Anne Landman's Collection
BIRD-I A STUDY OF THE QUIT - SMOKING CAMPAIGN IN GREENFIELD, IOWA, IN CONJUNCTION WITH THE MOVIE, COLD TURKEY
Abstract
In August 1969 all the citizens the town of Greenfield, Iowa attempted to quit smoking as a publicity stunt connected with the on-site filming of the movie Cold Turkey. Philip Morris (PM) surveyed the citizens 8 months after their quit attempt. PM used local Girl Scouts to hand-deliver the questionnaires to citizens (to increase the acceptance of the packets). They paid $5 to everyone who completed and returned a survey. The Girl Scouts were instructed to knock on doors and give a questionnaire packet to "every person who was 14 years old on Cold Turkey Day." This report contains Philip Morris' analysis of the success of citizens' efforts to go "Cold Turkey" and the unpleasant side effects they experienced. PM's descriptions are entertaining, chauvinistic and, of course, paint a dismal picture of quitting smoking:
"Even after eight months quitters were apt to report having neurotic symptoms, such as feeling depressed, being restless and tense, being ill-tempered, having a loss of energy, being apt to doze off, etc. They were further troubled by constipation...As can be seen from Table 3, the...differences among male smokers were sizable, but the female data are the most startling. The anti-smoking campaign failed to persuade the women to quit. We can only conjecture at the reasons for the failure: --perhaps it is because women are better at running their husbands' lives then their own... --perhaps it is because busy housewives are less exposed to anti-smoking arguments, or less responsive to logical argument, or less apt to participate in community affairs...It is also possible that [smokers who] wish to stay off smoking have learned from experience that alcohol weakens their resolve. A sad picture is painted of the quitter who used to enjoy himself at a party, now restricted to coffee, fruit juice and coke, turning his back on the swingers in the kitchen in order to hover around the candy and peanut tray among the staid old gossips in the parlor. After one or two such experiences he probably quits partying altogether...The net effect of the extra food at mealtime and the snacks of candy, nuts, ice cream and coke had its predictable consequence: the quitters report more trouble with constipation and much more trouble with weight gain. This is not the happy picture painted by the Cancer Society's anti-smoking commercial which shows an exuberant couple leaping into the air kicking their heels with joy because they've kicked the habit. A more appropriate commercial would show a restless, nervous, constipated husband bickering viciously with his bitchy wife, who is nagging him about his slothful behavior and growing waistline."
User-Contributed Notes
Fields
- Quotes
BIRD is our in-house code name for the study of an intensive anti-smoking campaign in Greenfield, Iowa. Greenfield (population 2100) made an effort to go "Cold Turkey" as a community, with all smokers in town encouraged dto quit smoking abruptly at the same time. The procedure paralleled the anti-smoking effort depicted in "Eagle Rock, Iowa" by the film, "Cold Turkey", which was released January 30, 1971, by United Artists.
...For present purposes it is enough to recall that this became a major anti-smoking effort among a group of mid-Americans noted for their integrity and sincereity; that is was accompanied by miscellaneous strong social pressures which rewarded abstinence from and discouraged return to the act of smoking; and that it was accompanied by a spotlight of regional and national publicity unprecedented in the lives of the people involved. There have been other community-wide anti-smoking campaigns, but this one was unique because of the parallels with the motion picture, the presence of the actors and camera crew, the employment of many citizens as extras in scene after scene, etc. For over a month the lives of the people of Greenfield became almost inextricably interwoven with the lives of their fictional counterparts in "Eagle Rock." As film star Dick Van Dyke was to comment later on national television, 'It was crazy...you couldn't tell where the film let off and the town began."...
PROCEDURE
Distribution and Return Rate
The questionnaire packets were distributed door-to-door within the Greenfield city limits by local teenagers, most of whome were Girl Scouts. Each packet consisted of a cover letter, a basic questionnaire...a personality test...and a postage paid pre-addressed envelope. The girls, who were seldom turned away from a door as strangers would have been, gave out 1592 packets. Their instructions were to leave a packet for every person who has 14 years old or older on Cold Turkey Day. We estimate that there are 600 children in Greenfield beneath the age of 14, and the preliminary 1970 census report shows Greenfield's population to be 2212, so that it appears we did contact almost al the eligibles (2212-600=1612).
