Reaction Paper

4.pdf

Home >Nursing homework help >Reaction Paper

The Effect of Ordinances Requiring Smoke-Free Restaurants on Restaurant Sales

A B S T R A C T StantonA. Glantz, PhD, and Lisa R. A. Smith, BA

Objectives. The cfFcct on restau- rant revenues of kKal ordinances requiring smoke-free rcstiuinmts is an impiirtant consideration for res- t;iur:iteurs themselves and the cities thai depend on sales tax revenues to provide services.

Methods. Data were obtained from the California State Board of Hiiualization and Colorado State Department of Revenue on taxable restaurant sales from I486 (1982 for Aspen) through 1993 for all 15 eities where ordinances were in force, as well as tor 15 similar control commu- nities without smoke-free ordinances during this period. These data were analyzed using multiple regression. including lime and a dummy variable foi whether an ordinance was in loree. Total restaurant sales were analyzed as a fraction of lotal retail sales and restaurant sales in smoke- free eities vs the eomparison cities similar in population, median in- come, and other factors.

Results. Ordinances had no sig- nificant etlcet on ihe fraction of total retail sales ihal went lo restaurants or on Ihe ratio of restaurant sales in eommunities with ordinances eom- parcd wilh those in the matched control ciimniunities.

Conclusions. Smoke-free restau- rant ordinances do noi adversely aflect restaurant sales. {Am J Public Health. |9'M;84:inSl-I085)

Introduction

As the evidence that environmental tobacco smoke endangers nonsmokers'"^ has aeeumulated. more and more eommu- nities have restricted or eliminated smok- ing in public places and workplaces. Several communities have enaeted legisla- tion that requires smoke-free restaurants, thereby protcctingthe public and, particu- larly, restaurant employees'' from the toxie chemicals in seeondhand tobaeeo smoke. Such legislation, however, is not in the interests of the tobacco industry beeause creation of smoke-free restau- rants is a highly visible statement that tobacco use is no longer socially aecept- able.' Thus, tobaeeo companies have sponsored front organizations like the Beverly Hills Restaurant Assoeiation, Res- taurants for a Sensible Voluntary Policy on Smoking. Californians for Fair Busi- ness Policy, and the California Business and Restaurant Alliance to mobilize res- taurants against local smoke-free ordi- nances.*''' This strategy achieved its first success in 1987, when the tobacco indus- try eonvineed the Beverly Hills City Couneil to repeal the first 10()% smoke- free restaurant ordinance in California on the basis of undocumented claims that business dropped 30% beeause of the ordinance.''•^ Because similar predictions for other cities have been published nationally, voieed repeatedly through pub- lie testimony, and regularly printed in news reports, we tested the hypothesis that the passage of a smoke-free restau- rant ordinance is aeeompanied by an immediate significant drop in restaurant sales.'"

This study analyzes sales tax data for the first 15 US eities to enact smoke-free ordinances affecting restaurants. The Cali- fornia eities of Auburn. BellHowcr (which repealed its ordinanee in Mareh 1992),

Beverly Hills (which amended its ordi- nance 4 months after it went into foree), E! Cerrito, Lodi, Martinez, Palo Alto, Paradise. Roscville. Ross. Sacramento, and San Luis Obispo, and the Colorado cities of Aspen, Snowmass Village, and Telluride have had such UK)% smoke-free restaurant ordinances in force long enough to assess their effects. We also examined sales tax data from 15 comparison cities similar to the smoke-free cities in popula- tion, ineomc, smoking prevalenee, and other factors."-" An analysis of restau- rant sales as a fraction of total retail sales, and of restaurant sales in eities witb smoke-free restaurant ordinances eom- pared with those in similar cities that do not have smoke-free ordinances, shows no significant effects on business.

