Tobacco Documents Online

. witvE.Mmvc MEwaNt' 12, 33a-3t1 (1903) . Hbwv Much Cani Business Expect to Profit from Smoking , Cessation?°,2 I`[AeVIN M. ICtusrtaW' Dt,parwtrnojFeaan.kar. StYNY. SlowrlrvoA: New Yo.k 1/79/.aadAiwnir.e 11r&hR ForeQhNOn. 3I0 Eese I3.d Srrnre. Nrw Ywk: New Yo.4 Wl7 Tbe dMa teviewed in ffiis: efsiele. ebr.w irom vsakwvrn. poWisbed sources. offer an evAmue of tl,e amaul eoatc d sawkieg to t>te averate eepioyer: TGese aosts tanae hom f371bwi601 ia Jaauaey 19W dollars perasrade smoking empoyse. 7De aaalysis is largef> bned oa y,pepue eodemiolopaal'dat. sed.,iu paenl.,tAe eapAsis is on tAe undcr- estimsting d'tlie eoaa to busineu. lt reports •'avnrqei "tsue tor lerfc populations aed. tierefiorR. it ao1 ladfviduWy aredietive: 7U ooA iaasers deab .dt6 ane isswanec. aedual. and iatlers sYCh a aitseaeeism. podued.ity. iavol,anary uaokug; and oceup.taaat healtl, tfsk: leeluded a a Aifewsioa of aa~ioyes~spo.eoeed s.wlciagxwrioa proQams and iee t~eeeifts tfersLraes. INTRODUCTION The Sul=eon General of the UtlitedlStates has stttitad: ••cipretu smoking is the single most important preventable envi7onmeata!!tactoe eontrlbuting to illness, disability. and deatb iin tbe United Staues" (78). In tbe mtlltiw.riate analysir of tbe . taslefor coronary heart disease, bstaed,on Framisl=hm detai cigarettes smoked is a dosrsecondto the number one risk factor (after ap). eboksterol level.,for men.' T6e c.mtnection between smoking and txsp'tratory, cancers--the only aujor t.m of wlcen trepstering an increasing mortality rate in secent years-is well estabhshed (1r*, 7s. 88,92). How much is this well+known heahA problem costin=business and ean some d these costs be reduced in a cost-effective mtulaer? A 1479 survey of 3.Q00 companies by the National Interagency Counc~ll on Smoking and Health repurted tbat less than 1% bad calallsted their costs due to stmokin4. 1 15% had heahh promotiac proRams involving smoking, and one-third wrisbad to develop or expand prograals for employees. Two-thirds of the com- panies had smoking >xstricdons in special '•bWe eollar•.•'areas; threr-qtlarters or awre had oo restrictions on smoking in general ($911 Tnirartick reviews the costs of:mokin=f prialanly dased on data from wdl•ieaown. published sources. in order to aid poGcy, tmakers in develbqft a rational eost-eltectiire policy in thirarea. The analysis is largely based'oaaggreg.te epidantioiopcal data; ireeports "avetaaes." tnue for large populetions; it is tlot individuatly predictive. The numbers are ••b.llpark•• fisllres. not seat, and aisle numbers. The cost centers dealt' with are • To wb- e.peint nasan eeadd be .ddna~: A.eiium li.dti Fowadrioa. 32ID Esn 43rd lteeetl Wm Yoet: NY 10017. aiuqporseA irp.ni by NQiS>< Graat 1J«4JJH9031/13-01 vd'USDHEM GeoaC2RI-7io16isAD toalse Aaeeiea. H..h! Fou.datioa, New Yort. NY. e Msad oo a Wk preq.rud!for tJse Naioetl Wee.p.ey t:eriei w Ssoolii.g swd Hs.hA. January IUiI. idmYmii: sTmFneinb.m lieev diKt'simee tbendoroa.d6eieob tfK..dnd ~ta) Ror,area of ail lyes are: tera.o. oss; forel,o+aweot, a4.: for aipesae ie.oiki.& 0.12, l.rsynoirt,iood pessws,. 0.39 i(7s): a09T-7/3"3 $3,00 C.P1.r1 ~0 ~110 ib AMMS w..: J.c. AJ IFJOU 1 w..YC.~ i. Inr. w.. if=..d. . M-7 >sa

TROF11rAa1LiTY OF EM[r11OtlSEE SMOKING CESS;ACTtON 359 insurance (medical and other);, absenteeisat, productiwity, involuntary smok= ing, and occupational health ri'sk. We lack meaningful ••case-eontrolled'" eompany comparisons, of experience with smoking employees vs, tvotrsmoking employees vs ex-smokers and the uapact on company costs: Therefore„we must work, from asgregate nationall epidemioloaical' data on illness and other costs associated witfi ismoking and attempt to apportion a realistic share to smoking employees. in general, the emphasis is on andrreszintciing, the costs to business: A sttmmary, of the, various costs of smoking ealculations, performed I in this, tnanner, indicates that the average one-pack-plus-per-day smoker mayy over his or her