- Company
- Philip Morris
- Author
- Ryan, Frank J. (PM Scientist)Philip Morris scientist (circa 1973), developed smoker puff-profiles, compiled data on compensation and lip occlusion of ventilation holes. Associate Senior Scientist with PM, c. 1987.
- Recipient
- Cullman, Joseph Frederick III (PM President & CEO (1957-1970))Executive vice president and senior marketing executive of Philip Morris in the 1950s. Exec. VP 1955-57. President in 1958, held that position until 1967. Chairman from 1968-1972 and acquired title of CEO. Chairman of the Executive Committee, 1979-85. On the Board of Directors from 1954-1985.
- Weissman, George (PM Chairman & CEO '79-84)
Vice President of Philip Morris from 1954 to 1956. Vice President and Assistant to the President in 1957. Vice President of Marketing from 1958-59. Executive Vice President of Marketing in 1960. Exec. VP Overseas in 1961, Exec. VP PM International 1962-66. President from 1967 to 1972. President and Chief Operating Officer in 1973. Vice Chairman from 1974-78. Chair and CEO from '79-84 and on the Board of Directors from 1959-84. "Mastermind" of Philip Morris' direction.- Millhiser, Ross R (TI Executive Committee, PM Pres, 1968)
Ross Millhiser was Vice President of Philip Morris in 1952, VP and Director of Marketing at PM 1961-62, President of PM USA in 1970-72, President of PM Inc. in 1977, Chair of the Tobacco Institute Executive Committee and Vice Chairman of PM Inc. in 1979, Chairman of the Board of PM in 1980. The above information is gleaned from correspondence found within the Philip Morris collection of documents. The assumption is made that Millhiser worked at Philip Morris the entire time between 1952-1994, based on his correspondence during those dates, however nothing has been found in the documents verifying his positions at PM during the gaps in time noted above. President of Philip Morris in 1968 Millheiser was with Philip Morris in 1983 in New York. Knew that profitability of PM derived from addictive nature of nicotine. Why risk multi-billion dollar business for your rats, Victor.- Cullman, Howard S. "Hugh" (PM Inc., President 1957-67, CEO '67-78)
Howard "Hugh" Cullman was the brother of Joseph Cullman Jr. He was president of Philip Morris, Inc. from 1957-67, Chairman of Board and CEO of PM 1967-78, Chairman of Executive Committee of the Board 1978.- Goldsmith, Clifford Henry (B&H (1953), PM Chief of Operations ('65) Pres of PM, Inc. (')
1953 Benson & Hedges. 1965 Philip Morris USA Chief of Operations. 1969-73 President of Philip Morris, Inc. 1978 PM Chief Executive. Served on Tobacco Institute Executive Committee, 1979.- Landry, John T. (VP of PM 1970-76. Sr. VP '77-83, Dir. of Marketing '84)
Vice President for Philip Morris, Inc. from 1970-76. Became a Senior Vice President in 1977 and held that position until 1983. Was named Senior Vice President and Director of Marketing in 1984 and served on PM's Board of Directors from 1973-84. Served as memer of the Social Acceptability Working Party of ICOSI, c. 1978- Bowling, James Chandler (PM; TI, Corporate Affairs Director & VP; Board of Dir.)
Vice President and Director of Sales at Philip Morris from 1967 to 1976. He was Senior Vice President of PM from 1977 to 1984 and on the Board of Directors from 1971 to 1984. Also worked for the Tobacco Institute. Attempted to improve the image of smoking in the face of negative health news.- Lincoln, Jetson E. (VP Philip Morris)
Director and Assistant Director of MR, VP of Strategic Research, Director of Marketing-Planning Tobacco Production-Consumer Product Division, and most recently VP.- Seligman, Robert B. (PM VP of R&D c. 1976-82)
Vice President of Research and Development at Philip Morris Richmond, VA 1976-1982. Reported to Senior Vice President of Operations. In 1982 transferred to tobacco technology group. Wanted to share ammonia and other tobacco technology with PM International companies.- Udow, Alfred (Consumer Research Dept., PM c. 1972)
PM Consumer Research and Marketing Departments. Famous for intemperate written comments. Told the truth.- Fountaine, S.