Methods

Data on taxable restaurant sales and total retail sales were obtained from the California State Board of Equalization" and Colorado State Department of Rev- enue'^ from the first quarter of 1986 through the first or second quarter of 1993 (depending on data availability) for the 15 eommunities that had smoke-free restau- rant ordinanees in foree. Included were cities whose ordinances were in force for at least four quarters during this period, plus Beverly Hills and Bellflower, Calif, whose ordinanees were repealed. Data were also obtained for 15 eomparison eommunities where no sueh smoke-free

The authors are with the Institute for Hciilth Policy Studies, Department nf Medicine, al the University of California. San Francisco.

Requests for reprints should he sent to Stant<m A. Glantz, PhD. Division of Cardiol- ogy, University of California, San Francisco, San Francisco, CA 94143.

This paper was accepted February 4, 1994.

Ju!vl'W4,V()l.K4.No. 7 American Journal of Public Health 1081

Glantz and Smith

TABLE 1—Profile of Smoite-Free and Comparison Cities

Smoke-Free and Comparison Cities

Aspen, Colo Vail. Colo

Auburn, Calif Oroviile, Calif

Beverly Hills, Calif Santa Monica, Calif

Bellflower, Calif Lakewood, Calif

El Cerrito, Calif San Pablo, Calif

Lodi, Calif Merced, Calif

Martinez, Calif Pleasant Hill, Calif

Palo Alto. Calif Mountain View, Calif

Paradise, Calif Red Bluff, Calif

Roseville, Calif Chico, Calif

Ross, Calif Tiburon, Calif

Sacramento, Calif Fresno, Calif

San Luis Obtspo, Calif Santa Maria, Calif

Snowmass, Colo Breckenridge, Colo

Telluride, Colo Steamboat Springs, Colo

Population (1989)3

5 049 3 659

10 592 11 960

31 971 86 905

61 815 73 000

22 869 25 158

51 874 56 216

32 038 31 585

55 544 67 460

25 408 12 363

44 685 40 076

2 180 7 532

369 365 354 202

41 958 61 284

1 426 1285

1 292 6 695

(

Inside Urbanized

Area

X X

Geographical

Outside Urbanized Rural

Area Nonfarm

X X

Type of Smoking

Restriction''

100% Some

100% None

100% Some

100% None

100% Some

100% None

100% Some

100% None

100% Some

Median Household

Income (1989)=

37 467 41 211

37 272 16614

54 348 35 997

32 711 44 700

39 538 25 479

30 739 24 727

45 964 46 885

55 333 42 431

22 954 19 474

39 975 19 005

84 414 75 864

28 183 24 923

25 982 29 492

39 107 33 259

31 968 29 363

%of Smokers^

23.5

24,1 23.6

21,8

21.6

22.9

24,1 25.1

22,0

19,7

23.6

24.1 23.6

21,6

25,2 25,1

18.9

23,5

Date Ordinance in Effect

10/85

10/91

4/87-8/87

6/91-3/92

11/91

11/90

3/92

11/91

8/91

9/91

1/90

5/92

8/90

5/89

4/88

=1990 US Census of Population and Housing," ""Some" refers to no more than 60% seating areas for nonsmokers, ^Tobacco Use in California (reported by county) '̂ for California and Behavioral Risk Factor Surveillanoe Study for Cobrado (statewide! 1 9 9 r ̂ "NumbGf of months for which data were available for this study, i=«<"oy.iuo, i ^ . ,

No, Of Months

in Effect̂

ordinance was in force or where no more than 60% seating availabiiity for nonsmok- ers occurred as a part of an existing ordinance (Table 1). Sales data for Aspen and its comparison city were collected from the first quarter of 1982 because Aspen's ordinance was passed in 1985. Data were recorded for "Eating and Drinking Places" and "Total Retail Sales." Published data for restaurant sales and total retail sales in the city of Paradise for the second, third, and fourth quarters of 1990 and in the city of San Luis Obispo for the fourth quarter of 1990 and first quarter of 1991 were corrected as in- structed by the Board of Equalization to

account for late-reported data (written communications from Robert Rossi, June 15,1992, and July 20,1993).

