Tobacco use and asking prices of used cars: prevalence, costs, and new opportunities for changing smoking behavior

BioMed CentralTobacco Induced Diseases

ss

Open AcceResearchTobacco use and asking prices of used cars prevalence costs and new opportunities for changing smoking behaviorGeorg E Matt1 Romina Romero2 Debbie S Ma1 Penelope JE Quintana3 Melbourne F Hovell3 Michael Donohue4 Karen Messer4 Simon Salem1 Mauricio Aguilar1 Justin Boland1 Jennifer Cullimore1 Marissa Crane1 Jonathan Junker1 Peter Tassinario1 Vera Timmermann1 Kristen Wong1 and Dale Chatfield1

Address 1Department of Psychology San Diego State University San Diego USA 2Joint Doctoral Program in Health Behavior Research San Diego State University and University of California San Diego USA 3Graduate School of Public Health San Diego State University San Diego USA and 4Moores Cancer Center University of California San Diego USA

Email Georg E Matt - gmattsciencessdsuedu Romina Romero - rromeroprojectssdsuedu Debbie S Ma - debbiesmagmailcom Penelope JE Quintana - jquintanmailsdsuedu Melbourne F Hovell - mhovellprojectssdsuedu Michael Donohue - mdonohueucsdedu Karen Messer - kmesserucsdedu Simon Salem - simonsalemhotmailcom Mauricio Aguilar - maguilarprojectssdsuedu Justin Boland - bolandjustingmailcom Jennifer Cullimore - jennijc55yahoocom Marissa Crane - marissacgmailcom Jonathan Junker - jjunker21hotmailcom Peter Tassinario - petetassinariogmailcom Vera Timmermann - veratimmermanngmxde Kristen Wong - kristenjanelle19yahoocom Dale Chatfield - dchatfieldsciencessdsuedu

Corresponding author

AbstractSecondhand smoke (SHS) causes premature death and disease in children and adults and thescientific evidence indicates that there is no risk-free level of exposure to SHS Smoking tobacco ina car can pollute the microenvironment of the car with residual SHS leaving telltale signs topotential buyers (eg odor used ash tray) This study examined (a) the proportion of used carssold in the private party market that may be polluted with tobacco smoke and (b) whether askingprices of smoker and nonsmoker cars differed for cars of otherwise equivalent value A randomsample of 1642 private party sellers were interviewed by telephone and content analyses of printadvertisements were conducted Findings indicate that 22 of used cars were advertised bysmokers or had been smoked in during the previous year Among nonsmokers 94 did not allowsmoking in their car during the past year Only 33 of smokers had the same restrictions Thesmoking status of the seller and tobacco use in the car were significantly (p lt 01) associated withthe asking price independent of a cars Kelley Blue Book value (KBB) Used nonsmoker cars wereoffered at a considerable premium above their KBB value (gt11) and above comparable smokercars (7ndash9) These findings suggest that community preferences are affecting the value of smoke-free cars New directions for research tobacco control policies and health education are discussedto further reduce smoking behavior to help consumers make informed purchasing decisions andto protect nonsmokers from SHS exposure

Published 31 July 2008

Tobacco Induced Diseases 2008 42 doi1011861617-9625-4-2

Received 19 June 2008Accepted 31 July 2008

This article is available from httpwwwtobaccoinduceddiseasescomcontent412

copy 2008 Matt et al licensee BioMed Central Ltd This is an Open Access article distributed under the terms of the Creative Commons Attribution License (httpcreativecommonsorglicensesby20) which permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Page 1 of 8(page number not for citation purposes)

Tobacco Induced Diseases 2008 42 httpwwwtobaccoinduceddiseasescomcontent412

BackgroundSecondhand smoke (SHS) contains more than 50 knownhuman carcinogens and has recently been classified astoxic air contaminant [12] SHS causes premature deathand disease in children and adults and the scientific evi-dence indicates that there is no risk-free level of exposureto SHS [1-3] A growing number of local communities andstates in the US and in countries throughout the worldare therefore adopting stricter policies to curb tobacco usein general and to reduce exposure to SHS exposure in par-ticular [4-6]

The introduction of stricter tobacco control policies isoften accompanied by health education campaigns aboutthe harmful effects of tobacco use on smokers and vulner-able groups of nonsmokers [7-10] The ultimate goals ofthese efforts are to improve public health through chang-ing personal smoking behavior community standardsand attitudes toward tobacco use and SHS [281112]According to the 2006 National Health Interview Survey[13] however steady declines in smoking rates since the1960s appear to have stalled remaining unchanged at21 since 2004 Thus tobacco control efforts remain ahigh public health priority requiring renewed efforts tofurther reduce tobacco use and SHS exposure

The present study examined the prevalence of tobacco useand asking prices of used cars in a community that hasexperienced extensive public health education campaignssince passing comprehensive statewide tobacco controllegislation in 1988 [14] We hypothesized that in such acommunity smoke-free cars would be offered at a pre-mium compared to smoker cars controlling for other fac-tors influencing the value of a car If this is the case futureresearch may be warranted to better understand the effectsof tobacco use on the value of personal property and howsuch consumer preferences could help further reducetobacco use and SHS exposure

