano ; ~dB./a~~tln+t>i~°;~adee ana tta.wdd~•19!{01; `tMhndNd Il Nd 1994x'; PoaBc 1y~~ ?npo and Bat dos (19P~' 9dase sea 1,99pf1 ~"e r2oalr; Te/lar w 1109311°: WaMeMld N d (2oa11's Sobs daa odor `[Aewr N d (1999)i° Empkryment `HIIdNd2t101i0; `Nyland and Cuaalws(19991°; lerla Hy1ad~~ 12001)'0; Flytsud Mdl9BOO1°; Booms NumMr d `end aW Csaaairgs 11999x' esb6llshmenh BonYrvpky daq Baums and Mokcmwn0' sswBSe to ssMd iwsdw b axlbseasn Sdr dab olhsr Pubco 2002" shldbe aasaluaod bM esgonbelMns or ossweeMe ~ Brno b Nu bbaem induty as°und eu dme d Nr elanly Taspbb nabs `Lilby M d 119961" rxotpb •MOSOei Nal (1991)'Of ShsdiN tsnsded by teharas sompade a• LsdsMry 90ou9e OOOPPa~ M Na sobaae ialasary Tazabb sale 'lovendiol N al raplpb 11990)'• Saba data odor Applied econanics 11996x' Empbrmenl `Lilby K d law4 119991'0' •lilby M al (1996)00 NumMr d 'Lilby N d 1999^ eaiobllahmeNs Bold type - psa wNOVnd. •Use dlwYa iaNw Ihm wnenuow dda bdom sal aFx Nw innoducNOn d polWes. TOnly rook eadsna dagwdbn vdlh dr sObarco induahy. criteria' 010 ° ° (p < 0.001). Eleven of these non- razely included an objective measure and were almost never industry funded studies have been published in peer reviewed peer reviewed. journals. In studies concluding a negative impact, the odds of using only a subjective measure was 4.0 times (95% CI l.4 [0 9.9; Study quality and eonClusion p = 0.007) and [he odds of being peer reviewed was 20 times Table 4 sets out [he fmdings of [hose studies meeting each of (95%CI 2.6 [o 16b.7; p = 0.004) [hat of studies concluding no various indicators of high quality: using objective outcome such negative impact Qable b) measures; meeting [his and Siege]'s other three criteria for quality; being funded by a source clearly independent of the Funding source and eontlusion tobacco industry; and being peer reviewed. There was a significant association of tobacco industry None of the 21 studies [ha[ met all four of Siegel's" quality support with negative conclusions of the study (p < 0.001) criteria reported a negative impact (table 4).In fact, four of the (table 6). Ninety four per cent (29/31) of [he tobacco industry studies report a positive impact on taxable sales receipts of supported studies concluded that there was or would be a restaurants, bars, hotels, or tourism.""' ° negative economic impact of implementing a smoke-free Only a handful of studies based on objective data conclude policy. The odds ratio for a negative conclusion associated with a negative impact. None of these meets more than one of tobacco industry support was infinite because none of the 60 Siegel's other three criteria for methodological quality. Only non-industry funded studies concluded a negative economic one peer reviewed study concluded a negative impact This impact. study relied on subjective data and was funded by a tobacco company. DISCUSSION Table 5, conversely, shows, for those studies concluding Lower quality studies were much more likely to conclude negative impact, whether each of the various quality criteria smoke-free regulations adversely impact the hospitality was met. Once again, studies concluding a negative impact industry, and weaker studies were much more likely to be www.tobaccoconnol. tom 16 5<0ll0, tot, Hyland, et al Table 2 Studies using subjective measures to assess the economic impact of smoke-free policies in the hospitaliy industry SNdiN tlrrlded tram NVreN Miler ItWn Kle Mlsoae IiMUelry Public seE In~1lKplr«aaNl hom Isa, 120tw1°~BieMrond patronage ofrosWwmh(ban ~ e ]493 iFT'CT~omm~^N d Il inn tmo Reseorch 120021f°' Omaer M d (1499 ~ ~ r m. Field RN I ; MYlmd and 'stun ripe 11999) • tom 1 S 2001 a; Mil and Kriwn (20041 ~ Miller an Kriwn 11002 Shapiro 2 1 Stydsg 12001 i Waken d 9 sa Proprieor pmdk4ws/ perupfirsm d sales II rM~tg~„1~I~.rMn1~ mid 09bKe chorgN 200 DrMSerM ~EdsWrdt 120Wt°;Huron count' Hwah Wrdt 1999'1'• llylalld'drsik Cuml9 /949}sOt; Jorlp etd~149W1(s's~ Morkham and twg 'i s~ wrty eFd l~r! ; #!eba: end Eewom (194J1i0; SdaNe (T4eb~'i (19966yy''; iM Corddance 90ard dCmkdu ] ,YorWdn Ads {20011'm Propriekr pwdkMons/prrcepHOm M war Mmiwa and QuleXe gOT~s; TFIe Conference Boord d Canodo (I99a^; Dos - . Cpmy CHIR (2001Ins Eslimoted numisas d ovNSeas Wibrs H Nand Masiill } sr StudiN ter svhkh (unr6ne puma's unknown Poprisor pmdlcBOn/ percepdons d wIN Economim Adel Grou 1998 'a. charges ~ 0 200 ' inn 2 1 'a SludiN cersdueid Iq{eslpmimKeru «canwMaMe wKh abort Rrlktte Mm bhoeID indusry around IIN Kale e(IM shills, Proprieb pre~cEOrwf'pipwpllpu d salty MmoRi N d (19911'Ot C I 101; Charkon Research (19941° changN SwdiN (ended hY lehacm cempaniN or klduelry grouq wppoeNd IfY the tebesw IMusley Pubik sell ~e~pred insdions «adud AusooE 1200416; Deaimo nle~9h 1198a1n fobnzio N d t1995J'a; KPMG B«Nh pahonogs of ndewanis/hon 119971a~ A1oA~ 119991°• Naslond Rubursull AeeoelMbn 11993 mN 19 ~ Public self reported spsrding/nme spNS Fobdzio Na1119951jp1; Monin Asaocialas 1999 a Proprieor predktlons/ peruplions d sales Adwn a MarkeRn Info 199 ° change Appi cawmiu 6 CCG~ 1995'a; h nResear~ Consul 1 nham Nsd MoA 0~"' dzk d " >++sryls' GambN (19911''; KPMG f'aM Morwtrk 114981'0; KPMG 120011i1; Alorbw 119941°~ Modow 996 MoewoDBlw Morkd~ararch ll''0° ~~gga~n. Aka Walws~g9w - 95~" Res[ ( EersiMn Grouo f 19971° Proprietor esfimaM d impaG w Ad~Inbgt A~rkNin~InLo+19971r0; empbymsm Applied Ecasolnfu (14961 s~rizio N d 119961'x. Ma 998 a; Pdce Wa tine RopN Starch I~-~•,; C~Ms I~bsr a(n Reswrc Consvhanh 1999 • omlmn Grouo 1199T°' Proprietor predkfiona/perception of wsl Sollars sl al 119991°0 Bold tYPe . r revlswd; undNline ~ study based w erhmabs of pradicled dwngN ratlwr Than astimotes of actual changN. Nar a random wrwy. iOnly weak NWence d wnnx4w wi4l the bbacco irrclNlry. ]Conhd far economic hands. funded by the tobacco industry. In addition, the industry passive smoking was not harmful to health; 74% of these studies were less likely to be published in the peer reviewed reviews were written by authors with tobacco industry affilia- literature. Almost all (94%) of industry supported studies, [ions. Among reviews written by individuals with no industry compared [o none of the studies funded by sources other than affiliations, only 13% (10/75) reached [he conclusion that pas- the tobacco industry, claimed a negative economic impact. live smoking is not harmful to health. In logistic regression These results are consistent with a similar linkage between analyses controlling for article quality, peer review status, arti- tobacco industry funding and s onclusions in reviews of [he cle topic, and publication year, the only factor associated with effects of secondhand smoke."° Barnes and Bero"` identified concluding that passive smoking is not harmful was whether 106 reviews of the relation between secondhand smoke and an author was affiliated with the tobacco industry (odds ratio disease. Thirty seven per cent of these reviews concluded [ha[ 88.4; p < 0.001). www, tobaccoconhol. con Economic effech of smokafrea polity in hospitmlily industry 17 Table a Qualiy of >tidisa wpporbd by the Aobaao i~usay compared wish thoas that are not ~x~ 1 ~~r r ' ~ ~ ~ ~a t,N •:.~z- 4maYlarsisa Il%MN'q°" eswrrw 1s t1:7rsriA:4J O.aab oW6enie liwmNw wla O%lfl/31t 35%Q71/bOJsmammr Ukdlr aiMim mrwwauaw Su6aCbpw rNiwv 3%tl/31(ei 3@%(79/6~saau.vv 183 (2.4b142.~ O.OQSS wawnw]raw wl lu CI, m4tdmcs. inbrvo6. Table 4 findings of higher quakky ahsdiea V ~ % y{v` 3 -114~:~'1~~ ~.u tr x _ =t~ a~rr 9ly~laYls. ll%~yl~lt t00ltCf l.Ye ]amww.M id.S3 a.009....~ Indudine on o6pcrtw maa]ua 19%17/371']av.1 Bi%i30/37taawauwa 7.64 0.006 i~ .n faldad sebp c4ody O%i0/40i 100%i60fb~aa+a mae t12.~ 0.000 InduM17 (n.4i01k. blsa[m m.aw«Im.lm lw-Im,m Pwrrw~bNgd tb.24i 4%tl/24iw 96%~3~aaau.aww 14.09 0.000 awwwa]w wl Ile Table _5 Auatiy of Ahldiea amalq d7oas dlat rep4rkll'o negativ4Jmpaat "hie„ sy jt A" 1@., MrAnY dllosr Shed allarb .~0%(0/39t 100%W.]`/33(mmba IS.I3 .,.0.000 a+l q.awlm.la l4 /~i~vp~__. ~_a ~lwww 7.6t 0.006- Im+gm tand~d ~a~yvycl~ap~~~y,~. a%tD/3$l ta0%f3u'/3!'~mwb]I t~.qp 0.000 efbt]amrs Ydua®r W]RwMla-Iq IN 1'Nr]eVlan6 3%il/J$M 9771 t.~13~mbaal 14.04 x.000 wwa a].lalu Table 6 Odda «ei~ fa indiegrora of bwer quaky among sNdiea wilh negative rnrleluatons compared ro spldNa Iwt wncludMg a negative (mpad Indadlne ~Y ~I~N,~ .~t1mw ..S.~fg?~+~wwwmws~~4.0 (t.1b 9.9) 0.007 aubom~msmwns aww w.w lu wwlgagla.]m na Fundd by dw lobacn 94%i29/31~]]]' d%(2f62i'°• InAnN~ 0.001 Indubry a o e~i aurporbd vel was Im Iv.a 11. 