To account for population growth. inflation, and changes in underlying eco- nomic conditions, the fraction (F) of total retail sales at restaurants was computed as follows:

were also compared with sales in compa- rable cities without ordinances as follows;

C =

Restaurant Sales in City with Ordinance

Restatirant Sales in City' without Ordinance

F = Restaurant Sales

Total Retail Sales'

If an ordinance adversely affected restau- rants, this fraction would be expected to drop when the ordinance was in force. Restaurant sales in cities with ordinances

Again, if an ordinance adversely affected sales, this ratio would be expected to drop.

Data were analyzed with linear re- gression'^:

y = bt^ + hf + h,L + hJV,

where y is the dependent variable (F or C), / is time needed to represent the

1082 American Journal of Pubiic Heaith July 1994, Voi. 84, No. 7

Smoke-Free Restaurants

underlying secular trend, and Z. is a dummy variable that indicates whether a smokc-lrce restaurant law is in force. The estimate nf the ctwtficient b, quantifies the annual rate of increase (or decrease) in the dependent variable, _v, each year. The dummy variable /. quantifies the presence ol a smoke-free restaurant ordinance as follows:

0 If No Ordinance W If Ordinance in Force for

1 Month of Quarter V^ If Ordinance in Force for

2 Months of Quarter 1 If Ordinance in Force for

Entire Quarter

The coefficient b,, quantifies the magni- tude of the ctfect of the ordinance on the dependent variable. Because all of the Colorado cities under study are ski cen- ters, the restaurant business is much stronger during the winter tourist season. To iilluw for this effect, the dummy variable H-'was included for the Colorado cities, set to I for Ihe lirsi quarter (the winter tourist season) and 0 otherwise.

Not only were data analyzed for each city separately, but nil the data on restaunmt sales as a percentage of total rcliiil saies for all 15 cities with ordinanees for the entire year period were pooled in a single analysis, including 29 additional dummy variables, to allow for between- city differences in the mean values of the fraction of total retail sales going to restaurants.

The variance inllation factors for each variable were computed to assess multicollincarity, and the Durbin-Watson statistic was computed to test for autocor- relatltin iimong the residuals. The vari- imcc inflation iactors were always well below 2, and the Durbin-Watson statistic never reached statistical significance. A change is considered statistically signifi- cant when/* < .05.

Results

Table 2 summarizes the results for total restaurant sales as a fraction of all retail sales (F), and total restaurant sales in cities with ordinances compared with tliose in the matched comparison cities (( ). The first column in the table is the mc:m v:ilue observed from I98fi (1982 for Aspen) to the second quarter of 1993 to provide n comparison with the magnitude nf the change associated with the ordi- nance.

Smoke-free ordinances generally had no statistieally significant effect on the

TABLE 2—Effect of Smoke-Fre« Restaurant Ordinances on Total Restaurant Saies

City

Aspen Auburn Belifiower Beveriy Hills El Cerrito Lodi Martinez Palo Alto Paradise Roseville Ross Sacramento San Luis Obispo Snowmass Telluride All combined

Aspen Auburn Beltflower Beverly Hills El Cerrito Lodi Martinez Palo Alto Paradise Roseville Ross Sacramento San Luis Obispo Snowmass Telluride All combined

Mean

Effect of Ordinance

Change, bi

Fraction of total retail sale:

24.8 7.5

13,1 12,8 12.7 11.7 10,3 15,8 14.9 7.1

43,5 13,9 12-7 49,2 29,6 18,4

1,1 ± 1,3 1.0 ± 0.5 1,5 ±0,6 0,6 ± 1,2

-0.4 ± 0,7 0.1 ± 0.6 2,9 ± 1,0 0,7 ± 1,1

-1.4 ± 0,8 -0.9 ± 0.4 -3,3 ± 9.1

0.9 ± 0,6 0.2 ± 0,6 6.0 ± 5.7 9.4 ± 4.7

-1,3 ± 1,0

i.F. %

,408 ,092 ,025 ,633 ,637 ,902 ,008 ,520 ,078 ,039 .715 -102 ,764 ,301 ,055 .210

Ratio of saies with comparison city, C

1.12 .44 .50 ,56

1.28 .90 .41

1.69 .71 .68 ,05

1.10 1.12 ,95 ,42 .82

.21 ± .12

.03 ± .02 -.02 ± ,02 - ,06 ± ,04 - 0 0 ± .08 -,01 ± .03

,04 ± ,03 .23 d

- 0 7 = -,02 =

,02 i - . 0 5 - - ,08 : - ,29 :

,08 :

1 ,07 t ,03 t .03 b.01 t ,03 !:0,6 t .20 t ,07

-.04 ± ,03

.106 ,186 ,347 ,171 .998 .742 ,194 ,004 .049 .562 ,196 ,091 .177 -193 ,282 ,166

Modei

,688 ,319 .313 ,033 ,100 ,005 .404 ,115 ,181 ,156 ,132 .102 ,082 ,374 ,197 .611

.153 ,082 ,036 .238 .053 ,270 ,329 ,416 .144 .089 ,302 ,403 ,154 ,584 ,372 ,828

000 ,007 ,008 ,646 .255 ,939 ,007 ,204 .075 111 ,243 ,233 .327 ,006 ,120 ,000

,071 .327 ,621 ,029 ,495 .017 .001 ,001 .132 ,300 ,028 .001 ,113 ,000 ,006 .000

fraction of retail sales that went to restaurants or on total restaurant sales in cities with ordinances compared with those in cities without smoke-free ordi- nances (Table 2 and Figure I). There is marginal evidence that the fraction of total retail sales to restaurants increased in two cities (Bellflower, P = .025; Mar- tinez. P - .008) and decreased in one city (Rosevilie, P = .039). In a comparison of restaurant sales in one city with an ordinance versus one city without an ordinanee, sales increased in one city {Palo Alto. P = .(X)4) and decreased in another (Paradise, P = .049). The iaek of consistent response suggests that these results may simply reflect random varia- tion, given the large number of P values that were computed. Analysis of ail the data in pooled regressions did not detect significant changes in the percentage of retail sales or sales in cities with smoke- free ordinances compared with those in cities without ordinances.

Beverly Hills is a particularly impor- tant ease because it has been used by the tobacco industry to support the claim that smoke-frcc restaurant ordinances are as- sociated with a 30% drop in business (Figure 2). However, data reveal that no such drop in sales occurred upon enact- ment, and that no inerease in sales followed repeal 4 months later. Likewise, despite the fact that the Bcllflower ordi- nance was repealed because of claims that business dropped, the ordinance was actually associated with a marginally sig- nificant (P - ,025) increase in business.

Discussion

This is the first comprehensive study that examines taxable sales data to deter- mine the eoinomic impact of smoke-free restaurant ordinances on restaurant sales. Using data from the California State Board of Equalization and the Colorado State Department of Revenue for pur-

Juiyl994,VoLHNo.7 American Journai of Public Heaith 1083

Glaniz and Smith

iin 'HO ini i»i mi

ino Itt' i«i iMi

INI in; itti it9< oee its' <me

Note. The quarters in which 100% smoke-free ordinances were in effect are represented by solid circles.

FIGURE 1—Restaurant sales as a percentage of total retail sales for the 15 communities included In the study.

t991 19921986

Note. Period of smoke-free ordinance Is indicated by the solid triangles.