Tobacco Use in CarsCompared to research on smoking restrictions in theworkplace restaurants and at home relatively little isknown about smoking restrictions in cars Existingresearch suggests that smoking restrictions in cars are lesscommon than those at home In California two out ofthree family cars had a complete smoking ban in 1996and 1998 [1516] compared to almost four out of fivehomes with complete smoking bans Among smokershowever only 29 had a complete car smoking ban and43 had a complete home smoking ban Similar patternswere observed in urban and rural settings of the US out-side of California Halterman et al [17] found that amongurban households with smokers and children sufferingfrom asthma only 64 had a complete ban on smokingin the home and 49 in the car Kegler amp Halinka Malcoe

[18] examined low-income families of children in ruralOklahoma They found that 49 of Native Americanhouseholds and 43 of Caucasian households bannedsmoking in the home but only 35 and 40 respec-tively banned smoking in the car A deviation from thispattern was reported by King et al [19] among African-Americans who found a higher percentage of nonsmokershad car than home smoking bans (84 vs 74) Amongsmokers however only 17 and 21 had similar bans

Residual SHS Contamination of Used CarsWhen tobacco is smoked in the confined environment ofa car tobacco smoke pollutants can reach extremely highlevels [20] Volatile SHS components absorb into surfaceswithin minutes of emission contaminating objects withwhich they come in contact Subsequently this residualSHS (also known as aged SHS or third-hand smoke[2122]) is re-emitted into the air over days weeks andmonths accumulates in dust and deposits on surfaces[23-27] creating a route of exposure for drivers and pas-sengers of smoker cars in the absence of concurrent activesmoking

Unlike mechanical or electronic defects detecting thesigns of previous tobacco use in a car often requires littletechnical expertise from a potential buyer Routinetobacco use leaves many telltale signs to prospective buy-ers Foremost is a distinct odor caused by the re-emissionof SHS contaminants from surfaces and dust that werepolluted during active smoking Matt et al [28] haveshown that SHS odor and ash marks are significantly asso-ciated with the residual contamination of dust surfacesand the air in cars It is this odor and other visible signsthat can signal to a potential buyer that a car has beensmoked in and that can be difficult and expensive toremove through cleaning or repairs

Private Party Sales of Used Cars in the USThe used car market provides a particularly interestingopportunity to examine the value of a smoke-free per-sonal environment because a large and diverse cross-sec-tion of the general population in the US sells and buysused cars Moreover a substantial portion of personalincome is spent on the purchase and maintenance of carsIn 2005 441 million used and 169 million new carswere sold in the US [2930] These transactions totaled$780 billion $367 billion of which were accounted for byused car sales Approximately 30 of used cars were soldby private parties [30]

Although asking prices for used cars are often informed bypublished pricing guidelines (eg Kelley Blue Book) sell-ers can advertise their cars for any price buyers can offerany price and the eventual sales price is subject to thelocal market forces of supply and demand Sellers typi-



cally begin gauging an asking price by establishing thestandard value of their used car based on its make modelage mileage and condition In addition sellers oftenlook up the asking prices of similar cars currently offeredfor sale by consulting the classified ads of local newspa-pers and used cars offered by dealers Sellers then applyadditional adjustments for factors believed to affect thevalue of a car in the community where it is sold that werenot ndash or not sufficiently ndash included in the standard modelThese adjustments may increase (eg chrome wheelssmoke-free car) or decrease the value of a used car (eg noair conditioning smoker car) Further price adjustmentsfollow if a car fails to sell and the seller is unable to reme-diate problems preventing a sale

The Behavioral Ecological Model (BEM) provides a theo-retical framework of the association between the valuesand norms of communities and the behaviors and prefer-ences of individuals Briefly put BEM postulates that cul-ture-wide social contingencies influence health practicesat both the individual and the community levels Changesin norms (eg tobacco use in the presence of nonsmok-ers) can initiate a cascade of social contingencies from thepopulation to the individual levels that affect the strengthof a given cultural characteristic (eg tobacco tolerance)[3132] Consequently changes in individual healthbehavior (eg car smoking ban) can be initiated bychanging social and economic contingencies at the popu-lation (eg smoker cars are worth less in the private partymarket) and individual levels (eg family members com-plain about stale tobacco odor in car) This study offeredan opportunity to explore hypotheses about emergingsocial contingencies with respect to tobacco We reasonedthat cultural changes regarding tobacco use should lead tolower prices for cars offered by smokers than equivalent

cars offered by nonsmokers This would create new socialand economic contingencies affecting tobacco use andSHS exposure among seller buyers and passengers

MethodsParticipantsThe target population was private party sellers (age ge 18years) of used cars who advertised between January 2005and April 2006 in the San Diego (USA) print editions ofthe Auto Trader magazine a popular weekly publicationof classified ads for used cars and with phone numbers inthe 619 and 858 area codes Approximately 3000 privateparty advertisements were published per week in the tar-get area codes

A random sample of 100 pages was drawn each weekacross all Auto Trader issues for different types of automo-biles (ie domestic Asian European sport utility mod-els newer and older models of trucks and vans) using arandom number generator The selected pages were sortedin the order in which the random numbers were gener-ated We called all eligible sellers on a page in the order ofthe sorted pages until we had recruited the target numberof smokers each week (1 2 or 3) Of the 2590 sellers whowere screened by phone for a study about the SHS con-tamination of cars 2081 (80) reported their smokingstatus 1667 (64) reported the smoking status of thecar and 1642 (63) reported both For analyses of ask-ing prices (N = 1425) we excluded cars built before 1989because car values could not be determined reliably Table1 provides information about asking price differences inthe print advertisements mileage age and make of carsby smoking status of the car and the seller The Institu-tional Review Board at San Diego State Universityapproved the research protocol

