Tobacco Documents Online

35 1.000 1.000 1.000 l.OO0 ! DO 1.00 45 0.986 0.')66 0.988 0.984 1.02 1.00 53 0.951 0.893 0.962 0.939 1.06 1,02 65 0.867 0.733 0.901 0.831 1.18 1.08 7'~ 0.689 0.466 0.760 0.630 1.48 1.20 85 0.336 0.159 0.453 0.289 2.11 1.57 Source: Derived from data in the American Cancer $odcty'$ Cancer Prevention Study II. ~ Probabillti~s arc ~hosc of a IT-year-old surdving ,o ~hc asc shown. shown by sex and smoking status. Probabilities of survival arc higher for females, but the same patterns are observed among smokers and never- smokers of both sexes. The benefit of lower death rates among never- smokers at all ages accumulates with age and creates a steadilF widening gap in survival rates. For example, whereas 87 percent of male never- smokers and 73 percent of smokers survive ro age 65, 34 percent of ncv- ersmokcm survive to age 85 compared with only 16 percent of smokers. In terms of relative survival, out of male ncversmokcu and smokers =live at age 35, 18 percent mo~e of the neversmokets survive to age 65, 48 percent more of the ncv~rsmokcrs survive to age 75, and more than ~vicc as many of the ncvcrsmokcrs live to age 85. Among females, the probability of surviving is 8 percent higher for ncvcrsmokers at age 65, 20 percent higher at age 75, and 57 percent higher at age 85. Smoking exac~ a much greater toll among males in terms of premature mortality. The disparity in mortality between male and female smokers reflects differences in cigarette smoke exposure (U.S. Department of HeAth and Human Services 1980). The mean age of onset of regular smoking among successive cohom of men has been less than 20 years since before 1900. It has declined slowly over time to between 15 and 16 years for cohorts born between 1951 and 1960. Among women born at the be- ginning of the century, however, the mean age at omct was 35 years. Although this declined steadily, it was not until the 1951-1960 birth TI310~07,..,9

| cohort, now 30 ~o 40 year~ of age. that it became heady ' ~ of men. ~ Exposure also depends on ~e likelih~d of qukting smoking{ : Among cig~c~e smoke~, quit ratios (~e pro~ion of ~cmmoker~ who ~e former smokem) have been inc~eming for ~th sex~ a?~imil~ rates since 1965 (U.S, Dep~ent of He~ and ~uman'Semice~ 1990). ~l~ough highe~ for m~ ~ fem~es, ~e gender difference only a couple of percentage poinu when quit ratios count for use of o~er'tobacco pr~ucm besid~ cig~et~es. In addition ~o age at Mifiafion ~d likel~ of quitting, smoke ex-: pomre depen~ on vmious d~emiom of the style of smo~ng, indud. ing type of cig~ette, depth of i~a~on, ~d fiaction of cigarette smoked. Wi~ ea~ n~ genera~on. ~e surgeon gener~ hm condud~d,~ men ~d women have become more ~i~m in ~eir smo~ng habiu. ~d ferule e~o~e closer to ~a~ of men. In ~e ye~s we c~ expect m~e ~d femfle mo~i~ from smo~ng ~so to become more ~ilm. Mo~fli~ rat~ ~d pmbab~fi~ of ~u~iv~ demonstrate ~a¢ n~er- ~mokets live longer ~ ~mokem. ~d .m~ more neve~smokets tea& those ye~ of ~e &~e~zed by high me'eft e~e expen~tur~. D~- ..... ference~ in d~ ~tes of smoke~ ~d n~u~e~s imp~t on medic~ cme cosu M ~o way,. On ~e one ~d. ~ere h a high cost ~ociated wi~ dyMg ~at is M~red ~er in ~e ~e sp~ for ~mokers ~d hm a pre~ent v~ue ~ed 1~ by d~couu~g, wher~ ~ose who live J longer in~r ad~don~ ~endimr~ in hter y~s bo~ ~o~ ~d d~edenu (fi~ze 2). ~e d~p~i~ in medi~ c~e spen~ng for older ~d younge~ per- sore, coupled wi~ ~e longer life ~pec~ of nevemmok~, r~ ' ~e of whe~er n~enmoke~ have higher lifetime medic~ e~endi- : rares ~ smokers, wi~ smo~g holding down medic~ cosu. In order to dete~e whe~e, smokem ox nevemmoke~ have ~gher lffec~e ex- penditure, it ~ n~ to ~e ~to account ~erences not only in ~e ~ect~. but ~o ~ me~ ~e ~d ~d e~en~ in.red during ~e y~ lived. :1! A g e.~p ~¢ifi¢ Expenditures In figure 2 we have age-spedfi¢ medical expendimre~ for male smokers and neversmoke~ a¢¢oxeling to whether one survives to the end of the T131060737