lifetime, cost his or her employer between S336 and SG01 per year, in January 1980 ddollars, in extra expenses as compared with an otherwise eqpivalent, nonsmoking employee. Adding about 40% would bring the medical-care cost fig- ures up tio ~ January 1983 dollan; adding about 25% wt71 i adjust the other dollar, figures tdmilarly. Preseat;studies suggest that mone than half of these costs anay, at kast' in pan. be recaptured in ~ the medium to short run i by smokinQ uessation, prognams at the workplace. One particularly striking resuh of the analysis which follows is that the annual nonhealth taedicai-care costs may be greater than the bealtlrcare costs. We also deal with the issue of substituting one type of cost for atatherw,t.e.,,the possibility that ~inereasittg pension fund costs may ovetwhelm the gains in terms i of savings on medical care qeendins and higher Qoss national product (GNP) due to lower morbidity and tnottality: Trble 1 suaunarixes the ettrnent totai bealth-ea:eeosts and indirect costs in lost GNP (due to morbidity and premature mortality) of smoking to the society as : a whole. METHODS To attempt a scientifically based answer to the question of how much i employee smokittg is aetually costing the typical Amerkan ~ business and by bow much this cost can be raditced would require studies and data we do 'noc now-and most. likeiy will never--piossess. Thus, it wottld' be necessary to have data rellecting at least a 2(~year follow-up of.at least one case-controUed study eompataetg,a com- pany with several thousand employees that had pursued a smoking cessation effort (perhaps as part of a broader health-protnotion program) with a similar company that has not pursued such a program. One would eoatpare.measures of spendinr on health insurance and other insurance coverates (includittg life, 5ne . and accident)i absenteeism, productivity. occupational heatth. etc., and the costs of the program. Tbereby,, one could estimate over time the sum saved (or not saved) by the company that had reduced the amount of'employee smoking: A 20-year fonow-up would be essential for a full'expression of the range of diseases related to smoking, given the dose-response liierature. as is the requiremenr for several!ttiousand employees for reasonable "effect size" given the incidence rates ofthe diseases involaed and the differentials in rates likely to be found. I+iothing in tlirliterature even approaches such data. There have been aggregatiive and special epidemiological studies and estimates of overall extrr illness rates of smokers vs nonsmokers and ivs ex-smokers, however. ln ~addidon, there is extensiwe literature on the relationship, between chronic diseases and smoking based on population and laboratory epidemiology. Alsu, there are studies,,generaily with serious re- M~-8

360 wAttVIW M. <r1tISTEITI Tuf,DLE' 1 To?A1~~ ECpNOWK~.CoiT! Oi SMOa1NG /N'.TMa UNIr[lf ST~ATif !l[a YLAa INJANUARY' I9!<0~ DouAas' (M/uaNs) Dirsat oals' l.dincl tostss' Tow i Nelrolasms $1,433 56.3101 57,763 Cirevlmory' Sa20 slt.2l4 23.3'50 Ziaspiratory 1,430' 12.0'" 164475' Toeai 11.023 j6.i6s s9sas ttr *duh sawker 204.00 6T1.00 811.013 /4t eisaeene sold 0.02 0.016 0.08 !w pac(ss0e sold 0.36 1.19 1.56' • Hased oe I:noe ald ScAmeitter, 1'/arct+ 9. 1Mds21',and aswmOtioe of St milhcn aduh'smoben' a.d 612 bdlioe!eiplettet sold per yar: • H.dieal!eare :speadift o. d"usuesrwhI to smokina: •'Lest praiop'due to embidiry, aad premauns eonaliry, (1liuaamed at'1%). search limitations, of various populations dealin=',witb measurts of productivity. absenteeism. ievoluntaryamolcins, and smoitinj and occupational health (76. 