Table 1 Asking price Kelley Blue Book value mileage age and make of used cars and percentage of used cars sold by a smoker and cars in which cigarettes have been smoked

All Smoke-Free Car Smoker Car

NonsmokerSeller

SmokerSeller

NonsmokerSeller

SmokerSeller

Sample size 1642 1274 95 77 196 of Used Cars 776 58 47 119Asking Pricea ($) 7636 8114 7684 5684 5602KBB Valueab ($) 6906 7213 7188 5330 5734Mileagea 69339 68019 72099 87400 70709Agea (Years) 60 59 60 60 67Make

American 50 51 40 48 48 Japanese 33 32 39 43 34 European 17 17 20 9 18

a geometric meansb KBB standard value of used car for private party sales based on Kelley Blue Book



MeasuresTelephone interviewSellers were identified as smokers if they reported smok-ing cigarettes every day or on some days during the pastyear Cars were identified as smoker cars if the sellerreported that one or more cigarettes had been smoked inthe car during the past year

Content analysis of print advertisementsData about the asking price year make model mileagecondition and special features were obtained from theprinted advertisement of the car or during the phoneinterview For each car the widely used KBB value wasdetermined given the information provided in the printedadvertisement using the online valuation calculator [33]If mileage information was omitted (15 of cars) theKBB value was determined assuming 15000 miles peryear While year make and mileage can be easily deter-mined condition appearance and special features oftenrequired judgment and interpretation The condition of acar was coded as good unless the seller listed specificnegative or positive characteristics in which case the con-dition was downgraded to fair or poor or upgraded toexcellent No KBB values are available for cars in poorcondition (eg salvage title major mechanical prob-lems) When this was the case a car was excluded fromanalyses (lt1) To examine the reliability of the KBBvalue determination by coders of this study intraclass cor-relations were calculated based on a random subsampleof 50 cars coded by each of five coders The ICC for indi-vidual ratings was 093 indicating that KBB values weredetermined with good reliability

Statistical AnalysesStatistical analyses were conducted using Stata version 92[34] Asking price and KBB values were log-transformed tonormalize the model residuals This was confirmedthrough graphical and quantitative analyses Because carssold by smokers and nonsmokers may differ in character-istics other than smoking status it was important to adjustfor these factors Multiple regression analyses were con-ducted in which the log-transformed asking price was theresponse variable and the log-transformed mean-centeredKBB value and its quadratic and cubic terms were enteredas covariates to statistically control for differences in theasking prices given the KBB valuation We then investi-gated whether sellers may have applied different weightsto the components of the standard valuation Thus weadded mileage year make and condition of a car as addi-tional covariates and retained covariates in the model α lt005 Finally we entered dummy variables for smokingstatus of the seller smoking status of the car and interac-tion terms of all covariates and explanatory variables Nointeraction effects were statistically significant (α = 005)We examined the robustness of model estimates through

sensitivity analyses in which different transformations ofasking prices and KBB values robust variance estimates[35] bootstrapped regression coefficients and alternativeregression models (quantile and robust regression) wereexplored Throughout these analyses overall model fitstatistical significance for smoking status of sellers andcars and effect sizes remained stable The reported find-ings are based on log-transformed variables and modelsspecifications outlined above

We derived maximum likelihood estimates based on oursampling design to estimate the proportions of smokersand smoker cars in the target population Briefly we mod-eled the number of smokers who were not recruited andthe number of smokers who were recruited using the neg-ative binomial distribution Variance estimates werederived via the delta method [36]

ResultsSmoking Status of Sellers and CarsOverall 177 (95 Confidence Interval 159 196) ofsellers reported themselves being smokers 166 (148184) of used cars had reportedly been smoked in and224 (204 244) of cars had either been smoked in orwere being sold by a smoker Among the nonsmokers57 (46 71) had allowed smoking in their car duringthe previous year Among smokers 674 (616 726)had allowed smoking in their car Table 1 provides addi-tional detail on the smoking status of sellers and theircars

Smoking Behavior and Asking PriceLinear regression models showed that the KBB valueaccounted for 84 of the variance in asking price (p lt0001) indicating that the asking prices closely matchedthe prices suggested by the KBB valuation model In addi-tion there were statistically significant quadratic andcubic trends (p lt 001 accounting for additional 1 vari-ance) Further investigations of the nonlinear associationsindicated that for cars with low KBB values (lt$2500) sell-ers raised the asking price comparatively more than formore expensive cars Sellers made another adjustmentbased on the make of a car that went beyond the standardKBB valuation model accounting for an additional 1 ofthe variance in asking price (p lt 0001)

Controlling for KBB value and make of car smoking sta-tus of the seller and the car accounted for a significant pro-portion of variance (F(2 1388) = 637 p = 0002)Because the two variables were highly correlated (r = 062p lt 001) neither accounted for variance independent ofthe other when entered jointly in the model When exam-ined in separate models smoking status of the car(t(1391) = 326 p = 0001) and of the seller (t(1389) =309 p = 0002) each accounted for significant propor-



tions of variance in asking price (approximately 01)independent of KBB value and make of car We addressthe practical significance of this effect below