FIGURE 2—The 100% smoke-free restaurant ordinance in force in Beverly Hills did not reduce saies by 30% (dashed iine with "TI [tobacco industry] claim"), as the tobacco industry had suggested; rather, it had no significant effect on saies.

poses of paying sales taxes has several advantages. First, the numbers rctlect alt restaurant sales in a community, not just those of a small sample of restaurants. Second, the numbers arc objective; they were collected through consistent meth- ods by agencies with no interest in the effects of smoking restrictions on restau- rant sales. Third, sales tax data can be expected to be reasonably accurate since it is a crime lo lie in reporting the figures.

The communities studied in the report are different from each other and represent a cross-section of communiiics that might enact legislation controlling smoking in restaurants; Auburn is a small Sierra foothills community: Beverly Hills is a well-to-do urban city; Bcliflowcr is a middlc-c!;issbcdnx>m community; FJCcr- rito and Martinez lie wilhin highly indus- trial areas; Ltidi is a rural agricultural center; Palo Alto is a large suburban university community; Paradise is a small, scmiagricultural community; Sacramento is a large city and the state capital; Sun Luis Obispo is a college town; Roscvillc represents a semirural bcdrixim commu-

1084 American Journal of Public Health July 1994. Vol. 84, No. 7

Smoke-Free Restaurants

nity; Ross is a small affluent San Francisa) Bay community; and the three Colorado cities arc mountainous, tourist resort areas. The fact Ihat there were no adverse eifects on business in these communities supports the conclusion that the results generalize broadly. Further, these 15 cities represent every ciiy that has passed smoke-free ordinances that have been in cfTect long enough to study.

This study covers a significant period of time. It is important to take into account long-term (secular) trends as well as the quartcr-by-quarter random varia- tion and short-term economic changes. We avoided short-term analyses because it is generally possible to reach any conclusion desired by seleetively picking the "corrccl" two quarters for analysis.'^ To avoid such hiases and increase tbe power of the statistical analysis to detect an effect of the ordinances, we used data for a 7-ycar period (12 years for Aspen Lind Vail). This length of time allowed us to obtain gixKi estimates of secular trends before evaluating any cllecls of the ordi- nances.

A common concern is raised about the possibility that patrons will dine in adjacent communities without such restric- tions. Our data address this concern because the cities examined in this study are not isolated communities. Auburn, Lodi. Martinez. Paradise. Roscville. and San Luis Obispo, while not in large urban centers, are all surrounded by unincorpo- rated areas that contain restaurants. Bev- erly Hills and Bellflower and their com- parison cities, Santa Monica and Lakewood, are all located in Los Angeles County, a major metropolitan area in which all communities directly abut other comnuinilies. El Ccrrito. Palo Alto. Ross, and their comparison cities all lie within I he San Francisai Bay region. Sacra- mento and its comparison city, Fresno, both large urban eenters, face competi- tion from several neighboring communi- ties. Although the skiing communities of Aspen, Telluride, and Snowmass Village are relatively secluded, other resort towns nearby that allow smoking would repre- sent viable tourist alternatives to these smoke-free cities. If people were leaving

these cities to dine in neighboring cities, our analysis would have detected it.

Another area of concern is the effect on b:irs since smoking and drinking are thought to go together. Revenues from bars and "full-service" restaurants are included in the sales tax data we used. The ordinances examined in this study contain different provisions governing bars independently and bars in relation to restaurants. Had there been a significant effect on sales in such restaurants, our analysis would have detected it. Further- more, an analysis of individual classes of restaurants (based on whether they sell different types of alcohol) for four cities in Calif()rnia previously showed no effect when full-service restaurants were ana- lyzed separately."*

Finally, the fact that the ordinances in Beverly Hills and Beilflower were repealed adds to the strength of our conclusions. Had the ordinances affected sales negatively, we would have expected to see an increase in sales following repeal. However, there was no increase in Beverly Hills, and sales dropped in Bell- flower after the ordinance was repealed. Thus, legislators and government officials can enact such health and safety require- ments to protect patrons and employees in restaurants from the toxins in seeond- hand tobacco smoke without the fear of adverse economic a>nsequences. •

Acknowledgments This research was supported by funds provided by the Cigarette and Tobaccti Surtax Fund of ihe State of California through the Tobacco- Reialed Disease Research Program of the University ol California (award 1RTS20).

