Tobacco Documents Online

PHILIP MORRIS U,S.A. RESEARCH A R C H C E N T E R DATE: March, 1978 ANNUAL EXIT-BRAND CIGARETTES: A STUDY OF EX-SMOKERS F. J.j Ryan SUPERVISED BY APPROVED BY F. Resnik A. Udow N'. Hblbert J. Zoler R. Sel i gman T. Osdene H. Daniel R. Fagan Dr. F. Gullotta Dr: C. Levy Ms. P. Martin Mr. P. Faust Iir. N. Johnston Day File (3) Central File (2) Indexer KEYWORDS: Exit Brands, Quitting Data, Nicotine Delivery, Low Delivery, ;Smoking Habit, Smoking Behavior :_;•~:~~ --at

.as } Z' 11tt,it' Mt)RItIS' U. S. A. ~ ° ' RESEARCH CENTER , - 'ACKNOWLEDGMENT RECEIPT - CONFIDENTIAL REPORT t.. ,z, . REPORT NUM$E7R • 7 $ -Q'75 - ~ DATE OF REPORT 3/ 7 8 nnE ; EXIT-BRAND CIGARETTES: A STUDY OF E):-~?SOE.~R TWs:eE°CF*~It:: V*:,ll.hll ro TI/E _}l'"fAt S1~ ~M~ 11/( i OMD~ANY;.11 i$/1(1litD ~ ~N~ cARtrnttY tNNNptttl kND IS~,NIt -'~ 1RiN?~t't~Rlt ~ . It. nd. Attn1RT tUS $tRYth~its~rt!R:~ Mn 4Np lL No WM, t1R' NFFPCD„ Rf 4tt RI I I IRN 11 165AU {14111 l1 1 0, tt/E Rt>f.kqidi i tNit R IJ 1ttE A1 V N1 kYN (N ltNtk+,l tlltJt.i,,.,:;'t":~t:':-.'. .

This survey of people who have q,uit.smoking,foundithat the proportion of quitters leaving the market via low-dellivery cigarettes is 1.4 times as large as the market share of those brands would suggest. The r_eason for the disproportion is unknown. 'fhe most likely explanation lies not i'n the delivery of.nicotine and tar but in the characterastics of the,people who have shifted to low delivery`.' people who are concerned about•their health and have already shown their concern by.shifting to low delivery. The brand most often used as an exit brand is Marlboro (full flavor) as .is befitting,the brand which is smoked by the most people. Winston (full flavor), the national!second best seller, was second most likely to be an exit brand. Marlboro and-Wiinston, however were less likely to be used as exit brands than their market share would have predicted. Among the l'ow- ters, making market share comparisons risky. A few brands appear to con- market share would predict. No other low-delivery brand had over ten quit- delivery" brand) followed'by Vantage, the second best-selling,low delivery. Kent was as iikely and Vantage more likely to have beenian exit tZrand than delivery products, Kent was most often mentioned~(it's the best selling "low- ; involved are so small •that to mention names in a summary would~be misleading. tribute twice as many quitters as market share would predict, but the numbers The princiipal reasons for quitting are health related, with the general health scare most often mentioned. Social pressure from husbands, wives, children, coaches and others account for about 100/ 0 of the cases. Specific illnesses thought incompatib1le with continued smokiing account for about 9c%, of the quitters. Pregnancy was mentioned by 40, doctor's orders by another 310 . The only other recurring explanations worth mentioning were cost (7`.) soo248os'7s the question "Is it easier to quit