T131060738

Although expenditures for females toad to exceed those for mzles, th r~latio~hlps ob,cr~cd for m~s. in terms of s~,~rvivors vergus smokers versus ncvcrsmokcrs, generally hold for females also (figure 31 Expected Exl~enditures Applying probabilities of surviving and dying to mrvivor and deceder~ expenditures yields age-specific expected medical expenditures (figure These are the di~:ounted expenditures we expect the average individu~" aged 17 ro incur during each age interval for the temalntier of his or he life, according to whether the person is a smoker or neversmoker. A year-old deemed to be a smoker is one who is or will become a smoke; probably within five years. Females generally have higher medical penditure~ than males, but the relationship of expenditures to smokin is the same for both. Through age 74, smokers have higher expenditure at each age, but after age 75 neversmokers have higher expenditure: Here we see the impact of lower mortality rates a~d longer life expe~ Amount (thousands) 8100 Nev~¢~moker ~urv|vor "~ AI| smoker eurvlv~r II + &li tmo~r ~ecJdint $40 " TI310~073g

$15 $0~ I I , I t I I I 17-34 35°44 45-54 55-64 §5-74 75-1~4 ~5-o~lr FIG. 4. Expected medical expenditures of a 17-year-old during age intervab according to sex and smoking seams, discounted at 3 percent ....................... ~- tancy of neversmokers. Smokers who do survive to older ages have higher medical ~are cos~ (figures 2 and 3). However, because of lower probabilities of survival, so many fewer smokers compared with never- smokers live to age T5 or beyond that the gverage, or expected, expendi- ture that will bc incurred is less per smoker than per neversmoker. Expected expenditure is the proper conceptual m~asure for estimating average lifetime expenditures and is the basis for the analysis in the re- mzinder of this article. The influence of discounting on monetary values far in the future apparent. Ahhough discounting decreases the present value of all penditures, the impact is greatest on more highly discounted expendi- tures in later years. Because future ~pendimres l~:orac les~ important relative to more current expenditures, the impact of high expenditures by neversmokers relative to smokers in the later years of/fie on the gap between smokers and neve~nokers before age 75 is diminished. The net effe~x on smoker versus neversmoker lifetime expenditures of higher expenditures for smokers up to age 75 and lower expenditures T[31050740

$12 $10 $8 $S $4 35 45 55 65 75 95 Age FtG. ~. Cumulative e~t~ed exee~ medi~ ex~endirarea for a ~noker from age 17 to s~ltcttcl age~ ~c¢otcLing to ~, d~:ountetl at ~ percent. after age 75 is shown in figure ~. Figure 5 plots the cumulative excess (,moker minus nevers~oker) expenditures that smokers incur from age 17 to the age shown. The cumulative differeme reaches a peak at age and declines thereafter. The total of medical expenditures inoarted by male a~d female smokers remains higher than for neversmokers through- out their lives; the gap n~rrow$ after age 75 bur remains positive. The net Lifetime exce.~ expenditures for smokers ¢omparecl with neversmok- ers is shown at age 9% Over their lifetimes male smokers average more than nevermaokers and female smokers average $10,119 more. Total Ia'f~time Expenditure Total expected lifetime medical expcndimre~ from age 17 are higher for smoke~ than neversmokers and increase with the amount ~rnoked (ta- ble 2). Lifetime expenditures for male mmtetate smoken (fewer than 2~ cigarettes a da~,) for hospital care. ph~iciam' services, and nursing-home care average $52,891 in 1990 dollars, wkith i~. $~,61~ and 21 percent higher than the $27,2~6 for neversmokers. Heavy smoke~ (25 or more TI31060741