78): This article attempts to mine available dlua to obtain the information soul;ht by starting with aweptive epidemiological data, largely for, the United States as a.vbole. and reduciag it to per eapita dats by dividing by the size of the': relevant poptlllstions: Estimates of the~ excess chronic and acute illnesses asso- ciated with sowicittg, i.e.. qnt;er, drculatony disease. and'respiratory disease; are: taitea'at the lowest range suggested in the literatwe in order to underestimate thee impacts of aal+okint. These percentages of excess ~iWless atx then used to calculate the costs of smoking by' multiplying them by the generally accepted estimated dollar costs of the relevant 171nemes. Old data is inllated by relevant price indices to obtain up4o•date dollar 5=uaes: Obviously. tnia= multiyGed estimates as inputss pt+owides even less aectnste estimates as otuptrt. However. every effort has been made to emphasize tbe ~ lowest ranW of estimates in the sensitivity analysis. Details of tbe atktllatiotts arn provided iu apecial ydereaces to Tabib 2. Tbus, the upper endpoint of the flarl! etafmate, i.e., that the average smoker costs his or her emQloyer about'f1600'per year in various forms, is aa underestimate given the basis from which it is derived. The lower eodpoint. about $340 per average smoker per year. is' an almost arbitrary rednctiott of the apper 5gure by about SW: 71his is a sensitivity analysis where the medical!eare portion is based on data fi~romm an early dnft of a paper on the health care costs of smolCini'renecting threxpenience of a, bealth1 m.iutenance organisation (HMU) ptlpul.tiott (79). When the results were =ener.lized,, they provided an estimate equal to about half, that of the Lucrand Schweitzerfi323 data. The wrorleers' eomperuation: absenteeism, and productivity low estimates are arbit>rary 50% reductions as wdl. rounded to the tyearesr lowest $5 figure. Some of the eost' of smoking estimates are based on statistical associations of the variable concerned and smoking; some are based on laboratnry epidemiology and controlled population studies. lUl who have studied statistics know that cor- relation is not causation, but correlations iupported by elinical, experdmental4 and laboratory data and tested epidsmiolopcal! theory come as close to cause and. M-9

2025684330 0 ~ A ~

362: MARVlN M. ICRIST'StN TAII:E 2-•Cotwi+wipd The b.ier'tsrp is baseA' on an annual salary or M0.000: the npOer'ra.Oc e# an annual salir) of 1115.000. (e) Absenteeism cosu are based on the literattm discussed ie the tsport' of the U.S. Surgeon Gtneral: Swtotinr,.nJ Xedrh'. 7aeuary 1g79: Chap. 3471. ?i). The estimate in the Inen+ure is that 81 million workin5 days atx lost per year due to smoking. When diiriabd by d0;tn....: • Workin5' a.mkers. this reduces to 2'ays per aooker'per year. valued at i/0 per day Sivine as the $!o hiyp ran{e. Tbe low range d ts0 is twc only ose-balf the high nrtOr bw also rtefleets the dilferenual laduction d approaiatately one ay per ete-smaker found in certain work site, beahh-promouon et;onsd~, S. 12. 41, i5)~ (fl As indicated in the te:t; erximate of the productivity cost of smokins is bwed on assuming ttrat I min per working btwr, or I tnin a day. are wasted by smokine,rituals.,etc. Actual emplo)Rr estimates are eloser to 30 min Per day (ti01,i1): The iar,tante /fSure lon produniw.ity losses due to amokitK,aerralbiam* ene-balf of the bieh euimate. rounded to the rwreat lower L5. (t)!TRe loss dee to itap.ets an Involuntary smokers is,raku4ted by taking one-68'i+ of the costs assipnd!1o'qakers. emittie5 6nc losses. 73is'is based on the White and Froeb int8ttt5',It3b that i.vahmtuy smokers aay setlfer breatbins eapaeity impairment equal to tbat tor bSltt smoters (onc to tea eipretres per day) atW that the eaear risk d fiebt sawkers is oaefiRb that of uaohen of one !ack 'Nes t+e* day (!Y): We believe that ttrrft relative nncer'risk as an estimate for Oencre1 impairm ttaeat a an tmdbetqim.te. tiaar eaeeer ia a telttivcty nre disease iavolvio5 a IonO-term nla+ionp abip ait6 emokieS. as compared with oardiovasenlr,r ditwus and respiratory diseases on which 1tn ibtpct'ef tmotitt0 is rsrter and ore immediace. Fnrtber.we assiSn,ody the short-term risks to brtinesaes. i.e.., on-Mb. of losses due to rorket s coapeasacioa., abseereeism., and pro. ~etiriy, or iorr,.n0e:S27: Yo rroKc. ssti. 01 21c, 57,2 liswe 1or oce.pationd Yealth is baad on t!e October H tl a. vd by ttte, Supreme Coun d NotY Caroiaa iwa bro~ra Imte ease of 57,000 per ysv for f ysars. x S 1a1.00D to an earty disabitity.retirsd etepbyee. If one etttimaus tbe,otaidenar of COLD or btorn lunr disease to tx ; tom7ar to IuaO cancer ib smoking onnwm tmaers aad asbestos .rorkers. then we are dealin5' with a ootential annual rate of abou! 10 per W000 per year (71). This probably involves an ineu- baios period of'20 years of'esFotro!o. Hovever, for an old company with a lbn5-titne smoking em- Floysrpopulatios, the 0.0D+( latuwal incidence rate aay be etcperienead every'ysar'for the fonesetsble Mtafre. Obe ari0ht. tiermfoe. view, tthe oewpatioeal health cost as a sbort-term cost. but since it ia leahb batied, here it is< dasadud as a ktarterm cost. Bwe .uhiply the Si5:000 award by the Ate dd0 per f0A00i(0!004)we ubtain a ooterroallanmml eost pa smokin5empioyee'ofiY:.,Atcording to Heyden and lRan ('!2). e4mioation o[,twrokin5 a the wrork site would virtually eliminate cotron dIrst,rsleted ttisFiratorry disease. Tbus. is tbis ease. one may assign the entire occupational health ftk award to a.mokia0,ia terms of economic analyris; e+ren tboush the caan viewed!tbis portion as .oenp.rio..l i>A.k: effect as;can be expectedlat this tirue. Forexampk..vecannot state with certainty that absenteeism will declirn if employees stop stnoking. but insofar as absen- teeisnl reflects illnesses (largely acute respinatory, in the short run) related too smoking, absenteeism should decline with a decline in smoking. In facr, thiss outwme has been observed in several tierltlrpromotion, efforts by various types of employers (3. 3. 12. 4'1,., 83).' For cost centers involving matters other than cost of fUness, the approach of applying estimaterof the net eQect of smokingon ~tae cost center, and Ito value the •T1e beNtb-promotion efforts of, various employers rtlerred to ia this contexc include: (a) qtpioyees in the New York , State Department , of ' Education. (b) en iAmeeican Health Foundation RroQam eartied oot.fw employees of the New York City Fasbio.OSrceaf'Skan. Roetwck i Cmn- pary. (c) American Telephone and Telbp.ph Cotapany. (d) Ne.' York,Teiephone Company. (e) Speedall Corporation. NS-11

fRCFITA'11LITY OF EMPLOYEE' SMOKLNCr CESSATION 363 difference by a(fenerally low) relevant current price has been used. Forexample, fire losses per smoking worker aro estimated by, taking dau.,on nonresidential'fire losses from the "Insurance Infortnation,Institute Fact Book." Atrestimate of the percentage of fi'res due to smoking is applied to this figure„wbich is then divided by an estimate of the number of smoking employees to oboain the estimated cost of $5 per smoking employee. Estimates ofiothertypes of insurance costs (for elcam- ple. We inswarce) ane based on the higher age-adjusted morwlity, rates for smok- ers vs nonsmokers that have been observedi The diffelxnaes between iluurance pre- tniums cbarped. retleciin= suc11 elonalily rue difrerences.,are compared. Productivity-loss estimates rdlect a consideration of a range of estimates from titee ~ and motion testimony to i the U.S. Congress involving a service industry (banking) (bi) to employer comments in the construction industry (R0): the lowest level of estimates is emphasized. While such estimates are not scientifeally ek- gant, we know thatthe smoking rituals take tialeand that them are added cleaning and ventilittion costs. In addition, there are more carefully developed studies of the effects of smoking on the mental and phy,sical capacity of smokers. Again, our estiluates try to reflect the minimal level of such costs. TWE' COST BY COST'CEN1fER! The usual procedure in cost accounting is to classify costs in terms of the adnunismative units. type of activity areas, or dtpanments which pve rise to the costs being studied. These are called cost centers. Since this article seeks to be as fenetalited as