0 r I \ ` INTRODUCTIOLN Are smokers.more'likely to-quit from one type of cigarette than from another? This question's importance grows every year as the pressures to` 44, quit smoking~increase and the'number and type of cigarettes proliferate:;; • . ., . . _ _: ... _.: t. We think that most smokers can be cdnsidered nicotine seekers, for the :. pharmacological effect of ni~cotine is one of the rewards that come from ~4 smoking When the smoker quits, he foregoes his accustomed'nicotine The ,~ . ,r _.._ :change is very noticeable,`he misses the reward, and so he returns to smoking, One treatment which antismoking agencies have recommended is a nicotine ~~ ;:~ r 2., ., ,. _.. . -. - . weaning procedure, Zn which the smoke rewards are gradually adulterated'until; , . - . . _ : : -,. .-. . . . . , _ . .: . . . . . they are barely noticeable.': Then the transition to no reward is easier r. because it is hardly missed.' The stop-smoking kit marketed as "one step at introduction .: Over recent years the gradual reduction in nicotine delivery of the , major brands, plus the introduction of many low-delivery brands and brand extensions, has lowered the nicotine available to smokers in a manner which .,....,.. . ,_ • If the industr s int y' roduction of acceptable lo~N nicotine products does make it easier for dedicated smokers to quit, ,then the wisdomtof~~~ ~the ~: ,, .F.. a time" is basedon this idea you smoke your favorite cigarettes through a series of ever more efficient filters until finally you are getting almost' no tar and nicotine.at all. Then, to the extent that nicotine is the dominant reinforcer in":smoke, complete cessation should be less painful than it would otherwise be. paral,lels the one-step--at-a-time procedure. The industry's.new found abilit. y : to make flavorful products in the low delivery range, like Meri't, has made it possible for smokers to switch to'low nicotine products which are acceptable in~taste and hence willl not be rejected out of hand by people who try them. But the nicotine deliveries of these products may be low enough to constiltute a partial weaning of the smoker. As he adapts to ever lower amounts of r'~ r'~.±~ nicotine, he may now fiind' it easier to quit entirely than woul-d have been =~- R. I ,

P. IN " a Therefore we+need to know whether it is easier to quit smoking from ~ . r. low-deUvery brands y The best way to study the problem would~be in a -controlled'experiment There are many possible experiments. For example, " we could randomly select a large number of Winston smokers, switch a randomA _ . .: hal f of them to Meri r ' t and half to Marlboro, keep them on the new cigarette ' for a while, and then urge all of them to quit.. The critical measure would be the difference in.the number of successful quitters from full-flavor, f~ high-nicotine Marlboro and flavorful 1ow-nicotine Merit. ~ agencies, fund'-raisi.ng organiizations,the American Cancer Society, Smoke Watchers, and a host of other people, have been,urging smokers to quit entirely. In contrast to the experimentall situation, all these phenomena , An approximation to this experiment has in fact been gaimg on in the ~~ marketplace for some time-Tobacco company advertising has been switching `) .many full-fl!avor smokers to low-dellivery brands, while still other advertisi~n .. . i.i7.~ .has beeniswitching themito other full-flavored brands or trying to keep them_ from leaving'their full-flavored brands. Meanwhile, physicians, government J .have been taking place concurrently, so the analogy is not perfect. Further- more, there is another very important difference between the marketplace and N the experiment: in the marketplace the people studied are not randomly selected, and they are not randomly assigned to groups.- Instead, they are M1 self-selected and self-assigned, assumedly for reasons of their own. It is l i kely, for exampl e, that the popul ar bel i ef that l ow -tar ci garettes are "hea1thier" than full-flavor cigarettes means that people who are concerned about their healith wil'1 be more likely to switch to low-tar products than st ~ people who are not concerned about their health. "People who are concernediabout their health" is not a good phrase, purposes. It is meant to encompass those who are hypochondriacs, those .. . . . .. . . . . • " -f" .. ~..~.fi'. x f .).a _ for we all-want good health, but the phrase will have toldo for present ..- p24809'78 , . xi R. 4- 'lY~a4 .~ W.'+L(r .- ,V;