- Pollack, Shepard P. "Shep" (PM President c. 1979-80)
Served on PM Board of Directors 1980-83, also served as PM chief of operations (circa 1980-82) and president of PM circa 1979-80. Served as CTR Treasurer.- Holtzman, Alexander (PM Asst General Counsel. 1975-85.)
- Wakeham, Helmut R. R., Ph.D. (PM R&D VP)
Vice President and Director of Research & Development, Philip Morris- Resnik, Frank Edward (Vice Pres., then Pres. and Chairman of Bd, Philip Morris)
TI Executive Committee. Prot駩 of Clifford Goldsmith. Vice President Philip Morris, Inc. from 1979 to 1984. President in 1984 and served on the Board of Directors from 1985 to 1989.- Thomson, Ronald H. (PM Europe President)
President, PM Europe- Fagan, Raymond (PM Principal Scientist c. 1968-83)
Principal Scientist at Philip Morris Research Center in Richmond, Virginia, between around 1968-84.- Eichorn, Paul A. (Worked with PM)
- Weissman, George (PM Chairman & CEO '79-84)
- Region
- United States
- Greenfield, Iowa
- Type
- Report
- Subject
- smoking cessation
Document Images
CGAIFlGZ1 J
T1A4
'e ruled out such"a procedure for the present study on the
grounds of expense and unavailability of interviexers. we had
rejected training local interviewers on the grounds that some
.
respondents would not tell local interviewers the truth, but
might tell the truth to an impersonal computer a thousand miles
away.
Therefore we wrote a simplt basic questionnaire we thougnt
Iowans could handle with relative tase, with built in skip-
statements for nonsmokers, nonquitttrs, etc. The basic question-
naire covered demographic data (25 items), a smoking history
(22 items), health and behavior change before and after ouit:{n;
(45 items), and a special section of 20 questions for students.
Personality scores wtee measured by a separate 137 item test,
the 16PF, which will be described in more detail in the next
report.
0
In return for returning the questionnaire,the cover letter
promised (1) to keep all names from the report and (2) to send
each respondent $5 and a brief summary of his personality prof';e_
(Appendix 7). Not everyone wanted the Personality Report, but
no one turned down the money: we received 1221 ballots from
citizens of voting age for our S5. In contrast, only 760 ballo:s.._-,
were cast in Grsenfield during the 1970 general election.
'ae might have done better using interviewers, but it's ha-:
to say. Our return rate was 90.1%. The re:;:rn ra:e of %a:ional
Analysts and the Opinion Research Cor9or3tion for the 1956 h::i
study was also 90.1:, an amazing coincidence.
~
02
C
00..
~
N
c.n
0 C
C.:
a
--k
N
2
_ 7 _ aa
0 c
~
w
1 ZZ
I
i
f
f
i
,
I
I
I
t
1
t
t
,
Processina the 4ata
The respondent's name did not appear on the questionnaire
when we received it, but the name was printed on the outside of
the return envelope. When the questionnaire was received it was
logged into a record book by name and number, the name was
written on the questionnaire, a check for $5 was clipped to a
thank you letter addressed to the respondent, and a code number
was placed on a Personality Report for the respondent.
The personality sst was then hand scored (using an ovtrlay)
and the personality scores entered on the questionnaire. The
Personality Report was then made out, the check was signed,
appropriate notice was placed in the record book, and the check
and Report was then mailed to Greenfieid.
The time required to process the data this far varied
accarding to the mait volume and work backlog, but it seldom
took more than five working days. A speciil effor't was made
to process the earliest questionnaires within 24 hours, for we
realized that getting a few checks back to Iawa quickly would
dispel any doubts there of the legitimacy of our research.;
In general this procedure proved effective, and we had a
minimum amount of correspondence and confusion about money and
questionnaires.
SA11 checks and corresoondence with Gretnfield bore the POL
letterhead. The sponsor's name was keot secret during the
investigation, although it was known in Grelnfield - and was
so reoorted in both the Adair County Free Press and in the
Des Moines Register -- that one of POL's clients was a tobacco
firm. No POL national panelist lives in Greenfield, al:hougn
ssveral live in Des Moines, and none of the Greenfield nar+es
and addresses were added to the national roster.
a-
.
iz3
CONF~~E11
TjqL
I
f
~
I
I
i
r
I
I
The questionnaireJs were next coded for keypunching. Eacn of
four coding experts processed a several page section of the
questionnaire, reading the answers and entering codes In appro-
priate "office use" spaces on the questionnaire.7 The cards were
keypunched and then verified, a slow operation.e
The data from each questionnaire filled four Hollerith cards,
Card I contained the questionnaire code number, the demographics,
and the special student section. Card II repeated the code-
number and showed the smoking history and some derived personality
scores. Card III repeated the codt number and showed the changes
in food intake, mannerisms, and health problems, plus transformed
personality scores. Card IV contained the code number, the
personality test raw scores, and several miscellaneous measures..