possible, the approach followed is classification by type of activity. Tbus, the costs associated with the areas of'insurance costs (major relevant ty,pes). absenteeisnt. productiivity. imvolrintary, tsmoking (effect on ttonsmoking others), and occtlpational i health are studied. The implications of ' the data for company policy are ezplorrd, In addition, examples of cost effectiveness of com- pany programs are discussed. Also, a company experience with work site; no- smoking policies is,neviewed (Illppendix 1) and a caveat oldered (Appendix 2). The evidence suggests that an aggressive antiamoking; policy is a viable company policy thu may be bighlyeost effective. 1. TJu InsuraRce Costa of SnrokragA, company's insurance expenses depend on the coverage it offers its employees. The vast majority of companies offer, health-care insnnance, largely paid for by the company, for employees and their families. Most companies also have fire, worker's compensation, and accident coverage for tbeir employees. There is well-known, reviewed litetature dealing with the excess Mealth•care in- surance costs associated!with smokers. The doUar cost figures used!here are based on the research of Luce and Schweitzer 1978 ' (32). which is, in tum, based on a 1976 study by Boden (11). These calculations rest on the judgments of'medical experts as to the percentages of certain major diseases which are associated with and related to smoking (6, 44, 45): neoplasms and circulatory (cardiovasculxr) system and respiratory systeln, diseases. The percentages emplbyediare, respec- tively: 20, 25„and 40% and some recent Iussearcb indicates that they may, be under- M-12' I

f 364' MN'ENJN M. KRIS7EM estimates (7d)' TUesr percentages are applied to the familiar Cooper, and Rice calculations ofithe economic (d~oilar) costs of these diseases (7; 18pand are then adjusted for inflation. The total! figure for the annuai excess-illness cosc (i:e.. direct' excess medica)-eare spending due to: smoking (in January 1960 dollars), would be over 5111 billion. The author has seen, an early draft of a toore' recent attempt at a national estimate4the zedical care cos[ of smokin.g based on the experience of~an Oregon Hh+IO sample of some 2,300 persons.(79) that anives at an upper-range figure of S+t.S'billion in 1960 doliars:"The study covers a nel*tively small sample for only 7 years. ln fact, the short-run quitters have a 26% excess usage as compared with quitters of over 4 years standing. One may use this figure to adjpst;the upper range estimate partially to refiect the smoker/ex-smoker difference which would be found in a longer follow-up (e.g., 20 years, which is the lead time for the relation- ship of respiratory caneers to smoking). This raises the Oregon HNlO estimate to 547 bl7lion: Seven years is barely enouglttime to establish a relevant minimum rate for bean diseases and respiratory diseases for the eoltorc; E2% of whom' were under~ 65. IUSo„ttie ex-stnoker group appeared to consist mostly off persons whoo qvit becausr of tllness, larsely beart disease and respiratory disease. How many of these costs may, be assigned t o the employed! population? How, tmany of'that populatiian's direct costs are borne by coatpany-paid health insur- ance? Dr. Jeffrey Harris has estimated'ahat there were S4!mtlfion smokers over the age of 17'in 1978 (18). Tbust dividing aggregate costs by population. one may say that the average adult stttoker. o+ves his or her lifetime. accounts for S 1A5-20![ of'the annual excess ttledical-care costs as compared with a nonsmoker. The indirect costs of awrbidity'and premature tnortality, i.e:, some 330,00!D premature deaths (78)n, which involve considerable personal losses to the itldividual„ his or her &m1y,and tlle;societyas a,whok: are sepameftom,the direct costs and average' over, two times the dollar cost of direct medicat-eare costs (7,18): They' are excluded from the discussion at this point, since they are picked up in later sections,of ttie paper on abseateeism, etc., aod do aat repnesent direct heatth-care costs to the employer. In the author's 1479 paper (40, be attempted an ~ahtraative type of cost calcuia- ttioia, differing from the l.uce attd Scbweitzer approach to the deta: Rather, it was • 7>ta Anaieas Qresr Soeisty aewaaes drt ss,okiss ia;anoeirsd .