who have real illnesses or medica1 probl!ems such as allergies, or pregnancies; those who have recentl'y been touched'~by the death or illness.of•friends or family, those who are+growing~older and beginning to~feel the physical symptoms of middle and old age, and probably the largest group of all, those who are.convinced that smoki'ng is dangerous to their health and who are torn: between a conscience that-urges themto quit and a hedoni'stic desire to _ .. ,.. _ . The very fact,'then,~*that a smoker.has decided to switch from a full.-flavo . .;,, cigarette to a lowLdelivery'cigarette tel,l's us something very important abou"t . .. . . .. .. . N _ - . .. . . . . . . .. .. . him: he is concerned about his health, and he is wi'lling to do something abou In the experiment used asanexample, the smokers would be assigned~ by chance to one group or.another, like cards being dealt from a well-shuffled ` deck. But in the marketplace, the deck is stacked. The people who are most ;~ rYTL. likely to quit have already assigned themselives to one of the groups, and it,_~ is therefore extraordinari'l!y difficult to draw any conclusions from the marketplace data. Unless, of course, the whole hypothesis is, wrong. In this,case, there ,; would be no differential quitting from the two types of cigarettes, and some , serious rethinking would'.have to take pl'ace. Initfie early spring, of 1977 in conjunctiion with the national tracking study, a telephone survey which monitors the frequency with which smokers change brands as well as the number who are brand loyal, we asked nonsmokers (inihousehol'ds without smokers available) whether they hadiever been smokers, and i f so, how 1 ong ago they had qui t, why they had qui t, and what brand they had last smoked, as well, as their age and sex. In discussing the answers, : _2 first we willi cover differential quitting from selected nonmenthol filter .brands, as measured by the "quit ratfo" (see below), then the trends in quitting in recent years, the quitting from different types of cigarettes during the past 5 years, the age at quitting, the reasons given for quitting and finally the role of nicotine in ease of quitting.

RESULTS AND ~ISCUSSIOPd The Quit Rati.o We will define a brand's Quit Ratio as the ratio between two percentages:.; I held by Brand A, then the Brand A's Quit Ratio is 1.00. If.Brand B has 12`0 of: share. If 6% of a1l quitters come from Brand A and 6% of the market share i's. the brand's'percentage of'the identified quitters divided by the brand''s market s the.quitters`and 4% of the market, its ratio is 3.00. If Brand C has 2,°~ of the .-„ quitters and 4% of the market, its ratio is .50. High ratios mean.more quit- .ters than expected from market share; low ratios mean fewer quitters than expected from market share. ' We took a close look at the nonmenthol quitters of the past year and' , .. broke the data out for selected brands. Then we found these brands' share of the nonmenthol market from the Maxwell Report. Because the numbers and shares are often small, there is consi,d'erable inaccuracy inherent in the counts and percentages obs'erved. This should be considered in evaluating the results which are presented'in Table 1. The first thing to notice is that almost two-thirds of the quitters come from full-flavor products with Marlboro most often named!as an exit brand. This is as would be expected from Marlboro's position as the national1 sales leader. Second most often named was, of course, Winston - the number two brand~. The most frequently named "1!ow-delivery" cigarette was Kent, the largest- selling iow-d'eiivery brand. (At the time of the interviews Kent had not been .repositioned to~ 13 mg~, but we havealways consIdered it asliowdelAverywhether it actualily de1ivered more or less than 15 mg.) The second'most often mentioned low-deltvery product was Vantage, the best-selling cigarette then under 15 mg~. It does not appear that any individual brand should be singled out as an exit brand~in the sense that significantly more people quit from it than would be expected. However, a few brands appear in the quit list more.often than

u True Merit Kent Golden Lights Do ra l Winston Lt. .Marlboro lt. •Vantage : ~ .. All Parliament All Other Kent A71 Other Low Delivery Total Low Celivery Fulll Flavor Al1' Other Marlboro All Other Winston B&H 100 Virni!nia Slim All Ot'herFul l! FlavorTotal, Full Flavor t ~ TABLE 1 NUAIBER OF QUITTERS FROMSELECTED NONMENTHOL FILTERS AND THE BRAND'S % QUITTERS AND % MARKET G~. v - . . . . ., " . .. All 1-3 4_11 1 in Mos. Mos. Yr. Yr. 7 10 13 6 5 5 16 8 2 1 ll 41 31 .19 91 28 24 7: , 59 18 21 15 54 .5 5 .8' 18 3 2 2 .7 12 9 11. 32 66 61 43 170 % Nonmenthol % Nonmenthol ~~ Filter Quitters Filter Market ~ Qui in Year in Year Ya~ Ra' 3.4 1.6 .4 .6 5.0 3.4 1.1 2.5 6.11 6.1 4.2 2.8' 34.9 24.5 s e 9® ® l , N , >