References 1. Centers for Disease Control. Office on

Smoking :ind Heaith. The Health Comc- quences of inmtiintary Smoking: A Report of the Surgeon Cenerat. RtKkville. Md: US Dept of Health and Human Services, Public Health Service: 1986. DHHS publi- cation CDC S7-S398.

2. National Research Council. Committee on Passive Smolcing. Envirtmmentat Tobaeco Smoke: Measuiing Ex[x?\uri's and Assessing Heatth F.ffects. Washington, DC: National Academy Press; 1986.

3. Respirautry Health Effects of Passive Smok-

ing: t,img Cancer ami Other Disorders. Washington. DC: EEivironmental Protec- tion Agency, OfTice of Health and Environ- mental Assessment; 1992. Publication EPA/fiOO/6-911/(KKiF.

4. Giantz SA. Parmley WW. Passive smoking and heart disease: epidemioiogy. physiol- ogy, and biochemistry. Circulanon. 1991;83: t-12.

5. Steeniand K. Passive smoking and the risk ofheartdisea.se. JAMA. 1992:267:94-99.

6. Siegel M. Involuntary smoking in the restaurant workplace: a review of em- pioyee exposure and health eflects. JAMA. 1993;270:49(M93.

7. Pithltv Attitiuies toward Cigaretle Smoking and the Tobacco huluslr\: New York. NY: Roper Organization; May 1978.

8. Samuels B. Giantz S. The poiitics of kxal tobacco control. JAMA. I991;266:2110- 2117.

9. Traynor MP, Begay ME, Giantz SA. The new tobacco industry strategy to prevent local tobacco control. JAMA. 1993;270:479- 486.

10. Press Alert and Fact Sheet: San Francis- co's Hospitality Industry Faces Smoking Ban. San Francisco. Calif: Goiden Gate Restaurant Association; February 17.1992.

11. United .Slates Census of Population and Housing 1990: Stmmuiry Tape tile 3A— Btireaii ofthe Census. Washington, DC: US Bureau ofthe Census; 1990.

12. Pierce JP. Faritas A, Evans N, et ai. Tobacco Vsc in Califomia 1992. A FtKiiS on Preventing Uptake in Adolescents. Sacra- mento. Calif: Califomia Department of Health Services: 19<)3.

1.̂ . Behaviomt Risk FactorSurxeithmeeSystem— Summary Prc\'alence Report. Centers for Disease Control and Prevention. National Center for Chronic Disease Prevention and Health Promotion: 1991.

14. Taxable Sales in California (Sale.i and Use TtLx). Sacramento, Calif: State of Califor- nia. Board of Equaiization: Tabie 5 (quar- terly reports 19S6-1992).

15. Sates Tax Statistics for Cotinties and Selected Cities and tiitailed Industry' ReporL'i by Coiinn- {January l9ii2~June I9Q2). Denver, Colo: State of Colorado, Department of Revenue.

Ih. Mendenhall W. Sincich T.A Second Course in Business Stati.itics: Regression Analysis. San Francisco, Caiif: Deiien Pubiishing Company: 1989: chap 9.

17. Impticts of Local 100% Smoking Bans on Restaurants and Related Businesses. Los Angeies. Calif: California Business and Restaurant Aliiance: i992.

18. Giantz SA. Smith LRA. Tfte Effect of Ordinances Requiring Smoke-free Restau- rant.\ on Restaurant Sates in Califomia. San Francisco, Calif: University of California Institute for Health Policy Studies Mono- graph Series; 1992.

July 1994, Vol. 84, No. 7 American Journal of Public Heaith 1085