~ Jt IF dltie ireidearce aed ! 41% ef tia Maths ans to aaopiaaau (1): Alio. 7096, o['the #atas due to clrpeic obur uctive pulmonar) Auaau .ee r.lar.d'to araki.{ Rs-7p: awdovaaedv ii.rases lCVDI io tse so: to s6year-old ltopalaiioa. aspaoiaily adde.4aati Mut,.aackan ate Ye.vily asaociated with tnroties. Swisucsl atwdiea Yave foaod ewokia5 to bs tYe variable tbat •'taptrini'" up to SO% af'tlre vatiarioe.in CVD simtahWem tiwr(+17). The upper tanse d mrotifWt imp.ct an major'diseafes a'probaI for a.6pras asepiauas. 35% for cardiowtaodir disea.es. 65% for rssyirxay diaeases. e'Vop a.d ScDMeieaa faoad a 294E atars aedicai are ayaEe by •'e6cir.* soatiauibs smokers vs aMaoters ie t!e Ore~ HMO (79). t3is lower ttre aay r.pesena:ecornocted eswn.te of,tee:3p K aaieWe raaatly'poltdI or it may, reflect the stwn'foUowwp period ofoely 7 ytan and the fact iAar tlw ea-smoha Vouo w keavn'ly weiSAtsd .rith petsoas wAo auI tbeir doctors' advloe, a e1lenin5 , tt,e *r.asaae ,d $aious diuase ~. M,-13

lROflTA[ri.I1fY OF EMPLOYEE SMOICING CESSATION 365 based on total U.S: health-care spendin8 figures and employed the well- establishedipoint in the literature that adultbeavy smokers, especially men, make greateruse of the hr.alth-carrsystetn, especially hospital stays, than do nonsmok- ers. Smokers use at least 30°Ja 1(78) more health care (48). reflecting their increased prevalence rates: ofi chronic and acute conditions and disability, days than do nonsmokers. The results of this calculation produced fi8ures: for national health eare costs of sawkine, which were strikingly similar to those obtained by Luce and Schweitter's 1978 study ~(32): Obviously, the Nealtb-cane expenses of smokers, individually. are concentnted at higher' a8tes; for example the average age for lung cancer is about 60 in the United States (92). This is derived, as one would expect, from the studies of the dose-mponse formulas neiittin8 the relative task of lun8 cancer to the average dat7y tar intake and ditration of smoking (S3, $7, 89). The avera8e age of first heun attack is much bwer and fallina; it is probably about 55. Therefore, only a iraction of health-care costs due to smoking should I be allocated I to : employers. This is conservuiivety estimated at tbree-fourths of the national lfigure, since 87%ofithe medica!' eosts associated with arnokin8 involve circulatory and respiratory dis- etises, which are significant at earlier ages than cancera. The Luice and Scbweitzer article (52) also contains estimates of fire accident costs due to seaknl8, which amount to $10 per smoker per year, includin8 resi- dential 6nes: As , estiinated I herein, nonresidential fire losses are $5 per smoking employee per year (see footnotes to Table 2'for details of the calculations). A S10 fire-loss figure is presented as an upper limit since there is reason to believe that the proportion of nonresidential fires due to smokinQi is =ready underestimated and many of the historically most dlsastrous iadustrial ifires have been caused by smoking (15, 22). The author estimates that other accident costs and the smokers' share of t!)rc annual worker's compensation costs add an additional $17, to 1 $34, per year per working snroker to the insurance cost category. These calculations are based on studies by Naus et o[: (56) and Yusta aad, Da Guevara (78, 94). which 51td' tbat smokens have two times the arcident rate of nonsmokers, due in part to loss of attention, smokin=,hand being occupied. eye ittritation, and cough. These higher accident rates are applied to the total cost of worker's.compensation to obtain the extra cost due to smoking. (See footnotes to Table 2 for detads , of the calcula- tions:) Many