Figure 1 shows the association between KBB value andasking price for cars sold by smokers in which tobacco hadbeen smoked and cars sold by nonsmokers in which notobacco was smoked Also shown is the reference lineindicating cars offered at their standard KBB value (ieasking price = KBB value) The figure shows that the askingprices of smoker cars were consistently lower than thoseof nonsmoker cars of equivalent KBB value and were onaverage within plusmn 4ndash5 of their standard KBB value Incontrast nonsmoker cars were consistently offered at a10ndash13 premium above their KBB value

Table 2 provides model estimates of asking prices basedon smoking status of car and seller for cars at different KBBvalues controlling for make model age mileage andcondition of car These estimates reveal that cars sold bysmokers and cars that had been smoked in were offered ata significantly lower price than equivalent cars offered bynonsmokers and in which no smoking was reportedBecause the regression models used log-transformed ask-ing prices the observed effects of smoking status of a carand a seller can be interpreted as differences in the per-centage of asking price That is given a particular KBBvalue and model of car the value of a car decreased by77 if it had been smoked in compared to a car that wassmoke-free The value decreased by 75 if a car was soldby a smoker compared to an equivalent car sold by a non-

smoker Finally a car decreased in value by 90 if it wassold by a smoker who allowed smoking in the car com-pared to an equivalent car sold by a nonsmoker who pro-hibited smoking

Table 2 also shows the percentage premium over KBBvalue that sellers asked for For a car of median KBB value(ie $7363) nonsmokers who had not smoked in theircar asked for a 135 premium over the KBB value Thiscompared to a 33 premium for an equivalent car soldby a smoker whose car had been smoked in

DiscussionTo the best of our knowledge this is the first study toexamine the association between tobacco use in cars andtheir asking price in the private party used car market Ourfindings show that one out of five used cars for sale in theSan Diego (CA) metropolitan area were offered by smok-ers or had been smoked in during the previous year Whilenine out of ten nonsmokers reported that no cigaretteshad been smoked in their car during the past year onlyone in three smokers reported to have had such a restric-tion Finally used nonsmoker cars were offered at a con-siderable premium above their KBB value and abovecomparable smoker cars In the following we will brieflydiscuss limitations of this study and implications fortobacco control and consumer protection in the used carmarket

LimitationsFindings from this study are based on a cross-sectionalsurvey and self-reported information Because of thesocial undesirability of smoking behavior and its likelynegative impact on the sale of a car we suspect that somemisreporting took place such that the proportion ofsmoking sellers and smoker cars are likely to be higherthan we determined

While plausible a preference for smoke-free personalenvironments is not the only possible explanation for theobserved differences in asking price The non-experimen-tal nature of this study raises the possibility that theobserved differences may be due to additional variablesthat affect both smoking behavior and the value of the carTo address this issue we included in our analyses as a cov-ariate the standard KBB value of a vehicle It is possiblehowever that smoker cars overall are in worse conditionthan nonsmoker cars and are so in a way that was notmeasured by our rating of the condition and appearanceof a car based on its print advertisement It is also possiblethat potential buyers use the smoking-status of a seller orthe tobacco odor of the car as a proxy of poor mainte-nance to negotiate a lower price and are not at all con-cerned about potential health effects or odor nuisanceOur study was also not able to examine whether smoking

Association between asking price and Kelley Blue Book value of used cars sold by smokers in which cigarettes have been smoked and by nonsmokers in which no cigarettes have been smoked (LOWESS fit lines)Figure 1Association between asking price and Kelley Blue Book value of used cars sold by smokers in which cig-arettes have been smoked and by nonsmokers in which no cigarettes have been smoked (LOWESS fit lines) Also reported is a reference line for cars offered at the KBB value



and nonsmoking buyers have similar preferences andhow smoking status of the seller and tobacco odor of thecar are considered in the purchasing decision

This study focused on asking price and does not allowconclusions about actual sales prices Although we expectthat the sales price differences will exceed asking price dif-ferences this should be demonstrated in future researchThis study cannot provide explanations for the processesleading to the observed asking price differences We alsocannot rule out that smokers differed from nonsmokers inhow they discounted asking prices for damages other thansmoking or special features of their cars

Tobacco Use in Private CarsOur findings support existing research that smokingrestrictions for private cars are less common than forhomes In 2005 53 of Californias smokers reported liv-ing in a smoke-free home[37] but only 32 of the smok-ers in the present study reported that no cigarettes weresmoked in the their car during the previous year The rel-atively low prevalence of smoking bans in the cars ofsmokers raises the question whether cars may havebecome sanctuaries for smokers to light a cigarette whiledriving to and from places with smoking restrictionsWhile this smoking pattern protects nonsmokers whensmokers drive alone recent research [28] has demon-strated that it leads to the pollution of cars with residualSHS long after cigarettes have been smoked Futureresearch is needed to examine smoking behavior andrestrictions in cars and how they may be influenced byrestrictions in a smokers home workplace and commu-nity