67 6q bkamo GMUSky IVot baine pw minsgd7 97%(34/331 1 fj1 63%(39/62(10 a ]a'1O 20 i2A b 166.7) O.OOd bYaV]A IO41a 1N ]]L ]a11 Y]M]atl V-]ONN wA la Ia Ia-11] vnvwlobaccoconhol.com 18 Scollo, Lal, Hyland, of al The possibility of publication bias always exists. It is pos5i- a Mangudan C, Bero a. Lessons learned ham the tobacco indushiei ble that studies by those sympathetic to public health goals e8ons ro present Iha passage of o work place smoking regulation. Aml that detect a negative impact would be less likel to be Public Heahh 2000;90:1926-30. Y 9 Dsmdew J, Bialous 5, Glantz 5. Tobacco industry manipulafion of the submitted for publication. On the other hand those funded by hospiroliy industry ro maintain smoking in public pl«as. Tob«co the tobacco industry would be similazly unlikely to release Connol jin pros). studies detecting no negative impact. We have made eve 10 AlbeHa Tobacco Control Centro. The economic impact o/smokelree ry reaburant byksws. Edmonton: Alberta Tobacco Conhol Cenne, 1999. effort to identlfy all studies done on the effects of smoke-free l l Myqxamsm 5, Chadesworlh A, Glantz 5. Print media coverage of laws and regulations on the hospitality industry. The fact that California's smokehae bar knv. Tobacco Control 2001;70 154-60. the tobacco industry has a strong motivation to publicise all 12 $ieRN M. Economic imp«t of 100%smokehee resburanl ordinances. In: Smoking and reskuronla: a guide kr poliq makers. Berkeley: UC negative studies adds to our confidence [ha[ we have no[ Berkeley/UCSF Pravenro8ve Medicine Residency Program; American missed a substantial number of studies concluding a negative Heart Ass«iotion, Califonio A(Fliob; Alameda Couny Health Care economic Impact. Services Agenry, Tobacco Conhol Program; 1992. 13 McMini T, Bero L. Paliqq maksri perspectives on tobacco control Siegel's criteria are a valuable tool for assessing the quality adv«alas' ruler 1^ ragulaton development. Tobacco Conhol of studies on the economic impact of smoke-free policies m 2001;1o21B-2d. 14 Scholland M, Bxo L Evaluating public commentary an uienfific the hospitality industry. Our findings suggest [ha[ pohcymak- widence submiBed in the dwslopment of risk assessment Risk Analysis ens can make a quick preliminary assessment of study quality 2002;z2:131-40. by asking three questions; 15 Mro L, Montini T, Bryan-Jones K, e/ al. Science in regularory polity making: case sNdies in the development of workplace smoking (I) Was the study funded by a source deazly independent of resnicriona. Tobacco Conno12001;1a:329-36. [he tobacco industry? 16 Scollo M, Lal A. Summary of studies assessing the economic impact of smokehae policies in Rio hwpitaliy indeshy - i«ludas studies produced (2) Did [he study objectively measure what actually happened, ro 31 August 2002. M•16ourne: VicHwlRi Cenne for Tobacco Connol; or was it based on subjective predictions or assessments? 2002. http://www.vck.xg.aa/publ/reports/ hospitaliy_popsr_sommary.pdF. (3) Was it published in a peer reviewed journal? 17 Boma DE, Bxo IA. ScisnflHc qualiy of original research articles on Of the 35 studies on this topic published chat concluded a anvironmenbl bbacco smoke. ioba«o control 1997;6:19-26. 18 Moselli L, Creecas P. The sNech of a ban on smoking in public places in negative impact, none have been funded by a source clearly San Luir o6lspo eaRfomia; t99t. independent of the tobacco industry, and none have both used 19 Auspell. Philip Morris Public Opinion Survey; 2000 January. an objective measure and been peer reviewed. In fact, 80% of 20 De<ima Research. F«us group report on project visa for Imperial Tobacco. 1988. Access dole: 24 September 2001. URL: these studies passed none of these basic tests of quality. With http://www.robaccod«umsnh.«g/ all 21 of the well designed studies finding that smoke-free view.cfm?CiNO-57726343GeRistArroyldx-.185howlmage s-yes. restaurant and bar laws had no negative impact on revenue or 21 Dyer C. ConBdanfial Tbacco d«umenh Doter public domain. BG11 1998;316:1186. jobs, policymaker5 can act to protect workers and patrons 22 Malone RE, Balbach ED. Tobacco industry d«umanrs: treasure Hove or from the toxins in secondhand smoke confident in rejecting quogmira7 Tobacco Conho12000;9:334-8. prediC[i0n5 that [here will an adverse economic impact. 23 K?MG Baronb Group 4C. The expxbd economic impact on Spain o! a ban in rmokingg in reabvranh. Washington; 1997 April, 24 Mcghee 5, Hedley A, Lom T. Does dte govemmenys proposal ro creaM smokabee cabrin lvciliees in matauronh, mfes, burs and korookea ACKNOWLEDGEMENTS influence the intentions of bviah b visit Hong Kong and b pohoniae ITrankS to Dr Mohalmnad Siahputh from [he VlcHeal[h Centre for cabring venom. Hong Kong HevlRs Service Research Group, Tobacco Control For advitt on statistical analysis, and Louisa Hoey for Daparnnant of Communiy Medicine, Univarsiy of Hong Kong, March assistance with classification of studies. 2002. 25 Hyland A. Before and alter smokelroe regulations in Haw bxable sale The VicHeal[h Crntre for Tobacco Control is funded by the Victorian 6a^ ev^^e arrd drinking plvcm in New York $bN. New York- Roswell Health Promotion Foundation [o conduct economiq legal, and sodal Pork Cancer IndiNb, Juns 2002. 26 Mlllx C, Kriwn 5. Smokahee dining in South Australia: surveys of research in tobacco control. Dr Hyland's work was supported by [he Communiy ottiNdes and ppr«licas vfM d and 18 months. In: Tobacco Roswell Park Cancer Institute NCI-funded Cancer Center Support Conhol researzh and evaluo8on report I99B-1001. Adelaide: Tobacco Grant, CA16056.26 as a member of [he Hiomathema[icsBiostatistics Conhol Research [valuation Unit, 2002. Core Resource. Dr Glan[z's wo[k was supported by US Natio_na_I Ca_n- 27 Pubco. OBicial figures ham Onbrio hrewxs confirm disastrous effects of cer Institute grant CA-61021. smorteingtian. Puh-6nd Sat COaRttan-of Ontario. 2002. Access date:2tr Augus12002. URL http://wwwsmokeinottawa.cam/html/index.hhnl. 28 Milbr C, Krivan 5. Communiy support for smoking bans in bar and " • • • • • • • • • ~ ~ ~ ~ ~ gaming venues In SouRs Ausnalio.ln: Tobatm Conhol research and Authors' afRliOfionS evvlua8on mport 1998.2(NJI. Adelaide: Tobacco Conhol Research M Scollo, A WI, VicHealth Cenne for Tobacco Conhol, The Concar Evaluation Unit, 2002. Council, Victoria, Melbourne, Victoria, Aushalia 29 Synng W. A sNdy ollha Foil Wayne (INJ resburant smoking ban: has it Impacted the resburonl buslnessB Indianapolis: Hudson nstiNta, 2001. A Hyland, Roswell Park Concar InsfiNb, Deparhnent of Cancer 30 HaysbB 1, Huang P. Impact of dean indoor air ordinances on reabvmnl Prevention, Epidemiology, arsd Biosrotisfics, Buf(vlo, New York, USA revenues in four Texas cifies. Bureau of Disease, Injury and Tobacco 5 Glantz, InsRNb of Health Polity SNdios and Cardiovascular Research PrwenRon, Texas Department of Health, 2000. Institute, Univarsiy of California, San Francisco, California, USA 31 KPMG. Proposed smoking bon: impvch on Hong Kong hospifaliy businevea. Hong Kong Colxing Indmhy Association, September 2001. REFERENCES 32 Dacima Research Inc. Public support grows for ciy's smoking bylaw. Decima Reswr<h Inc. 2002. Accea dale: 26 August 2002 URL 1 Chapman 5, Boland R, Scollo M, of al. The impact of smokefree hop://www.decima.ca/research/V/hahNew/index.aspllD-57. workplaces on declining cigarette conwmplion in Aushalia and the 33 Bourns B, Malwmson A. Economic impact analysis of the non-smoking United $rohs. Am J Public Health 199989:1018-23. bylaw on Rte hesp'tvliy indeshy in OMOWa. KPMG, 2001. 2 Fichlanbarp CM, Glantz SA. Eff«f o(amoke-hea workplaces on 34 Burbach W, Pope G. Pmliminarryy analysis o{the xonomic impact o/ smoking behaviour: systematic rwisw. BMI2002;325:188. Brooklirtes smoking ban. Massathusaes: Health Economics Research Inc, 3 Bumf D, Shanks T, Major J, of al. Restritlions on smoking in the 1995. workpl«e. In: U$ Department of H.olth and Human Serviua, Public 35 Badoseh W, Pope G. The economic eff«t of smokafrae restaurant Heahh Service, NaRonvl InaNNta of Hwlth, National Cancer InstiNle, polkies on restaurant bvainesses in Maswchusahs.l Public Heahlr eds. Populvnon bused smoking caxalion, Bethseda; 2000. Manag Pr«r 1999;5(1 j:53~2. d Samuels B, Glantz 5. The politics of I«ol bbacco Conhol. LAMA 36 Bialoua 5, Glantz 5. Tob«co Connol in Arizona. Institute br HealRs 7991,266:2110.7. Polity SNdies, Universiy of California. 1997. Access date: B August 5 Glantz S, Charlesworth A. Tourism and hotel rwenues before and aher 2001. URL: www.library.wsF.edu/bbacco/az. passage of smokefree restaurant ordinances. LAMA 1999;287:191 I- 37 California Slob Board of Equrslizolion. Report, 13 October 1998. 18. 38 City of Beuldo Colerodo. Tax receipt data, 1996. 6 Magzoman 5, Glantz 5. the new baMeground: ColiFornia's experience 39 Droner J, Beim 5, Lichlenslain E, el al. Muhiple impach of a bar with smokatres hors. Aml Public Health 2001;91:245-52. smoking prohibition ordinance in Corvallis, Oregon. Pacific Research 7 Ritch WA, Bagay'dE. Sponge bedfellows: the history of collaboration InstiNts, 1999. beMean Rse Massachuselh Restaurant Ass«iafion and Rse bbacco 40 Draux L Clearing Rie air. Wiuonsin: Tobacco-Frae Wisconsin industry. Am 1 Public HeaNh 2001;91:598-603. Coalifian, 1999. www. tobaccoconhol.