DLu:~am~d at 3 Fer~erit~ All SexN~e~oktl~ smokers smokers Males Lifetirn~ expenditures $27,276 $35,914 $32,891 $4o,187 Excem expenditures $0 $8,638 $5,615 $12,911 Ratio to nevcr~rnoker~ I.OO 1.32 1.21 1.47 Females Lifet Line expenditures $42083 $52.902 $48.918 $60.347 Excess cxpendkurcs $0 $10.119 $6,135 $17,564 Ratio to nevetsmokers 1.00 1.24 1.14 1.41 • VaJues are io 199o dollars. cigarettes a day) udlizc medical care costing $40,187, $12,911 and 47 percent higher than neversmokers. The average for all male smokers is $35,914, exceeding expenditures for neversmokers-by $8,638 and-32 percent. To the extent that self-reported dgazette consumption is under- reported (Hatziandteu et al. 1989; Wa~ner 1978), some moderate smok- ers may cor~ume more than 25 dgazet~es a day mad the least amount consumed by heavy smokers may be more than 25 cigarettes a. day. Females use mote medical car, at most ages and llve longer than males, and therefore have higher lifetime expenditures. The amount of smoker excess expehditutes is higher for females than males, but the ra- do of smoker to ncversmoker expenditure, is smaller for females. Excess lifetime expenditures are $6,135 for moderate smokers, $17,564. for heavy smokers, and average $10,119 for all female smokers. Lifetime pendimres are 14 percent h~gher for female moderate smokers thaa nev- ersmokers and 41 percent higher for heaw smokers, with average expenditures 24 petcent higher for all female smokers. To show the sensitivity of our results to the discount rate, we have also estimated lifetime expenditures discounted at 5 percent. At higher discount rates dollar amounts zre lem. but the disparity between smok- ers and neversmokers inctemes. At 5 percent, a,¢etag¢ smoker lifetime expenditure* are 37 percent higher for males a~d 31 percent higher for females. Tl310~07 2

The relatively smaller impa~ of ~moking on fem~.le expendimrr.~ co~tcn~ ~i~ lo~cr ¢~ ~kc ~rc ~ong fcm~l~ p~t. Lower exposure ~ulu in lower He~ and Humm S:~ic~ 1989) ~nd lower medi¢al et ~. 1986) relative to neversmokers ~ong femur. ~wer relative mot t~ity ~d medi¢~ ~¢ use in rush produce a smiler pro~a~oaate in creme in lifedme medlc~ expen~tures of ~mokers ¢ompited wit] neversmoke= for femur. ~ f~m~e ~posur¢ approaches ~at of tomes we ¢~ ¢~ect lffe~¢ =pen~mr~ of ferule ~okers to interne rela dye to neve=moker~. Excess Medical Care Versus Excesx Mortality Thus far we know smokers use more medical cite at all ages when ate alive than neve~mokers, whereas neversmokers live longer and us~ medical ¢uze over a greater number of yea~s. The impact of highe~ medi- cal care use while alive out-weighs shorter life expectancy and, on balance. male and female smokers have higher lifetime medical expenditure: ............. ihat~ ne'vdrsrnokers. The iati6 of smoker expendltures to neversmoke: expe,~dkuses in table 3 shows how smoker expenditures exceed n~er- smoker expenditures during each age interval up to age 75, where'a: neversmokers incur higher expenditures a£ter age 75. The separate contributiom of excess medical ¢ate and excess moruJiq can be appredated by comparing lifetime expenditures of smokers wirk lifetime expenditures of two hypothetical groups: (a) smoke= with meal. jtml care use of neversmokers (smoker expert, dirures.adjusted for medica. cite) and (b) smokers with mortality rates of nevetsmoke= (smoker penditures adjusted for mo~tallty). Comparing smoker lifetime expeadi- tures with expenditures adjusted for medical care, we "Observe the • cuntribution of higher medical r.ate use'to smoker expenditures. FoJ males and females, excess expenditures due to higher medical care me are highem in the middle Feats of the life span and falJ off rapidly dur- ing the later yeats. Excess medical ¢uze use of smokers in=eases thei: lifetime medical expenditmes by 43 perceive for mules and by 29 pereent for females. The impact of higher mortality rates on lifetime medical is obsen~ed from the comparison of smoker expenditures with expendi. tures adjusted for mortality. Up to age 6~ for males and age ~5 for re- T13106074.3