Tobacco Use and the Resale Value of CarsOur findings indicate that smoker cars have lower askingprices than comparable nonsmokers cars From the per-spective of the BEM such asking price differences are to beexpected in communities that value smoke-free personalenvironments creating a greater demand for smoke-freecars such that sellers can ask for a premium over the carsKBB value and over comparable smoker cars In contrastasking price differences would not be expected in commu-nities tolerant of or indifferent to SHS exposure andtobacco odor in cars and their effects on health and thevalue of a car Consistent with the BEM asking price dif-ferences contribute to a community-wide culture thatencourages car smoking bans and discourages overalltobacco use Thus even if the motivation for discountingasking prices is not a perceived health risk the financialconsequences may contribute to establishing community-wide norms for not smoking in cars This adds one moresetting in which smoking may become socially unaccept-able

This study was conducted in a community in SouthernCalifornia that has been highly sensitized to the healtheffects of tobacco use in general and SHS exposure in par-ticular over 20 years of public health education efforts[38] These efforts have contributed to reducing smokingprevalence in California from 26 to 14 between 1984and 2005 [37] and in 2002 93 of nonsmokers and83 of smokers agreed that any exposure to SHS can beharmful to your health [1239] Our findings support thehypothesis that changes in collective values smokingbehavior and attitudes toward SHS have influenced themarket place affecting the value of personal property and

Table 2 Adjusted asking prices for cars of smokers and nonsmokers at different KBB values

KBB Valuea Adjusted Asking Price in US $b

Decile US $ NonsmokerCar amp Seller

SmokerSeller

SmokerCar

SmokerCar amp Seller

1 2080 2893 2677 2670 26342 3223 4031 3731 3720 36703 4354 5187 4803 4787 47224 5674 6563 6081 6058 59765 7363 8358 7747 7715 76096 9186 10324 9574 9531 93997 11818 13200 12247 12189 120188 14900 16600 15409 15333 151139 20515 22823 21200 21093 20779

Difference Asking Price vsNonsmoker Car amp Seller

Referent -70 -77 -90

Difference Asking Price vs KBBat median value ($7363)

+135 +52 +48 +33

aKelley Blue Book (KBB) values in US currency for 1st to 9th decilesbAsking prices were adjusted for linear quadratic and cubic effects of KBB values and make of car



shaping purchasing decisions Thus monetary value ofsmoke-free environments in the market place may pro-vide a useful outcome to evaluate long-term effects oftobacco control efforts at the level of communitiesBecause the health outcome of exposure to residual SHSin a car are not well understood the observed differencesin asking price may be a sign that the concerns of non-smokers reach beyond recognized health risks and includeconcerns about the depreciation of personal property andquality of life This is consistent with the increasing publicdebate about drifting smoke in multi-unit housing [40]and recent changes in local ordinances that led to smok-ing bans in city parks on playgrounds beaches and side-walks

Future Tobacco Control Efforts and the Used Car MarketFrom the perspective of tobacco control policies theobserved asking price differences are not only importantoutcomes of successful health promotion campaigns butmay suggest new strategies to further reduce tobacco useand SHS exposure For instance future tobacco controlefforts could educate consumers about the effects oftobacco use on the value of used cars Our findings suggestthat many sellers are already cognizant of this effect How-ever it is unclear whether this is equally known amongsmokers and nonsmokers and across different educationsocio-economic and ethnic groups Health educationcampaigns could help motivate smokers to smoke less orquit altogether Such campaigns could also empower con-sumers to assert their interests in smoke-free environ-ments and in obtaining an appropriate discount if theychoose to tolerate a smoker environment Finally con-sumer education campaigns would provide incentives toprivate sellers and dealers to advertise the smoking statusof cars allowing consumers to make informed purchasingdecisions

A more drastic approach would involve a change in thevaluation model used by private parties car dealersbanks and insurance companies to value cars Although itcould be argued that such a step is unnecessary given theexisting market response it is worthwhile to consider thispotential path as an explicit recognition of how a commu-nity values a car that may affect the health and drivingexperience of drivers and passengers

Signs of tobacco use (eg odor burn marks) are currentlyimplicitly included among many factors that diminish thevalue of a car via their impact on appearance and overallcondition Our findings suggest that the smoking historyof a car affects its value as much as many prominent fea-tures of a car that the KBB valuation models does con-sider In November 2006 for instance a 2000 ToyotaCamry LE four-door sedan 4-cylinder engine automatictransmission 77000 miles standard equipment and in

good condition was valued for private-party sale in theSan Diego market at $7695 If this car had been offeredby a smoker and had been smoked in the asking pricewould have been about $700 lower (ie 9) For this carto loose $700 in KBB value the car would have to miss allof the following standard features air conditioningpower steering power windows power door locks cruisecontrol and the dual front airbags Admittedly a car miss-ing all of these features would probably sell in the marketplace for a much larger discount Still ignoring tobaccouse in the valuation model of used cars disregards a fea-ture of an automobile to which at least some communitiesappear to have assigned a considerable monetary valueFrom the perspective of the BEM such a recognitionwould introduce an explicit incentive that may trigger fur-ther changes on the community and individual levels toreduce tobacco use and SHS exposure

Authors contributionsGEM conceived and designed the study performed thestatistical analyses and drafted the manuscript RR andDSM contributed to the design and coordinated the studyparticipated in the data collection and data analysis PJEQand MFH contributed the design of the study and draftedthe manuscript MD and KM contributed to the data anal-ysis and drafted the manuscript SS MA JB JC MC JJ PTVT and KW contributed to the design the study and par-ticipated in the data collection DC contributed to thedesign of the study All authors read and approved thefinal manuscript