<am Economic affects of amokafree polity in hospitality indushy 19 41 Fiekhe J. An analysis clads bx recsiph Irom rebumnb wills ban 7d Medrhom V, Tug R. Reacfiom and anitudu ro hwlth smoking in and tree abrading bon in Chiro, Califomio 19951997. Colifomio a«losed pksce~ regulations 1999. 2001. Awx dab: 3 Augw12001. Dsporhnml d Heats Service, Toheeeo Conhl $eaion, 1998. URL: htlp~//«wh.org/Iibrory_h_sN.hlm. d2 Glenh S, $milli L. Tlr effect of ordirwnces requiring smokshas 75 5daeu 1, Ecknm M. EK.ch d a ciy ordinance regulating smoking in reburonb on ruburonl wle. Am l Public HeaBh 1994;84:1081-5. rerouronh and rsbil sbre. J Communiy Hsahh 1993;18:175-82. 43 Glanh 5. EKsa d amakehee bar low on bar newness in Califonia. 76 Sciacea J. A mordafay smokingg hen in reburanla: conurns vxms Tobacro Conho12000;9:11 I-2. expxisnus. J CommunNy HevMs 1996;21:133-150. 44 Glanh 5, Smith L. The effect of ordirwncu requiring smokshea 77 5lanwick R, Thomson M, Swsrlwne P, er al. TM response of Winnipeg resburanh and ban on rswnue: a fellow up. Am J Public Health retail stops and rsdauranh ro o 6yksw regulating smoking in public 1997;87:1687-93. pl«es. Cant Public Heahli 1988;79:226-30. 45 Geldabin A, Sobel R. Environmenbl bba«a smoke rogulalions have 78 Pony 1, Lmperbn H, Flanagan T, eL al. An swluation of the impod of not hurt ruburanl solo In North Carolirw. North Carolina MedJ the inhoduclion of'narromoking' arws on hods and cusromar 1998;59:284-7. saliaLocfion in 11 public houws in SlaRordahiro. Tobacco Control 46 Huanyy P, Tobias 5, Kohut 5, of d. Aassssment d the impact of a 100% 2001;10:199-200. smokshes ordinances on rodauronl aa4s -Wet Lake Hills, Texw, 79 Cord~ruru Beard of Canada. The aconomirs d smokelrea 19921994. MMWR Mort Morbl Wkly Rep 1995;dd:370-2. roabumnh. Toronro, 1996. 47 Hyland 0. Cumminrggs K. Redaurant am loyment before and afbr Ths g0 Adwntage Marketing Infemwlien. Rhode Islander's aniNde Towards New York Ciy $mdshss Air Act. J Public Health Manag Prod smoking in rosburonh and hotels. 1997. Accus date: 13 September 1999;5111:22-7. 2001. URL: ht1p://www.pmoptions.rom/images/Rhodehland.pdF. 48 Hyland A, Cummings K, Navenherg E. Analysis of roxable sales g 1 Applied Ecenemin. Economic impact of Nss Ciy of Mesa smokehes recaiph: woe New York Cilkh Smokekea Air Aa bad for busines7 ordinanu: working paper I . 1996. Accau dab: 8 August 2001. URL 1 Pvblic Heath Ma Pron1999;5~i ad-21. h8p://www.pmoplione.can/imago/EcononmidmpaaMewl.pelf. 49 Hyland 0. Vana C, Cummlrga K, of aI. Ths sHect of the Clwn Air Act of B2 Chamberlain R•aearch Ceruullanla. Smokingg issues in Wisconsin. Erie County, New Vark on reburonl employment. 1 Pvblk Heohh Manog 1998. Accea dab: 13 Sepbmbm 2001. URL: hnp:// Prod 2000;6161:76-85. www.pmopNons.can/imagsa/Wisconsin.pdf. 50 Hyland 0. Tuk J. Resburonl employment boom in New Vak Ciy. 83 CharROn Research Group. P«iBc Dining Car Resburonl and Southern Tobacco Conhd 2001;10:199-200. California 8winsu Ass«iaNon Survey. Son Francisco, 1994. 51 Monnq N, Sherwood D, Sbbbkbins W. The imp«f o! bb«w 84 Dunham J, Marbw M. Smoking laws and Nuir di8erenNal ef(ech on control ordina«e u rssbunnl nvsnwa in Cdlhxnia. Claremont, mburanh, Mn and bwms. Conbmporory Economic Polity California: TM Claremont InsHbb for Eronamic Polity Studio, 1994. 2000;18:326-33. 52 Pacific Atglylira I«. The arorwmic imp«b d IM propwd amsndmsnL 85 Gpmbw P. Economk imp«b olsmoking ban in Bellflower California. ro tM ETS regulatOn. 2001. A«eu dab: 3 Augud 2 1. URL: Califonia Busineu and Reburont Alliance (CBRA~, 1991. hnp://www.wakadebc.com/Prioriy/smoke/pdh/eroimp«Lpdf. Bb KPMG PM Morwkk W. Ths imp«f of Cdifomia's smoking bon on 53 Pope G, Burbach W. Egscl of hml amoblrsa rosburanf pdkW on 6aro, broms and nlPhl clubs: A avrvey of owner and manager. rebumnl rownw in MauachusMla. Center far HwNh Eronomiw Wadsingbn DC: KP 1998. Reearch, April 1997. g7 Monew M. M xammk amlysis of the Mvirn smoking bon: widenro 54 5<iacm J, RatliR M. Prohibiting smoking in nslaurana: effsas on from pahons and owner d bvaiusses California Polytechnic Srob rmburant aolu. AmJ Fleahh Prano1199B;17:176-g4. Univeniy, 1999. 55 Tayler CeruulNnq Group. The San Luir Obiapa smoking ondiwnu: s g8 Manew M. The economic elfecl of smoking Iowa en ban and bwrw. study of tiro eronomk imp«h of San Juis O6hpo rebumnb and ban. California Poyyechnic Uniwniy, 1998. San Luis Obispo, Califomia, 1993. B9 Manin Auociobs. Analysis of paawngsr axpsndiNre profile at 56 WakKeld M, Siahpush M, Scdlo M, of al. Ths effect of a amoks(ree Phoenix Sky Harbour InMnoKOnol Airport inconceuion survey. 1999. law on monHJy rsWUranl reroil Nrnowr in SouNi Awholia. Awl NZJ A«eu dab: 13 Sepbmber 2001. URL: hNp://www.pmoptions.com/ Public HsaM 2002¢6:375-82. image/PHOENIX.pd. 57 Applied Economic. Economic imp«I of the CIy of Mew smokeJrss 90 Mogn-Dixon MorkN RWarch. Maryland smoking regulation survey. ordinanu: working paper 2.1996. Accsu dale: 9 Augw12001. URL 1996. Acceu dab: 9 Auge12001. URL hnp://www.pmoptims.com/ hnp://www.pmopllons.com/imago/EcanonmiclmpaclMssa2.pdf. images/Marylard.pol. 58 Lawnthel L Henwath. Preliminwrryy analysis of Nis impaa of the 91 No erwl Mabunnl Asaedatlon. Smoking in reslauranls a consumer proposed Los A~ebs ban on smokier in rWOUrona. Lm Angale, aMbde survey. Wwhingbn DC, 1993. 1990. 92 Prize WrneHwun LL?. New Yod Ciy rerouranf survey. 1995. Accus 59 Liliq W, DeFronco L. 7M impod o/amoking rehi<fius on Nw bar and dab: 8 Augud 2001. URL hlry://www.pmoptions.com/image/ ravem indwhy in CaliWmia. Wadsingron: InContun I«, 26 Ocbbar NewVark.pdf. 1999. 93 Pdn Walerheuw. PobnNal effech of a smoking ban in the State of 60 Lilley W, DaFro«o L Reburont jobs in New Vork Ciy, 1993 through California. 1993. Accex doro: 9 August 2001. URL: hnp:// fiat quarter 1996, and the rerouront smoking Mn. 1996. A¢eu dab: www.pnropfions.can/Images/PofantiolEconomkEff«h.pdf. 13 Sspbmbsr 2001. URL: hNp://www.pmopfiom.rom/Images/ 9d Regr Shrdr. Nafiorwl survy of rssrouranh and bars/taverns: smoking Rasrourantloha.pdf. polity and regulaHOna. 1996. Accau dab: 9 August 2001. URL: 61 Liliq W, DsFronco L. Maas«huaeHS nalaumnt smoking ban 23 hhMp://www.pmapHom.com/imago/Nafional5urvey.pol. <i1lea/lava: impM on rulaaaM jots 19931995. 1996. A«ee dais: 95 50Ren D, krgram J• Ewnoesk imp«t of Nu reslauranl smoking ban in 13 Sepbmher 2001. URL: hlry://www.pmopNOm.rom/Image/ N. ciy d Manaehusena, May 5 1999. MauRerouronf. 96 TM Cniq Grwp Ire. Wet Vlglnia rasrourants and taverw fear 62 Auggual K, Br L. Support kx amoksJrss ban pnsws skugw in smoking bum wuW bun businsn, 1998. Califonia: Callfomla D• rmMt of HwINs $srvkes, 2000. 97 Ths Epppein Group 1997 stabwids hwpibliy industry Mnchmark 63 Biener L, Flhhggsrold G. Smoky ban and resburants: who maids them pelt. 1997. Acuu dab: 13 Ss~ ytimber 2001. URL hnp:// and why.1 PuWk Heath Manag Pr«l 1999;511 :74-8, www.pmopliow.com/imppu/Tezm.pelf. 64 Biene L, Siegel M. Behavior inbnNons o(Nro pu~li< a8sr bum in 98 BatHOaeh W, Pope G. TM effed of smokingg rehidions on resrourant rnsrouranh and bon. Am J Public F1sa1Hr 1997;87:2062-4. husines in Maswchuwm, 1992-1998. Tobacro Conrol 2002;11 ~suppl 65 Conun D, young C, Enz C. SMuW NYC1 mburobun lighbn ap7 I11:ii38-d2. EHeds of the dtJ~s Smokohes Air Aa. Cornell Hots Reburonl Adm O 99 Pubu. $epbmbw Pubco survy wnlirms awsra economic impact, IIoh 1996;37~2~:25-33. loess in ih member ban and pubs. 2001 Acwu date: 28 Nwambar 66 Cnmiws Oulletb P. Achsal and pxuivsd impacts of roMcco 2001. URL: hMp://www.ppuubcoalition.com/html/. ragualiu on raslaumnh and Brms. Tobacco Corbel 2001;10:33-7. 100 Allen K, Markham V. Pubic opiniom and oMludas rowords creafing 67 Decima Research. Onawa residmh hack smoking ban by hvo b one. xnokelroe ban in Weabm Aushalia. ACOSH, November 2001. 2001. Aaeu dab: 19 Ocrober 2001. URL help://www.decimo.co/ 101 Wakefield M, Roberts L, Miller C. PsrcepNons of the e8ecr of an rewrch/WhahNew/011017.asp. impending roelaumnf smoking ban on diningoul e.periancs. Prw Med 68 Edwards R. New Study: 76% of Ths NorM East Iwspitdipp bade bock 1999;39:53-0. smoke free crew 8 own 90% of publiwm recommend olhx pubs M 102 Verkshin Ash. Popularly and Impact on hods of smokehes one. 2000. Aaeu data: 8 August 2001. URL: htip://www.aah.org.uk/ accommodation In Ths hwpibliy bade in Yorkshire; 2001. brat/prey/000720.hhnl. 103 CCG ConuNLnq Group OmBed. Tha hwpiroliy sedan and a 69 Field Reawwh CerpoMion. A survey of Calibrnio adulh age 21 w Vanuuvx amokirrgg bm, 1995. Accau dab: 3 December 2001. URL: cider about smokinrgg policies and smoks{ree bars. California: California hitp://193.78.190.200/evhosp/rsslaur/roamur.hrm. Deparlmanl of HwINi Service, July 1997. I Od CCG CenuRinq Gnup. Ths food rervi<ea and hospitaliqq secbr and a 70 Field Rewarch Coryormion. A survey of California bar pahons about metro Toronro smoking bon. 1996. A<csas doh: 3 December 2001. smoking poicies and smoks(rea ban: California Department of Health URL ht1p://193.78.190.200/whose/rostour/resfauohtm. Service, March 1998. 