companies pay for, life insurance coverage for th* employees and have extensive disability and early retirement provisiOAs. Many employers offer group tetm ~ life insurance as a standard benef t worth at least I year's salary. Such, benefits may add an, additional $20 to $33 per year to the saNokinB, bill l'' These ' Tbe Caf(le.aia GASP INtwtita.r for lawwty 19l0 (1t) 1 repons: Stste Mutual l.ifs Assunnce Coaspaay of,Aatsrica haa {o.vved'peemiuaa per3l.tl(q ol tife insutanee coverqe froeo f3.16 to 5:.67 for aaateoiusa; Ml.ctinS the eompany'1 esperience that. betwees tba 6th aad 1(kh ysan of covcraSe for oersoes a{sd 30 ~to 40. the stnokarfacowtoku 'deatA rario for aA causes vas ~1:3: Also. a rypical 32-year-oid m.le taioker ws ssciamed to havc a lile eapecuncy aveeapaS 40 yean a: eaapared to 19' ycvs far rsoasawiier. M 14 KQ111

C 3" MIA'RV1N bt. KRISTElN estimates arrbased on data as to diNerential'.mortality of ctnokersvs nonsmokers reported by the State Mutual I.ife Assurance Company of America and!data in the author's 1977 paper'(20, 48; 57). (Sbe footnoterto Table 2 for details of,the calcula-, tions.) These eompany costs really reflect the extra cost to the business community as a whole. If' a company is experience rated or' self-irsured' (usually a very large company) then its premiums should reflect the high cost due to smoking: and it might~lbok for!ward to directly pocketing those savings by smoking cessatiom Smaller' companies are =enerally' community rated. Their' premiums reflect the extra~costs to its local insurance area in the society as a wholy due to smoking. Unless all such employers an&employees cooperate to reduce smoking. a single small company may not sueceed in reducing its premium costs even if all its own' employees stop smoking and!its medical-care spending declines sigrtifiicantly. 2. Absenttrrsrn The literature on the absenteeism record'of the avenie smoker vs the average taasmoking employee is extensive;,thene are published studies of experience in the United States and'other eonntries,covering rvariety of occupations, ranging from Connectiaut tekphoneoper'ators to members of the'post-World War II Ger- matrArmy (36, 71~ 78. 84). Smokers,have 33 to 45% excess absenteeism as cbm- paredlwith nonsmokers. These figures do eot reflect ease-eotttro0ed studies nor do repeated measurements of 6igh,R2 eonrellstioo!eoef[icients represent causation. In 5et, one may argue that higher rates of absenteeism,and smoking,b!oth relate too and reflect otbcr factors. However, studies of'work site health-promotion pro- jrams, including smoking cessation efforts at a number of ditferent, eompanies, Gnd'substantially reduced absenteeism rates in the shortirun'aftera year or two for participant populations vs bistoricaf and current t3outrols! Based on the 1974 Health Interview Stuvey of'the I+latio.oal Ceaterfor Heahh'Statistics it was,esti- tatated that tbrexcessabsenteeism,associated with smoking in the United States ataountcd! to 81 million lost work days per year,: or about 2 days per smoker (78). Absenteeism among smokers is partieuLrly' concentrated in the 1'f-s4' age group. If, valued at $40: per day, that means that' the average smoker may be costing his or her employer as much as $90 per year in excess absenteeism. To deal with the issue that all : smokers' excess absenteeism may not be dite entirely'to smoking, and!tdat ex-smokets may newer attain tite kvelt,of nonsntok. ers, we ofter, a lower-ran8e absenteeism cost figure of 'S,0' per year per smoker. These calculations do not include the costs of temporary replacements' and I the effects of absenteeism on the: prodtutivity and morale! of co-workers. , •An bwendt study by, tbe aunift sernrice of'Swah'ern NewEnSked DeUlTieiepAone found tha, sMoken','ibseece duelo edds ialone coa an .renSe a('S3o per sowker per re.r norr than nonsmok- ps' abseucs due taooltlj. (i;epotsed at ttte Conetotiow LwS Assowtiow Coofersooe. November 13: tm. Ne.r Haven. Coseecticut.) M-15.