AcknowledgementsThis study was supported by grant 13-IT0042 from the California Tobacco Related Disease Research Program (TRDRP) and intramural sup-port from the Center for Behavioral Epidemiology and Community Health at San Diego State University

References1 State of California Air Resource Board Technical support docu-

ment for the Proposed identification of environmentaltobacco smoke as a toxic air contaminant Part A [httpwwwarbcagovregactets2006ets2006htm] Retrieved January 132006

2 US Surgeon General The health consequences of involuntary exposureto tobacco smoke A report of the Surgeon General Atlanta GA USDepartment of Health and Human Services Centers for DiseaseControl and Prevention Coordinating Center for Health PromotionNational Center for Chronic Disease Prevention and Health Promo-tion Office on Smoking and Health 2006

3 Repace J Exposure to secondhand smoke In Exposure analysisEdited by Ott WR Steinemann AC Wallace LA Boca Raton Tayloramp Francis 2007201-235

4 Americans for Nonsmokers Rights Smoke-free lists maps anddata [httpwwwno-smokeorggoingsmokefreephpdp=d13|p140maps] Retrieved June 11 2007 (2007 422007)

5 Fong GT Cummings KM Shopland DR Building the evidencebase for effective tobacco control policies The internationaltobacco control policy evaluation project (the ITC project)Tob Control 2006 15(Suppl 3)1-2

6 Semple S Creely KS Naji A Miller BG Ayres JG Secondhandsmoke levels in scottish pubs The effect of smoke-free legis-lation Tobacco Control 2007 16(2)127-132



Publish with BioMed Central and every scientist can read your work free of charge

BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime

Sir Paul Nurse Cancer Research UK

Your research papers will be

available free of charge to the entire biomedical community

peer reviewed and published immediately upon acceptance

cited in PubMed and archived on PubMed Central

yours mdash you keep the copyright

Submit your manuscript herehttpwwwbiomedcentralcominfopublishing_advasp

BioMedcentral

7 Gilpin EA Emery SL Farkas AL Distefan JM White MM Pierce JP TheCalifornia tobacco control program A decade of progress 1989ndash1999 LaJolla University of California San Diego 2001

8 Fong GT Cummings KM Borland R Hastings G Hyland A GiovinoGA Hammond D Thompson ME The conceptual framework ofthe International Tobacco Control (ITC) Policy EvaluationProject Tobacco Control 2006 15(Suppl 3)iii3-11

9 Kadowaki T Kanda H Watanabe M Okayama A Miyamatsu N Oka-mura T Hayakawa T Hishida K Kita Y Ueshima H Are compre-hensive environmental changes as effective as healtheducation for smoking cessation Tobacco Control 200615(1)26-29

10 Harris F MacKintosh AM Anderson S Hastings G Borland R FongGT Hammond D Cummings KM ITC Collaboration Effects of the2003 advertisingpromotion ban in the United Kingdom onawareness of tobacco marketing Findings from the interna-tional tobacco control (ITC) four country survey Tob Control2006 15(Suppl 3)26-33

11 Pierce JP Gilpin EA Emery SL Farkas AJ Zhu SH Choi WS Tobaccocontrol in California Whos winning the war An evaluation of the tobaccocontrol program 1989ndash1996 La Jolla CA University of California SanDiego 1998

12 Gilpin EA White MM White VM Distefan JM Trinidad DR James Let al Tobacco control successes in California A focus on young peopleresults from the California tobacco surveys 1990ndash2002 La Jolla CA Uni-versity of California San Diego 2003

13 Centers for Disease Control and Prevention Cigarette smokingamong adults ndash united states Morbidity and Mortality WeeklyReport 2006 56(44)1157-1161

14 California Tobacco Control Section Proposition 99 and the leg-islative mandate for the California tobacco control program[httpwwwdhscagovtobaccohtmlabouthtm] Retrieved June 112007

15 Norman GJ Ribisl KM Howard-Pitney B Howard KA Smokingbans in the home and car Do those who really need themhave them Preventive Medicine 1999 29(6 Pt 1)581-589

16 Rohrbach LA Howard-Pitney B Unger JB Dent CW Howard KACruz TB et al Independent evaluation of the Californiatobacco control program Relationships between programexposure and outcomes 1996ndash1998 American Journal of PublicHealth 2002 92(6)975-983

17 Halterman JS Fagnano M Conn KM Szilagyi PG Do parents ofurban children with persistent asthma ban smoking in theirhomes and cars Ambulatory Pediatrics 2006 6(2)115-119

18 Kegler MC Malcoe LH Smoking restrictions in the home andcar among rural Native American and white families withyoung children Preventive Medicine 2002 35(4)334-342

19 King G Mallett R Kozlowski L Bendel RB Nahata S Personalspace smoking restrictions among african americans Ameri-can Journal of Preventive Medicine 2005 28(1)33-40

20 Ott W Klepeis N Switzer P Air change rates of motor vehiclesand in-vehicle pollutant concentrations from secondhandsmoke J Expo Sci Environ Epidemiol 2007