105 Economist Arlviaery Grwp Lld. Tha pobnlial economic impact o{ a 71 Ryland 0. Cummings K. Consumer response ro 1he Naw York Ciy smoking ban in resbumnh. The Rssburant Ass«iation, Sepbmhar smokehee air «t. l Pvblic HeoBh Manag Pr«I 1999;5 I x:28-36. 1998. 72 Hyland A, Cummings K. Ruburabur repom of Tha economic impact of 106 The Publievn Newspaper. Reading the smoke signals. Market Report the New Vark CilIyy Smokehes Air Act. 1 Public Heahir Manag Prod 2001: smoking 2001. p 22. 1999;5111:37-42. 107 Fabdzio Mcloughlin and Aueeialw Inc. Impact of smoking bans on 73 Lam T, Chung 5, Tam E, el d. Public opinion on smoks{rse reburanh. amoksn dini~sg out paNems derived from national survey of adults Hong Kong: Dsparhnanl d Communiy Medicine, The Univsrsiy of Hong smokes. 1995. Accsu dale: 3 December 2001. URI: Kong, October 1995. htry://www.forces.org/widenu/avid/bans.hhn. www. tobaccocontrol.com 2p Scollo, Lal, Hyland, et al 108 Eobrizio McLaughlin and Aueciela Inc Survey of New York Ciy 113 Hild C, Larson E, Weiss L, el al. Reviaw of Municipally o/ Mchorage Rastaorahurs. 1996. Acau doh: 3 Dawmbsr 2001. URL: Chapter 16.65-prohibition of smoking in publk places. Imfihh For hnp://193.78.190.200/svhosp/fabirysn.htm. Circumpolar Health Studies, October 2001. 109 Hodges I, Maskill C. Asxsslrg dta pobnh'al impacs of resrouronf and 114 Enterprise MorkNinq and Rewerch Amcn. Ths effecs of recent bar smoking bons on visiron ro New Zeobnd. Auckland: Healhsearch, evenh on ha Tasmanian hotel industry, November 2001. I B January 2001. 115 Jena N, Wakefield M, Turnhdl D. Aeihdes and esperiances of 110 Douala Gounly CemmuniM HwNh Improvement ProIM. Douglas resrourahun regarding smoking bans in Adelaide, South Aushalia. Count' Bwiness Smoking Survey, June 2001. 1 17 Huron County Wdlh Unil. Hunan Count' Meath Unil resrourom Tobacco Conho11999;/:b2~. survey, June 1999. 116 lama DE, Bero Vs. Why nviaw adides on he heath effects 7 12 Shapiro T. Butt of he law hin rssrourants. Busineu Day, 15 August of passive smoking reach dilerent conclusions LAMA 2001. 1998;379:1566-70. t e 9 s e r e' ~ ~ ~ Ciy of Kingston, Canada: Smokehee By-law Campaign 2002. ®Kingson, Fronlenac and Lennox & Addington Health Unit -J. Chenier. wvrw.tobaccoconho.<om The Dunham and Marlow "Studies" on Smokefree Restaurant Laws More Tobacco Industry Hot Alr on Smokefree Restaurants August 23, 2000 The tobacco industry's primary argument against smokefree restaurant and bar laws has become the claim that these laws will hurt revenues for the hospitality industry. These Gaims have been studied in nearly 100 different communities using objective sales tax data and found not to be true.' In an effort to keep this argument alive, John Dunham (of Philip Morris and Michael Marlow (of Califomia Polytechnic University, San Luis Obispo) have published two papers2~ claiming that these laws hurt business even though there are no overall revenue effects. These papers are a great example of how one can do a sophisticated looking statistical analysis on a meaningless data set. There are many problems: * The entire analysis is based on a survey conducted in 1996 for the National Licensed Beverage Association (a major tobacco industry ally) of 650 restaurants and bars asking what the owners predict will happen if they are required to be smokefree. Such surveys have almost always shown that people predict ill effects; the actual data on what happens has never confirmed these predictions. " The analysis separates states that the authors claim have laws restricting smoking in restaurants and those that do not. Of the 32 states that Dunham and Marlow say restricted smoking in 1996, only 5 had specific requirements for the size of the nonsmoking section, and some preempt local ordinances.° Given the large errors in their assessment of state laws, the results of their complex statistical analysis are meaningless. " Dunham and Marlow criticize the earlier research on effects of smokefree laws on revenues on the grounds that they do not study profits, yet Dunham and Marlow do not present any hard data on profits, just opinions on revenues. * If 38% of establishments experienced lower revenues (and only 6% experience increased revenues) as Dunham and Marlow predict, this effect would be large enough to detect using aggregate revenues. The aggregate revenue data based on sales taxes shows no such effect. " Dunham and Marlow ignore the fact that the State of Califomia did an analysis of the effects of California's smokefree bar law at the level of the individual establishment, and found that the law did not adversely affect business, and was associated with an increase in business s " Dunham and Marlow criticize the early work by Glanfz and Smith e'' which showed that smokefree restaurant ordinances do not affect revenues on the grounds that they studied all the cities with 100% smokefree restaurant ordinances that had been in force for long enough to get data on actual effects for at least a year as opposed~o a random_sample..IncLudina all the _citie_s_ is always better than a_ sample. T_ here are no_ biases when one considers all the cities. Considering all the cities is not "cherry picking." * Dunham and Marlow state, incorrectly, that the communities that enacted smokefree restaurant ordinances have "relatively few smokers." This is not correct; the adult smoking prevalence in the cities Glantz and Smiths included varied from 18.9°k to 25.2%, well within typical smoking rates at the time. * Dunham and Marlow rely heavily on an critique of Glantz and Smith's earlier work prepared by Michael Evans for Philip Morris' National Smokers Alliance, which was never subjected to peer review. After the National Smokers Alliance submitted Evans' critique to the American Journal of Public Health as part of a complaint about the fact that the journal published the study on the effects of smokefree restaurant ordinances on restaurant revenues, the journal subjected Evans' critique to peer review and the editor observed the following: Instead of a compelling critique, however, we find a melange of scientifically inadmissible manipulations of data to obtain a desired result. These are conflated with a Flurry of suppositions as to what could be. Suppositions are then translated by mere assertion into factual "serious" Flaws. Even if Evans critique can pass in the field of economics for legitimate science -something I do not credit-it cannot pass in this Journal. Certainly, no economist has submitted so tendentious and meretricious an argument during my tenure.e " Dunham and Marlow claim that one reason that there is not an effect on revenues is widespread civil disobedience of the law and cite two newspaper articles to support this claim. They ignore the results of the California Tobacco Survey, a large scientifically valid study, which showed that 75% of respondents reported fast food restaurants, 75% of family restaurants, and 70% of full service restaurants were smokefree. This is hardly "widespread civil disobedience." e, r' "Z Dunham and Marlow ignore the conclusions of fellow Philip Morris employee David Laufer, in a presentation to other Philip Morris political operatives: Also, the economic arguments often used by the industry to scare off smoking ban activity were no longer working, if indeed they ever did. These arguments simply had no credibility with the public, which isn't surprising when you consider that our dire predictions in the past rarely came true. 10' P' 28 Overall conclusion: The Dunham and Marlow "studies" are just repackaged versions of earlier tobacco industry claims that have been proven wrong. Policy makers are still left with the choice of basing decisions on what actually happens (using sales tax data collected by neutral tax authorities) or what people predict will happen (using biased surveys conducted for the tobacco industry). References: 1. Glantz S. Smoke-Free Restaurant Ordinances Do Not Affect Restaurant Business. Period. J Pub Health Mgt and Practice 1999;5:vi-ix. 2. Dunham J, Marlow M. Smoking laws and their differential effects on restaurants, bars, and taverns. Contemp Econ Policy 2000;18:326-333. 3. Dunham J, Marlow M. The effects of smoking laws on seating allocations of restaurants, bars, and taverns. Econ Inquiry 2000;38:151-157. 4. Welch C. State Legislated Actions on Tobacco Issues. Washington, DC: American Lung Association. December 31, 1996. 5. State Board of Equalization Statistics Section. Comparison of 1st Quarter 1997 and 1st Quarter 1998 Taxable Sales of Selected Eating and Drinking Places with General On-Sale Licenses. Sacramento, CA: State Board of Equalization, Agency Planning and Research Division. October 1, 1998. 