their market share would suggest. It could be said that over twice as many people quit from Carlton as the market share would suggest. The trouble is . . . ~ ~ . . . . . . . . . . .r fi. that you're talking about 9 people instead of 4.2 pebple. Such ratios are bound to be misleading when based on very smalil numbers. Further, it is. ;' ._probably not right to infer the number of smokers from the sales market share. :ihat would be a particularly bad error if people who are apt to quit are also apt to smoke only a few cigarettes. In the Dresent case it would ` be a.mistake if low-delivery smokers either typically smoke more or fewer . : . . . . _ . ,~; cigarettes per.day than full-flavor smokers, because that would mean the., brands actually had fewer or more smokers than the market share would sugges Witfi .this caveat in mind, we can see that the low-delivery nonmenthol -brands in generaT were about 1.42 times as likely to be represented in the quitter column as their sales market shares would suggest: 34.9% vs. 24.5%. (Im a recent R&D study of POL brand switchers and quitters Peggy Martin'. data i'ndicate about 50 low-delivery nonmenthol quitters from 1808'contacts* " and~112 full-flavor nonmenthol quitters from 6748'contacts. Among these 162 quitters 30.9-% came from low delivery, as did 21.1% of the contacts. The ratio of 30.9 to 21.1 is 1.46. Thus we have two studies cond'ucte6in very a different ways, which yield about the same quit ratio. This may be chance, ~- but it is more likely to reflect a basic common phenomenon.). There were 459 people who had quit smoking within the prior year, of whom 455 offered enough usable data to be consideredi. Of these, 182 had quit within the last 3 months, 93 had quit 4-6 months earlier, 54 had quit 7-11 months earlier, and 126 had given up cigarettes "about a year ago." An additional 689 smokers had originally quit between 1 and 5 years earlier, yielding 1144 ex-smokers who had quit duriing the period marked by the rapid growth,of "low-delivery" cigarettes. There were 1882 other ex-smokers identified, all of whom had quit "more than five" years prior to polling. That made 3042 ex-smokers from the 9789 nonsmoking contacts, or 31% of the nonsmokers.(See Tabl es 2 and 3.) 1002480982 . ., - .. . . . ~ 'Y k: *A "contact" was a person sent sample cigarettes. In this case the,"quit ratio" is based on the proportion of people who said they had quit divided by the proportion of contacts for a brand.

TABLE 2 , NUMBER OF QUITTERS IN RECENT YEARS Time Period - ` N . % Total r: f Within a year 455 15.0 1-2 years ago 203 . . . ' ^M 6 7 ~ „ 2-3 years ago 202 . J 6.6 3-4 years-ago 151 5.0 ~ 4-5 years ago 149 4.9 `~ ~ Over 5 years ago 1882 61.9 ; Total 3042 100,1 V~ 0 Obviously a large percentage of all quitters were recent quitters, and there seemed over twice as many last year as in recent,years. But it .is difficult to interpret these numbers, because experience in dealing with;'quitters" suggests that most people who quit smoking will. resume after a while. 'Hunt and Matarazzo show data suggesting that 50% of quitters resume smoking within 3 months and 70% resume withiln a year (p. 76, Learning Mechanisms in Smoking, W. A. Hunt, ed.). Their percentages are 'based on follow-up studilEs of smokers who attended clinics o\were given various antismoking psychological treatment sessions. Such quitters obviously have more at stake than our (probably untreated) population, yet many of them ultimately resume.* Certainly the ranks of the 182'who had quit within 3 months contain many who would not still have beenex -smokers after a year. The apparent increase in number of recent quitters is, we therefore conclude, deceptive, and shcuTd not be interpreted as meaning that the actual quit rate had' increased during the time perilodijust before poliling. Q .3 © Turning to the subject of the brands and delivery levelis, we can compare the proportion of quitters among all! 80, 85 and'100 mm filter *On the other hand,, they may have represented the most difficult cases, else why would they seek help?' ,C..y, :~:;Y&-;~~-'-5~l-L ~ . y _ Am M , a