21 Aitken RJ Kenny LC Soutar A Measurement of personal exposure toPM10 in the non-workplace environment using passive sampling techniquesEdinburgh UK Institute of Occupational Medicine 2001

22 Szabo L Babies may absorb smoke residue in home USA Today2006 (August 6 2006)

23 Daisey JM Tracers for assessing exposure to environmentaltobacco smoke What are they tracing Environmental HealthPerspectives 1999 107(Suppl 2)319-327

24 Daisey JM Mahanama KR Hodgson AT Toxic volatile organiccompounds in simulated environmental tobacco smokeEmission factors for exposure assessment J Expo Anal EnvironEpidemiol 1998 8(3)313-334

25 Destaillats H Singer BC Lee SK Gundel LA Effect of ozone onnicotine desorption from model surfaces Evidence for het-erogeneous chemistry Environmental Science and Technology 200640(6)1799-1805

26 Matt GE Quintana PJ Hovell MF Bernert JT Song S Novianti N Jua-rez T Floro J Gehrman C Garcia M Larson S Households con-taminated by environmental tobacco smoke Sources ofinfant exposures Tobacco Control 2004 13(1)29-37

27 Singer BC Hodgson AT Guevarra KS Hawley EL Nazaroff WWGas-phase organics in environmental tobacco smoke 1

Effects of smoking rate ventilation and furnishing level onemission factors Environ Sci Technol 2002 36(5)846-853

28 Matt GE Quintana PJE Hovell MF Chatfield D Ma DS Romero R etal Residual tobacco smoke pollution in used cars for saleAir dust and surfaces Nicotine amp Tobacco Research in press

29 Bureau of Transportation Statistics (222007) National trans-portation statistics New and used passenger car sales andleases [httpwwwbtsgovpublicationsnational_transportation_statisticshtmltable_01_17html] RetrievedJune 11 2007

30 Mannheim The 2005 used car market report Atlanta GA Mannheim6205 Peachtree Dunwoody Road Atlanta GA 30328 2005

31 Hovell MF Roussos S Hill L Johnson NW Squier C Gyenes MEngineering clinician leadership and success in tobacco con-trol Recommendations for policy and practice in Hungaryand central Europe European Journal of Dental Education 20048(Suppl 4)51-60

32 Hovell MF Wahlgren DR Gehrman CA The behavioral ecologi-cal model Integrating public health and behavioral scienceIn Emerging theories in health promotion practice and research Strategiesfor improving public health Edited by DiClemente RJ Crosby RAKegler M San Francisco CA Jossey-Bass 2002347-385

33 Kelley Blue Book Values of used cars for private party sale[httpwwwkbbcomkbbUsedCarsdefaultaspx] Retrieved June 112007

34 StataCorp Stata statistical software Release 92 College Station TXStata Corporation 2006

35 White H A heteroscedasticity-consistent covariance matrixestimator and a direct test for heteroscedasticity Economet-rica 1980 48817-838

36 Armitage P Berry G Matthews JNS Statistical methods in medicalresearch Malden MA Blackwell Science 2002

37 California Department of Health Services California tobacco controlupdate 2006 The social norm change approach Sacramento CACDHSTCS 2006

38 Traynor MP Glantz SA Californias tobacco tax initiative Thedevelopment and passage of proposition 99 Journal of HealthPolitics Policy and Law 1996 21(3)543-585

39 Independent Evaluation Consortium Final report Independent evalua-tion of the California tobacco control prevention and education programWaves 1 2 and 3 (1996ndash2000) Rockville Maryland The GallupOrganization 2002

40 Semrad S A new arena in the fight over smoking The home News YorkTimes 2007 (November 5 2007)



BackgroundSecondhand smoke (SHS) contains more than 50 knownhuman carcinogens and has recently been classified astoxic air contaminant [12] SHS causes premature deathand disease in children and adults and the scientific evi-dence indicates that there is no risk-free level of exposureto SHS [1-3] A growing number of local communities andstates in the US and in countries throughout the worldare therefore adopting stricter policies to curb tobacco usein general and to reduce exposure to SHS exposure in par-ticular [4-6]

The introduction of stricter tobacco control policies isoften accompanied by health education campaigns aboutthe harmful effects of tobacco use on smokers and vulner-able groups of nonsmokers [7-10] The ultimate goals ofthese efforts are to improve public health through chang-ing personal smoking behavior community standardsand attitudes toward tobacco use and SHS [281112]According to the 2006 National Health Interview Survey[13] however steady declines in smoking rates since the1960s appear to have stalled remaining unchanged at21 since 2004 Thus tobacco control efforts remain ahigh public health priority requiring renewed efforts tofurther reduce tobacco use and SHS exposure

The present study examined the prevalence of tobacco useand asking prices of used cars in a community that hasexperienced extensive public health education campaignssince passing comprehensive statewide tobacco controllegislation in 1988 [14] We hypothesized that in such acommunity smoke-free cars would be offered at a pre-mium compared to smoker cars controlling for other fac-tors influencing the value of a car If this is the case futureresearch may be warranted to better understand the effectsof tobacco use on the value of personal property and howsuch consumer preferences could help further reducetobacco use and SHS exposure