6. Glantz S, Smith LRA. 71re effect of ordinances requiring smoke-free restaurants on restaurant sales. Am. J. Pub. Health 1994;84:1081-1085. - - 7- - -Giantr S; 3mimz: E., atop, ~ ;or'~he effect of ardinaneesrequiring smoke-free restaurantson-res•" ~.a~. o.^.~r - sales. Am. J. Pub. Health 1997;(in press). 8. Susser M. Goliath and Some Davids in the Tobacco Wars (editorial). Am. J. Pub. Health 1997;87:1593- 1595. 9. Pierce J, Gilpin E, Emery S, Farkas A, Zhu S, Choi W, Berry C, Distefan J, White M, Soroko S, Navarro A. Tobacco Control in Califomia: Who's Winning the War? An Evaluation of the California Tobacco Control Program, 1989-1996. La Jolla, CA: University of California, San Diego, 1998. 10. Walls T, et al. CAC Presentation #4.: Philip Morris Tobacco (www.pmdocs.com). July 8, 1994. Report No. Bates 2041183751-90. Prepared by: Stanton A. Glantz, PhD Professor of Medicine Institute for Health Policy Studies University of California San Francisco, CA 94143-0130 email: clantz(7o medicine.ucsf.edu • DRAFT FOR DISCUSSION PURPOSES ONLY Corrected Document Executive Summary: Smoking La nd Their Differential Effects on Restaurants, Bars, and Taverns By John u tetra d Michael L. Marlow July 20 ~ ailt~'' Q'Y`aM/-(~ OVERVIEW OF STUDY: ~ ~ty ~ IY1`~_ V Critical review and analysis of three published studies that conclude smoking bans present no adverse economic consequences to restaurant, bar and tavern profitst. This paper concludes that these three studies are based on faulty assumptions and methodology. Looking at data from a national survey, this paper determines that the effects of smoking bans are differential. (or. smoking bans exert different effects on different establishments.) FINDINGS INCLUDE: The three studies in question have concluded that smoking bans present no adverse economic consequences to restaurant, tavern and bar profits, however: • Comparisons between cities with smoking laws and those without did not control for non-ban featttres such as tourism, weather, tax-rate and employment growth, all of which could influence business. • The sample groups were not random. The fvst 15 cities to enact bans were used to collect data These cities have relatively few smokers, giving the authors of the three studies, biased samples, reflecting relatively few adverse economic effects to business profits. With few smokers to begin with, it is no surprise that profits didn't decrease in these sample cities. • As observed in other studies, ver~ rarely are smoking bans uniformly enforced by restaurants, bars, and taverns. This makes it difficult to make predictions about how other businesses might be impacted by a smoking ban. • The studies focused on revenue, nor profit, leaving out owner compensation For lost profits, such as raised prices, lower quality and increased advertising or promotions. Citing a nationwide telephone survey taken by Roper Starch for the National Licensed Beverage Association, Dunham and Mallow determine that smoking bans exert different N effects on different establishments. v The telephone survey used owners from 650 restaurants and 650 bars and taverns, ~s ~ ~ drawn at random from national lists. The owners were asked questions about seating d~ z p allocat%ans, attitudes towards smoking laws, strategies towards non-smoking and ~ U^~ • Paper baud in part on a study conducted Cor Philip Moms Management Corp. ~ ~ t,~,, ' Glantz and Smith (1994), 6oldatein and Sobel (1998), Sciecca end Ratliff (1998 ~L ~ 1" ~Blankenstein(1998),Canto(1998) V.'~,,n 4' ~~i DRAFT FOR DISCUSSION PURPOSES ONLY smoking customers, revenues and their projections on effects of smoking bans on revrnues. FINpINGS INCLUDE: • Restaurants with lower percentages ofnon-smoking seating allocations predicted losses. • Restaurants with higher percentages ofnon-smoking seating allocations predicted gains. • Bars are twice as likely to predict a loss in revenue. N O N V O DD N O N Smoking laws and their differential effects on restaurants, bars, and taverns Joy i~ a~ ~ 1,. Marlow xeprioua s~ co~emporory Ecom~mic Policy, Volume 18 No.3 (July 2000) pp 326-333 i I i 1 i i N O O~ (i~ J e0 ID V tD original a~ticlc pnblist~ed by Oxford University Press, Great Clan;ndon Std, Oxford OX2 6DP, UK. Tel: +44 (0)1865 267827, Fax: +44 {0)1865 267774 www.cep.oupjournals.org SMOIQNG LAWS AND THEIR DIFFERENTIAL EFFECTS ON RESTAURANTS, BARS, AND TAVERNS JOHN DllrIHAM assd MlgiAfi L MAF[tAW' 77rrs article eramirrts the effect of restrictive smoking taws ore nsraurantr, ban, and taverrtt Supparos of these lotto often argue shat they do not harm ffrmt and . etKrr ~P~~ Oppartantr argue that owners Dater to crtrtomer smo ' and lance man ~t4pnferenees, a ~ ~+ae/t' impact Pmhtc 77+Ir anicte provides neuq and danantrrates dmf of ~`~a Iowa amt as busi- mtanbtg~rd undersWrrdigg of the erorronltc dated brwux ~ ~ a fircnt eJj`ectr on dtffenatt front. 7Ae dirbfbut~co dose amp talus seer dif- of eQeetr 4 erambted trsirtg dam from a nationwide stoney of 1,30'0 rerlauranR baF and taata7r owners Wlrite acne sub- sea of frrrna arc predicted ro suAFr nt+rmu declines barn arc predicted ro be more Uutn navice ar Ike!) to experience Java as rataumntx An important implfwtlon V drat the incnaslrtg lever of governrnenfa! resnicrions on smokirrq in dre haspitaliry sector could gradually impact the typo of service avaltab& ro the pablie. (JEL RZ, Ho} t. N,1tAOw1CnDN their establishments, others to forbid all resVicting smoking in restaurantr smoking, and shit others to aeoommodate ers argue that thex laws do not Itarm~firms sepaza~anas ode in~g in ~rtition8 and may even raise their profits. Recent smoking patios or rooatq and aQ filtration studies fmd that bona on stnokirtg in eatieg and drinking pkaas have not impacted lbe little economic research has been pub- profitability of these establishments, sug- lished on the effects of smoking taws on the gearing that smoking bate either do sot ~~~ry of rtxtaurants, bass, sad taverns, reduce detaand or offset sales losses by and almost none has been directed toward towering costs. Opponents of smoking restric- the issue of how these laws may exert dif- tions argue that soma owners would find ferential effects on businesses. This articie it profitable to allow smoking throughout a framework for ezambrittg how smoking laws impact establishments based on a ntunber of business attn'butes. Hypothe- "Fhh u araised rossion or a paper presenuA u seS regarding the effect of smoking laws are ehe Weaere F.woomse Asodatimt international ammat tested using data from a nationwide survey of ~nN ~ ~ >h`a°. G. ~tM~a°`tM~~ p 1,300 restaurant a~ bar owners. The empiri- nmax aonduned [a cam. Vk thank PKNiun aoyea, pmt chatoupta, Bcith cal t:vidence indiwta that smoking laws exert Warner and three anwryttmos rc&rees fa tbe'a helptnl differential effects on businesses. commend Dunhassc Masagec of Fool Issues, PhiSp Marra Matmgemrnt Cora. Marloac Rolemy Dcpanmcm olFaonaaiq falltornia tt Gfif11CAL Rt~EYV OF THE U7EgATURE ON 1'olytuJtnic State Univessiry, San Iafe Obrspp, C.t SMOKING LAWS Fat 917-66I-5379, EmnN john.dunham~ypmmc.com 1. Smoking lows hwc been impoxd on testae- A. Smdia Canrlude Thal Smoking Laws rams in Alnlu, California, Canxe[ieat Delaware, DO IVOI Ffarm Front F9oAda, Hawaii, Itlabq INbwis, Iowa, I4tws, Lou6ian; Ma1ne, Ma~achweus, Midiigan, Mitmaata, Musauei, 1Vtu7e a &tll examination of the Welfare Montana, Nebtasta, Nevada, New Hampshire, Ned effects of smoking laws must consider all racy. Ncsr Ymk North Dakma, Oklahoma. Orogon, Penmyhatua, Ahode [shod, Utah, Vermont Virginia, members of society, our literature review Washingm,r and wam„sia focuses on the economic effects of smoking cask 2085579980 mporary Economic Policy ASSN 1W43529) Vol. 18, Na 3. ]ssly ?0IX1, J2bJ3J O Western Fionomic Assoriation Intcmational nunntnat g awtlow: sMOxutG rites ]z7 laws on the owocrs of firms.' Three published Their random survey of tt00 owners rnttdutled studies have conducted that btraioesses do not that, white 52% last tmsirtess, 6%gainW iwsi- suffer redrrcedsales as a tesuR of berm. GLmtz ness, and the rear did trot know. Of the 37 and Smith (1994) compare 15 cilia wig t]amets with favttnbL: effah, 32% had out- smokinglaws with IS matdted control group door seating, and 57%hadbarseating-areas cities Thry cronclude that "legislators sect exempted at that time- govcrnment officials can eoact such bealth The studies that find no adverse effects are and safety requirements to protect patrom unable to detect differential effetxs beattse and employes is resaurants from the lox- they aggregate erg firnta in a community im in second-Land tobacco smoke witbout together. This approach conveys contusion the fear of adverse economic cotuegerences" about effects m¢rted oa iodivldwl 8rtos. gut (p. 1085). In their smtk of smoking Taws in an uodetstanding of differential eBaxa is North CaroGn4 Goldstein sect Sobel (1998) eaaential for a comprehetrsive :acid welfare conclude: "Even in the number otu tobacco- analysis and wtwld lie necessary in detertnin- produciog state in [he US., E7S regtdations ing appropriak leveb of compemation for present ~ adverse ecoaamk impact, sect injuttd paRka within the broader framework ~ there is no need for exceptions to the ordi- of social welfare aoalyoa. i natters based oa such hers' (p. 286). Stdana ~ and liatliff (1998) conclude in their study of ~ Arizona firms that "[t]his stttdy seems to indi- C. Problems with Comm! Group I care that prohibitirtg smoking in e0 Flagstaff Mrrhodology restaurants has had m effect on total tmttan- Com~trsom between cities with smok- rant sales... , If tLeae findirtgs sre true for (ng laws and those without reveal efftxts communities throughout the United States, from the Taws only whin stadia properly then other cids rxn enact simiLtr lava, whkh control for nonban factors that affect setae protect patrom and food servioo arorkers from Previous stadia matcL dries on the basis tobacco smoke, without conarm chat restau- of poptdstion density sect sme and trtcdian rants will lose buaincss" (P• 1~)• income. However, it is unclear whether these factors control for differences k totuiamt, B. Previous Studies Canna Uncover weather, tax rates, and empL>yment grt7tvdr, { DifJ'ertntia! fiJjeces all of whidr could also influence businesses.' Momtmr, studies compare ratios of restau- These three studies all rely on a metbodol- rantlbar sale to total sales to examine eco- ogy that examinee aalea rat revenues in dries nomk effects of laws on firms. GIaMz and that Lave enacted smoking bare by competing Smith (1994 for exunpk, state tLat, "If an them with control cities that have rwt adopted ordinance adversely affected restaurants, this such laws There arc seritrm methodologial traction would be exaxted to drop when the problems with this approach.' Otte problem ordinance was in fora" (p. 1088). However, is that gross rotail Saks or tea revenues are when Worthen factors are not controlled for, not useful measures by which to gauge ecro- this ratio cannot reveal independent effects nomic effects on a diverse set of firrss. For of Taws on firms. example, on an aggregate basis, firms subjeM to a ban may show no net tffect on sales even though one-half of firms showed gains and the D. Sample: Are Not Rwdom other half athibited losses. It Is not portico- p~m studies have not used random tarty ttselul [o conclttde that nothing oavrred. samples For example, Glantz end Smith 1?vidtnce of differential tffeeU was uncov- (1984) chest "Ihe first 15 US dries to enact eyed by a Peat Marwick (1996) study o[ [he smoke-free ordinances affecting restaurants" 1995 smoking ban in California restaurants. 108 This sate bn osea a (P• P • 8 P problem ]5a impan nt suns leas on cmnmu and because these locatiom Lave relatively few producer surplus mttW ahq he eaammW; see 73eeler smokers since otherwise they would probably and awrpnr (199]) and 7h:mMav aa1 t}emh6r (1995). £wan's (1997) renew or Glaah and Smah (1997) d. Glantz std 5mah (1994) xketed control sties paints out numeroar agtirdral laws and caneludtl that that did apt nlaMate smoYina tiara; however, Evans findings of no etfeas arc uawsrranteJ. (1991) touod that coon than IwlTwert mec4ssificd 2085579981 320 COtrIBMPORARY ECONOFIIC POLICY not be first to legislate bans A sample aelec- G. Summary tivity bias arises because such locaions world Previous studies offer little information tend ~o experience relatively little harm from about the economic et[ects of bans on retau- bars. Moreover, local legislators arelttely to tanrs and bars. Sample selection bias and a be optimistic about Iota! economic condttiora statistical methodology based on community since thry are unlikely to promote laws that average and control group rnmparisons do jeopardize tax revenues. The authors then not clearly isolate eoemotnic effects. No infor- appear to have handpicked laattorm vvltere motion on bow such efecrs differ between economic downturns in the restaurant and bar etablishments is roported. Many nonban fa~- xctorsare tmlikeiy following passage of bans. tors that affect revenues are not properly - Pmptrical reaotts arc then subject to two contro8ed for, leading to serious questions xkction biases Fast, authors have "cherry- about whether studies tell us much about picked' samples that wi8 ezbtbit zelatively (he effects of smoking laws on firms. Impcr- few adverse effects from laws. Second, sear fen eaforcetnent and sample selection prob pi'tag calls inm question txsadtreiom that teens suggest pproblems with predictions of adoption of similar taws to other Itlwtiaas how laws wotttd affect other cities. Finally. will yield shnilar etts;cta. GStie with laws and studies focw on revenues, whereas profits arc those without (control groups} are disa;as0ar dtc more ttsetul indicates of economic effects when tt comes to amokiag tsataes, and it 15 on lttns. tnlikely that adoption of bans to odor titles It is also important to realize that fiad- would go as stnootitly u in titles that were togs of no negative effects oa revcnttes in the first to adopt them. aggregate could be consistent with a state is which firms have not reached an optimal mnc of smoking and nonsmoldog seating. How- E ~ Baru Art Nor Unffonnq ever, this situation does not ntxessarily mean that banning smoking is effrcieat because Smoking bats arc not always folly change in community attimsks would most enforced In California, for example, liltely ]end m seating changes by firms that widespread tan? disobedience appears tom- ace morn optrnrsal than from govemm~ent reg- mon (Jilanksteia, 1998; Canto, 1998). This ulatitm. No atlvetse revenue elfats does not indicates slut some pertx:ntage of fume wotdd nay mean that future revenues would fie adversely affected by fuU twmplianca- not be lower than they would otherwise have thus validating claims that smoking bans been as firms adapt [o changes in smoking adversely affect some firms. Noncomplying Preferences. However, lack of adverse effects firms would also lead to suffer groatet effete may simply be a result of serious statistical from fuller wmpliana. Differences is cam- flaws in previous studies- pliance rata compound the problem of nnak- ing prediciom about how other locations s. ~ ~ tADDO. OF SMOIONG LAWS might be impacted Coax (1960) provides a general frame- work for how toms efficiently deal with the F. Studies Focus on Revenues, Not 1'rofu smoking issue in the hospitality industry, and demonstrates that private anarlcets internal- Owners may attempt to overcome lost iu extema!(ties when [here are trivial trans- business by raising prices or lowering product action costs and property rights are dearly quality. Owners may also iacreax advertising assigned ~ Although this study does not focus or provide promotions that add to costs. if so, on the externality issue per se, applica- profits will change. Previous smdics examine lion of the Coase theorem has impoRant only revenue, which do not indicate profits. implications about how owners allocate air spas within etablivhmenu and therefore 5. Duntum and Marlod (2000) anger ital. whde how smoking laws will affect profits.' restaurants aul openk m states Nat hwe exacted smoking restdniom also tend to arttt a larger slave 4 See Boyer and Marlow (19%) [m a more detailed or Iheu sestina to noxsmotitg rue, tltia hu litre Io do dimaaioa. wish presenq of a smoking Int. Rather, their eastooren 7. Ban proponents argue dun tnxpayets subsidize simply have atroager prerertngs fm snch noting. higher heahh cart cents of m,okcrs Howcvcr, Ise 2085579982 ntrtatfut a kt~ltcow: skfoetntc uws 349 Although the air spas is commonly cvattnner bases and marginal coats ~ acwm- viewed as a public domain resource, this is modatioa Another testable hypothesis s that move within the oottfmes of private toms. adverse effecffi are higher fw firma wteting Whether m ownu caters wkly to soak- to relatively msnp stnoke~ as flmos with taw ers w to nommokers, w aooommadatm sackers haw ahrady found it profitable to both, depends on wsnvner prefcrerMes and rolmtar0y meet ngulationa. margiml cotta of ttccommodation.