The data from the cards was transferred to a memory disc in
the XOS computer's disc file.
Computer printouts were then made for a number of variables
in which the number of panelists with each score on each column
of the card were tallied against several different classificat'on
codes present on other cards (see Appendix 8). This produced
such outouts as the number of college students who smoked
tht number of smokers with incomes over $10,000, et:.
7The System brokt down at this point - some coded answers on one
part of the questionnaire proved inconsistent with answers in
other parts of the questionnaire. A singie coder, rlading all
Questionnaires,wou1d have caught most of thest inconsistencies
at once and placed them aside. Instead, they continued in the
system causing all kinds of troublt later on.
30ur second major error was in believing that the verification
process was accurate. It proved not to be so, and we wert still
finding keypunch errors as tate as Oecember, 1970.'
9 -
...
.:
25a 12C10350
1 Zq
}
C
I
I
t
I
I
l
I
I
!
t
These outputs were examined carefully for inconsistencles and
impossibilities, and the inconsistencies were then resolved.9
Processing and~interpreting the data took much longer than
we had expected, even with the aid of the high speed computer.
We did not truly appreciate the magnitude of our task, nor the
wealth of output wt would mine, when w* began the study, for
there was no R & 0 computer then, and, in retrospect, we Nere
optimistically naive about computer capabiTities. The tolerance
and good will of the computer stiff, who were engrossed with
the problems of installing their unit and preparing it to
oaerate equipment all over the Research Center, was truly
remarkable.
0
9Txo examoles of inconsistencies: First example: Oo you smoke?
"tlo." What brand? "Winston." Second example: How tall are
you? "7 ft. 6 in." How much do you weigh?."314 lbs." Do you-"
particlpate in any high school sports? "No." In both cases
the cards had been mispunched and the entries wert a column
out of line. In more difficult cases the res0ondents had
actaally made the conflicting statements attributed to them.
We exoecteC to get some fictitious replies from the high scrool
S:udent3 as a practical joke, but we detected none. When we
had a conflict we attemoted to resolve it on the basis of o:her C
directly related questions. If we felt we really didn't xnoW
what the rtsoondent mtant, we threw his data out (thert wtrs
anty a few of these). If we had a bias it was to classify ttit ~
rss:ondtnts as Cold Turkey Day Quitters. Most of the questior.s ~
were quite straight forward, as were the answers, and over 99: C~
reouired no interpretation by us. When our intervention was
r4ouired it almost always turned out to be a case of an impro:er Z
punch.
10 Z501 ZDQ351
Ii5
i
I
f
,
I
I
R E S U L T S
Imolica*.ons of the Results
Thtre are three points to make before presenting the
findings.
First, we have data on such a large proportion of 6rsen-
field's population that our group averages can be considered
population averages rather than sample averages. In a very
real sense our data include all the knowable scores, for it
is unlikely that our non-rtsportdents would tver respond to
anyone else's questionnaire if they didn't respond to ours.
This means that it is unnecessary to employ inferential
statistics in describing the behavior of the people of
Greenfield, for we know their population parameters. It also
puts an extra burden upon the writer and the reader, for if -
all observed differences are real,.it is necessary that we
agret on the importance of a very small difference. In
gentral, we will ignore the trivial differences and put
emphasis only on those of some size.
Second, we are not convinced that it is always reasona5le
to generalize from the Greenfield data to Such theoretical
populations as "all American teenagers" or "male cigarette :
Smokers between the ages of 30 and 39" or to "people who give
up smoking." Greenfield is a small town. It can't be comoared
to the vast megalopolis which stretches from Boston to
aashington. Its people in no way represent a microcosm of the
U.S.A. Further a significant part of the treat:nent variable
involves participation in the movie, which is not reproducibie.-
On the other hand, to the extent that Greenfield is like the
rest of America, it may be possible to generalize to the rest
of America; to the extent that "Greenfield smokers between the
ages of 20 and 29" resemble "smokers in the smali towns of the
West ttorth Central U.S.A.," a more reasonable extrapolation,
Z5012[1a352
12-(-,
f
C
.