Smokers Adjured f~t ~¢~ M~di=I ~t~ Us:, ~d Smeke~ Adjmtcd for ~ce~ Mothy, by Age and S=x Age Ncve~moker Smoker Smoker mrdieal ~a:c monalhy Males Lifetime 1.32 1.43 0.95 17-34 1.27 1.27 1.00 35-44 1.65 1.64 1.01 45-54 1.55 1.5I 1.04 55-(~ 2.27 2.21 1.06 65-74 1.32 1.33 0.98 75-84 0.74 1.02 0.73 85 and over 0.48 1.01 0.47 Females Lifetime 1.24 1.29 0.97 17-34 1.31 1.31 l.O0 35-44 1.67 1.66 1.00 45-54 1.56 1.54 1.02 55"-~ 1.20 1.15 1.04 65-74 1.12 1.08 l.O,l 75-84 0.92 1.01 0.90 85 and over 0.69 1.08 0.64 • Smoker expenditures adjusted for excess raedf~M ~r~ ate expenditures of smokers assuming neversmoker, medicaJ care use. Smoker expenditures adjusted for excess raortdit? ate expenditures of smokcn a.muning nevenmoker mortality rates. Dixount rate is 3 percent. , males, smokers have as high or higher expenditures as the hypothesized smoker with ncvcrsmokcr mortality rates. This can be attributed to the high cost of dying, which, for this period of the life cycle, outweighs the smoker's lower probability of surviving to each age and incurring expert. ditures. At older ages, however, the cumulative effect of higher smoker mortality rates has so reduced the probability that a smoker will survive to these ages that expected expenditures arc much lower for smokers. Smoker expenditures decline rapidly with advancing age ~cladve to cx- TI31060744

smoker mortify,, given by ~: result for ~L[ smoker l~e~e ~endimt~ by ~ p~r ~d f~ smoker l~e~e ~endimt~ h~" ~ permit. Up to age 75, smoker ~en~mr~ exceed n~enm~ker ~most ~lely became of ~gher ~oker me~ ~e ~e; ~c~ mo~ ~d ~e high co~ of dy~g m~e a smdl coa~budoa. petted smoker ~n~mr~ ~e mu~ 1~ ~m tut~. At ~e older ag~ ~c~ smoker medi~ ~e ~ m~ a positive but gready reduced con~budoa to ~oker ~ndimr~, md ou~eighed by ~e redu~on ~ ~ndimr~ r~dng horn ~e ~pact at ~ese ages of ~e ~uladve effe= of ~c~ ~oker mo~i~ ~ prior ye~s. ~e net effe= ~ higher ~e~e e~eadimrm for me~ c~e for smokem Population Expenditures In the preceding sections I cx~nined individuals'.medical ca~e expendi- tures. These results enable us m ~.ssess the aggregate burden imposed by - dga~cttc smoking~ Each ycar~ mo~c than one million young pc~om in. the United States start smoking (Pierce ctal. 1989). And each yc~, ap- plying the results in table 2, decisions by young people to take up smoking commit the health care system to extra medical cazc expcndi- t'ures totaling $9~4 billion (discounted at 3 percent), spread out over the lifetimes of each new aop o£ smokcm Medical cxpcnditmc~ expected over ~c remaining lifetime have been csr.imatcd by smoking stares and sex for each age g~oup, from yca~ of age to 8~ yca~ and older. Applying these profiles ~f remaining lifetime expenditures per pe~oa to the civilian noninstitutionalizcd population 25 yeats of age and older residing in the United States in 198~ gives future expenditures attributed to the baseline population. The baseline population will generate medical expcndit'~rcs for 6~ years, at which time the last surviving members will bc at Ic'~t 90 yca~ old and the process is trunca¢cd. In addition to fut't~c expenditures attrib- uted to the baseline cohort of smokct~, we ~ calculate excess smoker cxpcnditurc~ or the amount by which ~nokc~ eapcndimres cxcccd pcndimtcs that would bc incun'cd if smokers had the medical catc use and morraliW of ncvctsmokcts. Emm these ~Iculations wc dc~vc the re- sults that follow. TI31060745