° 'I}ansarsion costs are not prohibitive w o~slanort oc anra because owners act as inttamedierks between smoking and wt>:[roking customers. Akgoti- Survey data co0octed by Roper Stench anon tlccurs via owners who determine what for We National l,iceosed Beverage .gaso- air space allotstion is consistent with matt- ktvs iofiu ban. The~xk~ imt>m proflffi, and is the way, Private mar- P~ surveq keffi tt:nd toward intemiliaticm of enter>mli- was ta>adoRad during the period atF Septenr ties. Iaaortxt decisions displease patron and bar 5-12, 1996. The sample mmieted of ' lower 8ta owneratmmegers ~ 650 restaurmffi sad 650 Smo~ng regulatiwa shift ownership of the bara~taverns across the United States. Seen- i air apex from reaffiurmt and bar ptnpd- P~ wen drawn m a statistka0y rmd~ ' etora ro iadividtnls who prefer that reshio- mercer from nstitmel lists provided by Sur- ~ bons take plan. Options facing atstomets °C1 Sampling, Inc., and an applicable to all sorb eatablahmenta with a mashaum cam- i may ehsng°, and amotmg tatstamers may j choose to spend more time at hoax ether Pfing nor of aPP~*~' ~ P~t- Wmgoing out, whit wtumokera may apcad ~ Points. The survey imtrument iodudes ~ more time at firtas when they believe that 9°OariODS pertaining n seatirtg aeration, 1 laws im etrvlromtentsv attitudes toward smoking laws, atmtegiw to Supplyp ~id~a eHeds arise when laws change deal with kmg atammcts, tpet by requiring tbat firms atpend more reventxs, and ptojeedau of effects of amok- s retastssces eccoasmodarittg ntmsmokaa m ffig laws on raventxs. The sample v tlivLied the Lospitality industry, Proms are deter- into stake tviW and withtait ~d laws, mined is a relatively free market, with a ktge ~S the dasei8ation system of the Center number of suppliers snd twtoaxra10 Owe- fw Disease Control that de8ow amotiog bras l ers detetrnioe hryela of different attrt'6trtes, as laws alloying w nquiriug aoaantokiog ~ec- incltrtHag pda, tptality, qumtity, cuisine. and tiartc in restanrmnatt smoking aoeomtttodation to mathnbe prof- The sample provtties many advmmges its. 1[ is m empirical gtuation av ro Low overWceetaedinpreviouasmdieaThissmdy profits will change following stnolmsg laws '~Q tLm-level data, fac0itatiog the analysis of A testable hypothesis is then that reatrictiom dishrbuhonal impacts. Thu is a rmd~ sam- ple and dos aw sttfkr from selectivity bias eurt differential cffects m pro8ffi. F.ffecffi problems Tlx random naln[e d tLe selec- should also Yary between restataaat end bar tiaat 1 oho m:nimiron. potential fw bias industries, simply because of tSffereoces In Ming to imperfect cpmpGance issues. Thts (t99t a tbu bus ao not a0ae4 ate ~Ple does not [hnoae firma on We basis while AI ~tvn( ) usves dre soaey tu.. naLt of wLether Way operate order amokmg laws, thek euemal err bsot ~ also and question were pond to owners under aoa Zimmcrmm (1991), who emu tWt paaFe amolre We essuvrpdon of fu0 enforcement. Bu[, a9 tick bas been verenfmared by OSHA with previous studies, this date set focuses on a rrs impormmw otamommodanae nomaokem h reventrea An emminadm of profits wotdd be evident io indastry trade tugati0a. Sec [a mumge, preferable, but we lmow of no data that cm- wWer (199<) utd Fcachtmao 11992). 9. Cmua a aL tt996) (nand teat smoked diced out nee profits with aaioking Iaws. However, the less hequmt[ly toaowing New York (Sys hers d 1995. On avenge, emokers onnpwtt nonwakars by x2138 per 11. I.xd wmmooide may oho irapeae has, but week thus are mt ioelodN beer bmure we do sot knew ter. Pisauraot sad Inr Durkee aro sot perfatb menu beatiomr of theta Howewx, bal laws me ku mmpetitivc heause produen ate not 6omgenaour common b senora w(fipot So>abng ttws-fin the rtau lwvmver, bxause of ease of mny, ptoaa might go ro meawn provides ¦ Nod manna of reahicdam Lad nonnd N the bug run. b9 buvinaeea 2085579983 3l0 CONTEMPORARY ECONOMIC POLICY TABLE 1 reason to fear economic impacts. But, assur- Erpected Revenue Changes antes of no effect appear somewhar mis- M t1rAS Snoki¦a No placed green that 38% of owners [hat have ra. G.n already operated under some form oY a law predict that virtual bens lower revenues. Percentage of Restaurants Table 1 also displays expected revenue Frigher 6x sx rx changes by owners of bars. From the entire Aevanues sample, a ban is predicted to raise revenues Loa¢r 39 sa bz by2%ofawgers,lowerreveauesby83%, and Revenva rodutt rto ells No Ctunge 51 51 49 p nge by 13% (2'Yo provided no ooa•r Xnow a s 3 prediction). Again. using a X test, responses Observnnrm 650 444 ~ do not differ signdicaaUy when brokrn dowq Peroenuge of 9¦rs and "r¦varm into stales with amt without laws, thtu indi- Fllgher zg, 2x lx eating that firms that have already experi- Rerenaa enced smoking laws offer sinillar predictions. Ltnrer ga SS 75 A X= tert also indicates that responses Raveauq sigtti6ceatly differ between restaurant and No Cttsnge is 71 18 barhavern owners. Owners o! bars and tav- DoNt Roav z z 6 ecns are more than [win as likely to predict obratvatlum aso sza i~ decreased revenues, indicating that laws are expelled to impose effect Wat differ across 1ndLL5friea. focus on revrnues allows for the direct com- parison with results from previous studies. It is also importagt to note that economists tend VI. ARE OWNER RESPONSES CONSISTEM to be stupiclous of survey data asking about wtrt+ T~ Ecotlowc tnooEL7 probable effect. We attempt to overcome this A potential problem with this analysis problem by comparing responses of owners stems from the survey nature of the data who have already tern attbjeU to smoking and that predictions grey not reflect actual restrictions with those who have not, and we events-a problem characteristic of survey are inclined to rely more heavily oq responses analysis of possible future events This short- by fums that already here experiegced amok- coming is ameliorated m large degree by ing restrictions, determining how responses vary by current seating allocations. That is, do firms that pre- y. OISTAIBIIRON OF ECONOMIC FFFEC7'3 diet adve[se effects tend to allocate relatively few seat to nonsmoking use-or cater to Owvers were asked haw revenues would relatively marry smokers? If so, those firms change Yollowing laws requiring that all firms catering m relatively few smokers and allorat- become virruallynonsmoking, Table 1 displays ing relatively many seat to nonsmoking use expected effects on all rtstatuant, as well should predict revenue gains or no revenue as Fnoken down into states with and with- effects. Moreover, az disrnssed above, these out smoking laws. From the entire sample, a owners may now attract smokes that pre- restrictive smoking law is predicted [o raise viousty were not pleased with their policies, revenues by 6% of restaurant owners and and establishment that previously catered to lower revenues by 39%. Just over 504b of smokers may now compete with other firms owners expected no change, and 4%provided on the basis of price or quality. no prediction. ONy data on restaurants are examined A X test indicates that Use pattern of because of UtUe variation in stating allora• expected outcomes reported by owners in tiotu of bars }nd taverns. Table 2 indicates smoking law starts does not differ signifi- that 40 restaurants predicted revenue gains, cantfy (5% level) from those in no-faw states. ~q predicted revenue losses, and 329 pre- Managers with experience of dealing with dieted no changes." Relationships between laws then offer identical predictions as those without them. This 1S interesting ltecause ban 12. 2k<nty seven owners did not Imow what to pro- advocates often argue that owners have no diet and have been atdudcd from Wit analsSis. 2085579984 DtlNtrAM 3 MARI.OW. sntptcota [.AWS ri1 predictions and seating allocations are coatis- where teat wi0t erpectations. Fums predicting losses have rektivtly low seating allocations to non- NS, =percentage of seating allocated smoking use (38%). which strongly indicates to nonsmoking use, that they stet to large populations of amok chain; = 1 if fvm is a member of a era. Firrsr predicting gains report relatively corporeec chain and 0 oUterwix, high seating nonsmoking aflotationa (7096), which suggests they expect to attract smokers a8er =Years Srrrr ~ been in bttsineas, who previously did not like rhea nonsmok- Bizet =number of employees (full-time ing policies Finally, 6rnu predicting no rev- ~ p~_t~) erne eReets have aeathrg aflacationa (54%) much higher than those predicting losses. and = L if bu/tavern and 0 otherwise, slightly below those predicting gaGts. and Tabk2alsodisplaysseatiog acationsbro- law;.1 if state smoking law is present ktndavrtintoreataurantsthatoperateinatates and 0 otherwise. with and without smoking laws Again, under 6odt oondidons, owners preditting revenue The dependent variabk change; = 0 if owner ' gains or no change tended to albcate relatively expects either ao change or a rise in rtwenues, marryseabrortottuaokingrrse,andownerapre- andlifreventresarepredictedtofell. dieting revptue lasses twded ro elkrcate vela- The percentage of seating that owners tivelyfewseatsrorrortsmokiaguse. allocate to noasmokfrtg use 1QS; is expected In sum, predictions of revenue changes to exert a negative influerrx on the probabi4 are corrsiatent with the economic model of ity that revenues svifl fag since this variable seating alhxxtians and support our empirical indicates how rttany nonsmokers are served. esddenoe demonalrating [hat kws exert differ- Whether an establishment is a member of cults! effects. a corporate chain is expxted ro btflueace I Ibtential for bias deurvca reiteration responses if