L
t
1
it may be possible to generalize to the latter population.
Furthermore, if Greenfield's people do differ from some other
population, that difference is unimportant for our purposes
unless it can be shown to be related to smoking in some way.
For example, if Greenfield's people are taller than the U.S.
average, but smoking happens to be independent of height, the
admitted height difference would not preclude making general-
izations about their smoking. It can be seen the arguments In
favor of and against generalization from our data art not simple
and require soa+e prudence and common sense on the part of the
generalizer.
about quitting, or about personality.
Third, where it does seem reasonable to consider the
Greenfield data as a sample of the U.S., and we make inferences
about a population from this sample, we must be particularly
careful about making "Type I errors." (Youmake a Type I error
when your statistics lead you to believe two groups are differ-
ent, when In fact they are'not.) For example, we have 20
personality variables, 30 subtables for each of the personality
variables, and over 20 column averages per subtable. We could
compare over 120,000 pairs of column average scores if we were
so inclined. By chance we would expect 6000 of these differ-
ences to be so great that they would yield significant F, t,
or x2 statistics at the .05 level of confidence. Obviously,
then, we cannot just browse through our tables looking for
differences among column means, for many of these will be due
to chance alone. Instead, we must restrict ourselves to test-
ing hypotheses founded on some prior notions about smoking,
r
IZ?
CC,ilF
Jr~q~
r
I
I
I
I
I
I
I
t
I
I
Smoker Psrticioation.
Among the 1335 usable questionnaires we classified 444
as active smokers ia May a figure slightly higher than the
426 referred to in a Des Moines Register article (February 22,
1970), and just below the 461 estimated from a telephone survey
tiken on Cold Turkey Eve. (Our non-usable questionnaires seem
to show an extra 11 smokers, for a total of 455.) 'ahat
happened to these 444 smokers is shown in Table 1.
TABLE 1
Smoker Particioation in Camoaion
Total Number of Smokers in May, 1969 444
a) Signed pledge to quit 178
Did not pledge 214
Didn't answer pledge question 52
b) Made a real gffort to quit 137
Made half-hearted effort to quit 145
Made no effort to quit 159
Didn't answer effort question 3
c) Consider effort successful, stopped for while 96
Smoked much less 35
Smoked a little less 79
Smoked as much as ever 115
No answer (117 of whom didn't try to quit) 119
d) Classified by deed, rather than effort or pledge:
.100:
40r
43.
.
12:
Didn't try to quit completely 273
Ouit in June and July 29 ';
Quit on Cold Turkey Day 137
e) Total Nho ouit in June, July, or on CTD 165 i^.0 ; ~:; ;
?esumed .+i tt+i n 3 weeks 74 4s. S; C 1';
Succsssfut Earned 10 button 32 55.=; (= 20.7;
Successful - Off for 8 months or more 47 2,3 ,.:;CJ 10,:;
f) Cigarette smokers in may, 1969 394 100:
Didri't try to quit 231 64:
Ouit in June or July 23 5:
Ouit on Cold Turkey Day 120 :0;
Resu~+ed ~+i:hin 3 ~reekS 47 12;
:.arned 10 button 73 1a..,
7ff for 8:nonths or more 'a 9.7;
v
2501200354
t
Z~
F
I
I
I
t
I
i
CONFjCE,vTIqL
not all smokerswert willing to sign the piedge to quit
smoking. We have Identified only 173 smokers who answered "1es"
tq our question, "Oid you sign a pledge?" The remainder either
said "no," or did not answer the question. Some of the latter
may have honestly missed the pledge question, as it was embedded
near some pipe smoking questions, or they may have been avoiding
an unpleasant admission that they had broken their pledge. At
any rate, 214 of our respondents deny signing a pledge. These
figures do not jibe with the earlier local figures whicb
indicate that only 127 of the active smokers did not sign a
pledge. We doubt that the local figures were Inflated inten-
tionally. It seems more likely that the local totals contained
some former smokers who were pledging not to resume, etc., and
that some of our respondents have either forgotten signing a
pledge or do not now consider that they actually made a pledge,
whether they signed one or not. Sometimes it''s hrard to take
seriously the act of signing a paper pushed At you by a pretty
young girl in a'crowd of holiday ceTebrants.