lROF11LAKtLtTY'OF EMPLOYEE SMOKING CESSATilOI+ 367 C 3. Prodattisitr In addition to the impact on insurance costs and absenteeirm, there are direct productivity losses due to :ucll' factors as time lost to smoking tituals, extra dean-up costs, extra' damage to equipment, furniture, and fixtures0 extra, inefT- eiency and errors, discussed in the literature as dite to the effects of higher carbon monoxide levels in anokers, eye' irritation, and measured lower attentiveness, lower cognitive and lower exercise eapacity functioning. These costs have been estimated in 19b'9 Congressional testimony andlhave also been updated over time (9, 63, 86). A reasonable anderestimotr of such productiivity losses per year per average smoker for the'average business mould be S16G (in lanuary 198ID'dollrars).tq This is based!on a minimal estimate of a loss of 1 min per'workinQ hour, or 8 ntin per day, for' a 250-ds<y working year, at $40 per day. AB,ain,, to allow for' the possibility that exsmokers may not attain the performance levels of nonsmokers we offer a lower range productivity estimate of'S80 per year per saloker.. 41 Iwvotyntan Smoking Thretieets of smokin=on nearby nonsmokers have been studied. Forexample, the children of smokers have higher rates of respiratory diseases, adjusted for' relevant complicating factors (Z'7. 73. 78)1 The ttonsmokin='wives of smokers have been found to have higher rates of'respiratory cancer than would be expected amon8,nonstuokers (3'L. 74, 7'lL There has been much controversy on the topic ofindtsced eancer in involuntary smokens (2!3. 49; 38)." Therefore, our cost esti- mates emphasize impact on acute illaess and!work capacityrather than on cancer. In addition to the considerable physical and psychological irritation thati nonsmokers sufrer at the hands of smokers, nonsmokers' work performance abil* ity is neduced by the effeets of nearby smoking. White and Froeb ('30. 8'3')~ re:pon that the'damage to breathing capacity as a result of iz+voluntary smoking is equal to that of 1i81tt stnokiAg(i.e., one to ten cigarettes per day)': It is reported that there is a doublittg of tltie rate offunctionally'impairingastbmafor children ages <1-17 with smoking mothers,, statistically significant at the 0.01 level, as eompared' with nonsmoking ewthers' children (27). Also, disability days were 30% higher. even when controlling for the asthma and'brprnlutis: IiFone applies the White and Froeb s t~uce D:1+Ikifoe. Sestian Emoret of the EeerSx !lmnieS and Dtretb4mess OQke ditAe Sute of Itiltttsson has sstiwsted tiw'ehe eatta viesnilatioe twaqtunaeats. bned eaa s~esdatiom d tiu Articaa Society d'HeatutE. TislYiaeratinS. sad AkCon0tionias Eap'eeen (ASH1tAE')1 dw to aaokin{ at itle .roekptacs'~ ouW add ;aboul f;W pa yeat per a.qker to ttie etsee=) eosts af operatin5 :a plant (q: -'Wuh the praem reoommtsdatiosa (of ASHRAEJ. waken'wauW laccanet for 17F', of the Motand fbr'vantilt+tioar e.ef taott{h ttKy'conduste'less ttwt aae-third of the adult populttion i' l13). Ttis 1fSn+r .Moutd be added to the yeoductivity estisute' Si ren in the teat. since tirc latter is simp4 based ao '•time waaed'- and related aspects of sookis>s rituall. "Eetpioyen .rp6 eoasmokiaS!arias at the .+otk sile tsponted ttnt tAey'prarided tw or awre dri). amoki+tt,btxaks, a.etaong ,/3 mio. for.wrkkers is ttuse areas (9), " TAe Japanese study of tAa hrer cancer e:perience of aonaawkiaS ais~es otirawkets is supporaed by a Gseek audy ~(3t.14): The luter'finds a hiaAer tsbytive risk lot lua5 caacer'amosr<ttte twoesmokm5'. .ires atsa+okers in Cnsece. dose-sesponse related to the amouat'of smokby by the busband i similu to the AedinSs for Japan. Passive smoking Gnftk womea bad only a slightly io.rer tisk than activelb saroking l3neek women. A U.S. stttdy'failsd!tohnd this seiatioad6ip (Z3): T!e statistiaJ ralidity of the Jap.nese uudy has also txen queuioned (49). A r.oent umeetins at the NatioaN CNCer l.satutt'on ttiis topi+e ended meonedtsivdy (3i). M-1b