drain member offer greater Owners may oppasc smoking laws for pex- accommodatiat of smoketa and nonsmokers Bona! reasons and, as a result, exaggerate as alt element of ovawfl corporate s[categy. [heir expected coars.ts One m' [imagine that If true, chain rnentbers would be less liltely smokers ate interested in odter amok- to predict revenue falls and the expected sign era, sad they may exaggerate coals of laws. on chain; is negative. However, we have no information on the years in business, age;, is expected to likctihood of Ihi4 event, and know of no data pt~iv~, affect probabtlitles of revenue set that contains actual profit effects and drops. Aocommodation costs may fie post- seating allocations. Moreover, while owners tively related ro age of buildings, and older may have personal views, it remains audear firms aeoortunodate Icss, given that they whether views regarding smoking wouM over- (,end to cater ro mom established end sta- ride prekrencta for manmmng the value ble customer bases than newer businesses. of their firms. Whether they profit-meximi¢e Firm size, size,, is rotasured u number of or cater to pctsottal preferences regarding employees and is expected to exert a neg- smoking rcmatns an issue for faints research. alive influence when Beak economies exist - in accommodation. For example, larger firms VII. LOGR MODEL OF l7fP[.CfED may find it easier or cheaper to separate eetEeate etlat[t~ES smokers from oonstnokers. A positive sign A qualitative choice model estimates the is expected on bay following the discussion probability that an owner wtith a given act of that bars taverns find it more costly to sep- attributes predicts that bans lower revenues. orate smokers from nonsmokers and their The following logic model is estimated. customer bases have less preference for non- smoking xating. Table 3 displays logit (I) change, = f(NS„ chain;, age;, ma[ions for all firms, restaurants only, and size„ bar;, law,) barsJtavetns samples, Esthnetian based on all films supports !3. see aerrcns et at. (I99r) and KerkNiet (1994) many of the eapeetationa. Higher shares of for wncems wqh svcVey mt.. nonsmoking seating lower the probabflity that 2085579985 332 COPtI'EMPOIUAY ECONOMIC pO]ICY TABLE 2 Porcentage of Seatilig Allocation to Nonsmoking Use by Restaurants Predietlw Mee MaNaa Std Daa Nember Ali stars Itcveoue gads 70% 95% 39.9 40 Revenue liana 36% 37% 36.1 Z54 No c6arya 64% 75% 36.7 329 Smoking Law Staoea Rtvenue gaits 72% 10091 39.1 27 Revenue kma 41% 45% 37.1 167 No changq 68% 85% 39.3 229 No-[.aw SWa Revenue gslm 66% 80% 39.7 13 Revenue loxes 31% 2591 33.6 87 No changes 36% 6096 381 100 TABLE 3 Logit Model of Revenue Change Ail Parma RerlaufaNa BaNraverns Constant 0.26 033 1.39• 1.36 139 535 NS -OA2• -0.02• -0,01••• 6.L1 7b8 1.66 Chain _0.29••• -033••• -0.25 1.86 Ln oe4 Age no1•• ool•• ool 220 1.95 D.96 Size 0.01 OAI••• -.00t 138 1811 053 13ar 1.30` 7.69 Law 0.13 0.03 0.43 0.94 o.lr 1.62 Log Likelihood -648.73 -380.86 -26552 Ohs with Ikp .0 465 769 96 OM with Dap =1 793 294 539 Obsavarlom 1256 623 675 Dependant wuiabta = 0 ((owra expects ao c]'anae. or a revenue iocrcau, = l i(owoer expena a revenue detrcaac. Arympmtic r shtitlm am given hekrw ateoattd me~icienta. . dcoou sigtrianrrt et .Ol, A9, and .10 kvak rupectively. owners eapect adverse rt:venue effects, chain ence of state smoking laws do not signifi- members are less likely to txpetx revenue cantly influence the egtratitm. reductions, older limo are mote likely to Estimation based on onty restaurants expect revenue derlineR and bar ownets are supports the same significant relationhips, more likely 1o expect revenues to fall than except of worse the hart variable has been are restaurant owners." Firm size and prey- removed from the equation [n addition, firm 14. Ahemstive spadficetbos checked (or model # Size is estimated to exert a weak (.10 Ievel) hilhy between smaa and large auhlehmanx, yonog positive effect on the probabflity that an and otd aubOsbmeon, and chain versus independent owner expects a revenue loss. firers. to all use; the percentage o(rronwrokiag seating Estimation based on onty bars/tavems rcmahud sigelmnt end o(the same sfgn. Tho smorng gilds ust one si ificant factor: the share of law variable remahted irdgni6nat. 37re other variably Yi J Bn rerotixd the same propcnin nonsmoking seating exerts a negative effect, 2085579986 r , DIJNHAM R MARI.Ow: SMOKING [AtVS 333 as expected and is rntzsistent with the other Boyea, WsBiam l., and MieheeJ L Mubw, "7Le Puhac 111 estlmatl065. '[7119 effect is Daly weakly' sigaif- Demand [a Staeltiog t3am," lhh8c Charcc, 88:1-; icant at the .10 level, bu[ estimation «sults luly 1966, 57.-a7. suggest that most barltavem owners expect Cenro, Minerva, -Califormrc Putt on Despite Boo.' that a virtual smoking ban will lower revenues. R4shirgkrr ltw. Sa. a septeasber 24, 199e, p. 7. Ctlase, RomWrl, "rise Problem otSoeiLLCal,"3ourrud afLaw a,d Eawmssurx, 3, Omaber 19fiQ 1--M. ' VIII. (x)NCLU310N Corson, Dnvid L, Cheri a Yonoy and (Sthy A Erq "Should NYC's Restaurateurs IJI,']san Up7` Carrell -[his article examines the revenue effects Itaet a„d Itamaanr Adnsirdrwosoa 37.x, that smoking laws exert on restaurants, bars, Aug ]gy4 and taverns as predicted by 1,300 owners of Dunham, Johq and Michtd L Madw, "1'he B6tav these firths. While a subset of Hans is IHrely oy S~~a Ipso ~ ABOCatbos ~ Renau- to suffer adverx revenue effects following a ,onto, Itars, sad liwetw," E¦vvmntir rnyrsuy, ia:l, smoking ban, bars and taverns are predicted Jaousry 2000, ]st_]sy. to expetience adverse effects more than twice Earn, attehael 1G "A Review M Zye EBed of as often es rtS[811r8nts. '[bin is consistent tOtdigaotxs Rcgddng Smokeliee Rataunuta with "reality checks" that indicated that rev- an Reauurattt Salp," wwwspeatap.osglpd4 emu predictions reflect owrar perceptions mttcom.pdt, Marro 199'7. of customer preferences regarding smoking. E~sdstman, Ilavla J• "staakirsg is Reataunels- Adverse revenue effects were most often pre- ~^na ~ Haan*al~ty ~ Cadgvs:g, dined by owners allocating rclativety little br:7, July 1992, 65. seating to tttJnsmokiag rise, and ptrsitive or GIm4 StaNOO A" and Iha R a Smk4 "The F3rect neutral eHecn were most often predicted by d ordimocea ~nlS°a sin >wt+~c.nb owners who allocated relatively much of their oa Resuraaat Saba," Mrkan fotorso! °f Public seating to nonsmoking use. July 199a, loci-1tB5. 77u interesting public polity issue con- EHeu d ordiatrors ~'1"h'as smoke' Gems the distnbution of effects, and not the Foe Remmuta and Bars on Rereaues: A Folbsv- average e$ect exerted on a particular Indus- Up," Mtedmsa Jomrrt 74dba xe¦las, n:to. Osdnher 1997. t6Br-76?J. try. It m true that 51% of restatlrent owners poldxelo, Atbm O., atsd Radial a Stsbei, "Bm4.ron• predicted no revenue changes, but It is inap• r„an~ 7hhsooo Have Not Hun propriete to rnuchtde the[ no effect would be RestauaM Sales io North Caroliro,"Nod, Coro74m exerted on this industry sins 39% of owners ,vemadreunsa<, s9~, sepiJOCL 199x, 2a4-788. predicted that revenues would fell following Gnvellq Iona c. sad 17esob Zismoermaq "Gaueua a virtual smoking boa. Similarly, not all bars 7tres to Pond HWth Gre 7letomc M Eaanank and taverns are predicted to experience iden- Aoalyfir,• CRS Ripon for Congreq Mud, e,199a. tical effects, Imt the distribution or predicted Iredvaet, Joe, °FaitsutOg ¦ Logit Model with Rassdom- effccts is much tighter (83% predict revelmes lad Dan: 'nee sae tlr Cbcaioe Uae,• .rrwanaan to fall) for this industry. loumd oJsmdadn, 36:t, ApN 1994, 9-16. 1<e, Dwight. R., "EMirorsmentd Fanomirs ¦tsd the Ststi¦I Coat of Smek»g,' ConmrQprory Ibtiry Rl:FE8E~C6.S Lauer, 9:1, lanwry 1991, 8192 Peat Muwick IJP, for the Southern Caldomia Busmen Allen, 7<remiah M., "Environmental Ecastomia sad tht Associetio4 "IM lmpau d the Csnrent and Pro- Sndal Cost of Smtrkiag: A Comment." Cardarspo- roryfbliryGnrcr.l0.l,Iaouary 1992, 101. Pow Smokatg Bam w Rvstauants and Bars in Betker, Gary 5., ¦nd Kevin M. Slurphy, °A S(mpk 7M- Calaomiy" Apnl 19%. ory at Advenlsing as a Good or Bad," QYOnnl3 Sciacn, Jdsn, sad Michael 1. RaniII "Prohibiting Journal a/F.:onwniu, 108:1, November L993, 941- Smakiag to Restwrsati EQeas w Rcstausaas 964• Saks," Amolcan lnwnd o/lleahh Pvnimrton. 123, Barrens, Robert P., Alok K Bohata, and Joe Kerkvliet SeptJOcY. 1998,176-184. "p Randomised Respome Apprwnh to Dichoto- 7remblsy, Card ti., and Ypxor J. Tarnblay, '77se moos Choitt Continaem lytualkn,'Amrricaniow- imPam d Clgatene Arlverluiog on Conmtner Sur- u aJAgn'culmml Erevsamir7. 79:1. 14bnury 199'7. ph's• Pro9t and Soda) Weaate,' Caurrnrporsvy Blanksicin, Andrew, `Entorcement Cbuds Issue of Economic lbliry, l3:l, Janswy1995, 111724. Smoking Ban,- La Ar(gelrr 7hnrr, htetro See. 1Vaiter, Kale, "Design No-Smoking Section," jturaa- September 21, 1998, p. 3. mru Bndiwa, 93;1, lanwq 1, 1994, 1t4. 2085579987