The smokers split fairly evenly into three groucs: those
who recall making a real effort to quit, those who recall mak4ng
a half-hearted effort, and those who made no effort at all. for
appearance sake any of these smokers may have cut back their
public consumption out of respect for the image the town was
trying to attain, and some in fact acpear to have concentra:ed
their effort not on quitting but on cutting back their consu:na
tion. Many smokers appear to hive taken the HE'r!recommenda;i :^ Cj
about "smoking less if you can't quit." Still ottiers were
unable to find their exact case in our questions and ans:Ars.
We find, for e~amote, that 6 people who say they didn't stao on
Cold Turkey Day atso.say that they made a real effort to vui:.
Ti,ese ptople are anomalously categorized as "Didn't try to qu':."
-14 -
2501200355
1z5
Cc^~~.~~E,l TJqL
I Table 2 showS the relation between success, degree of efsort,
and stopping among people who were still smoking on the day before
Cold Turkey Day.
TABLE 2
I Effort, Stoooina, and Success
-
~ .
I
Didn't Go Stopped
On Cold Resumed
Within Earn ec
IQ
Cold Turkey Turkey Oay tks
3'a 9u ?~
I Maonitude of Effort ~
Real Effort to Quit 6 103 42 61
i Half Hearted Effort to Qu i t 112 32 28 4
No Effort 158 1 1 0
No Answer to Effort Question (2) (1) (0) (1)
~ Total 278 137 71 ' 66
~
Success of Effort
~ I Quit Completely CTO
1
6
I Smoked Much Less 1 22 19 3
r I Smoked a Little Less 53 25 23 2
~
~ I Smoked as Much 97 13 1a 0
flo Answer to Effort Question (117) (2) (1) (1`,
Total 278 137 71 65
I It is obvious that respondents had different interaretat~ons
,
of our question 065, "During the camoaign, was your effo.r:
successful?" Some people who elsewhere had reoorted that they
had quit for the month here said that they smoked a little less
but bypassed the answer that they had quit comaltt:ly. Our ca::-
r-in their effort was less than a complete success.
~.r
.C
C
C
C..?
button and subsequently resumed smokinq, so that in the long .:~k
3ory for peools who quit for a month is "Earned IO 3ut:on." ,te
don't know whether these people really earned or claimed they
earned an IQ button. We suspect that they did earn their IQ
C
Z501200356
,30
In light of these comments, the actuarial data of Tabie I
on the success of the campaign should be modified slightly:
considering those who did not stop. but made a real or at least
half-hearted effort to cut back thefr consumption, as well as
those who stopped early or on Cold Turkey Day, we find that 284
of the 444 smokers, or 64: of the smokers made some effort to
quit. Of these, 166 actually stopped on or before Cold Turkey
Day (33.4% of all smokers and $8.55 of all who were trling to
quit.)
In reporting participation to the Mayor of Greenfield and
the editor of the Adair County Frea Press we equated Cold Turkey
Quitting with stopping. We.should recognize that particiaation
in the campaign was greater than the stopping data alone would
suggest.
0
On, the other hand, if the percentage of participants was
higher than the ttoaping dati Suggest, the percentage of IQ
button earners was lower. The data of Tabie le show that 56wo
of the quitters received IQ buttons. Based on the larger
number of peopie who made at least a half-hearted effort to.
quit, only 32.4: received IQ buttons. Similarly, if we examine
those still not smoking after 8 months, the figures must again
be modified. In this case we would have to use a differllit ~
Count, for some 16 of those who didn't quit during the campaign
either subsequently quit cold or,tapered down to the point where
they could quit. This brings the total "sti11" off to Gas:er :o
63, (,22.2t of thoSe who made any effort to quit.)
~..
As in the case of the optimist who sees the Cake portion of
NOW
the doughnut and the pessimist who sees only the hole, .rhich set W
of fi?ures we wish to emphasize will depend in part on wna: we .:.
Wan: to ses. ~
c
L
~ft
i3)