Tobacco Use in CarsCompared to research on smoking restrictions in theworkplace restaurants and at home relatively little isknown about smoking restrictions in cars Existingresearch suggests that smoking restrictions in cars are lesscommon than those at home In California two out ofthree family cars had a complete smoking ban in 1996and 1998 [1516] compared to almost four out of fivehomes with complete smoking bans Among smokershowever only 29 had a complete car smoking ban and43 had a complete home smoking ban Similar patternswere observed in urban and rural settings of the US out-side of California Halterman et al [17] found that amongurban households with smokers and children sufferingfrom asthma only 64 had a complete ban on smokingin the home and 49 in the car Kegler amp Halinka Malcoe

[18] examined low-income families of children in ruralOklahoma They found that 49 of Native Americanhouseholds and 43 of Caucasian households bannedsmoking in the home but only 35 and 40 respec-tively banned smoking in the car A deviation from thispattern was reported by King et al [19] among African-Americans who found a higher percentage of nonsmokershad car than home smoking bans (84 vs 74) Amongsmokers however only 17 and 21 had similar bans

Residual SHS Contamination of Used CarsWhen tobacco is smoked in the confined environment ofa car tobacco smoke pollutants can reach extremely highlevels [20] Volatile SHS components absorb into surfaceswithin minutes of emission contaminating objects withwhich they come in contact Subsequently this residualSHS (also known as aged SHS or third-hand smoke[2122]) is re-emitted into the air over days weeks andmonths accumulates in dust and deposits on surfaces[23-27] creating a route of exposure for drivers and pas-sengers of smoker cars in the absence of concurrent activesmoking

Unlike mechanical or electronic defects detecting thesigns of previous tobacco use in a car often requires littletechnical expertise from a potential buyer Routinetobacco use leaves many telltale signs to prospective buy-ers Foremost is a distinct odor caused by the re-emissionof SHS contaminants from surfaces and dust that werepolluted during active smoking Matt et al [28] haveshown that SHS odor and ash marks are significantly asso-ciated with the residual contamination of dust surfacesand the air in cars It is this odor and other visible signsthat can signal to a potential buyer that a car has beensmoked in and that can be difficult and expensive toremove through cleaning or repairs

Private Party Sales of Used Cars in the USThe used car market provides a particularly interestingopportunity to examine the value of a smoke-free per-sonal environment because a large and diverse cross-sec-tion of the general population in the US sells and buysused cars Moreover a substantial portion of personalincome is spent on the purchase and maintenance of carsIn 2005 441 million used and 169 million new carswere sold in the US [2930] These transactions totaled$780 billion $367 billion of which were accounted for byused car sales Approximately 30 of used cars were soldby private parties [30]

Although asking prices for used cars are often informed bypublished pricing guidelines (eg Kelley Blue Book) sell-ers can advertise their cars for any price buyers can offerany price and the eventual sales price is subject to thelocal market forces of supply and demand Sellers typi-










NonsmokerSeller

SmokerSeller

NonsmokerSeller

SmokerSeller



































SmokerSeller

SmokerCar


1 2080 2893 2677 2670 26342 3223 4031 3731 3720 36703 4354 5187 4803 4787 47224 5674 6563 6081 6058 59765 7363 8358 7747 7715 76096 9186 10324 9574 9531 93997 11818 13200 12247 12189 120188 14900 16600 15409 15333 151139 20515 22823 21200 21093 20779




+135 +52 +48 +33





























BioMedcentral













































NonsmokerSeller

SmokerSeller

NonsmokerSeller

SmokerSeller



































SmokerSeller

SmokerCar


1 2080 2893 2677 2670 26342 3223 4031 3731 3720 36703 4354 5187 4803 4787 47224 5674 6563 6081 6058 59765 7363 8358 7747 7715 76096 9186 10324 9574 9531 93997 11818 13200 12247 12189 120188 14900 16600 15409 15333 151139 20515 22823 21200 21093 20779




+135 +52 +48 +33





























BioMedcentral



































































SmokerSeller

SmokerCar


1 2080 2893 2677 2670 26342 3223 4031 3731 3720 36703 4354 5187 4803 4787 47224 5674 6563 6081 6058 59765 7363 8358 7747 7715 76096 9186 10324 9574 9531 93997 11818 13200 12247 12189 120188 14900 16600 15409 15333 151139 20515 22823 21200 21093 20779




+135 +52 +48 +33





























BioMedcentral

























































SmokerSeller

SmokerCar


1 2080 2893 2677 2670 26342 3223 4031 3731 3720 36703 4354 5187 4803 4787 47224 5674 6563 6081 6058 59765 7363 8358 7747 7715 76096 9186 10324 9574 9531 93997 11818 13200 12247 12189 120188 14900 16600 15409 15333 151139 20515 22823 21200 21093 20779




+135 +52 +48 +33





























BioMedcentral














































SmokerSeller

SmokerCar


1 2080 2893 2677 2670 26342 3223 4031 3731 3720 36703 4354 5187 4803 4787 47224 5674 6563 6081 6058 59765 7363 8358 7747 7715 76096 9186 10324 9574 9531 93997 11818 13200 12247 12189 120188 14900 16600 15409 15333 151139 20515 22823 21200 21093 20779




+135 +52 +48 +33





























BioMedcentral































































BioMedcentral















































BioMedcentral





































Documents

Tobacco use and asking prices of used cars: prevalence, costs, and new opportunities for changing smoking behavior