Abstract
This internal Philip Morris (PM) memo from 1975 shows the extreme importance of youth smoking rates to sales of PM's flagship brand, Marlboro. The author describes the importance of "the teenage years" to the cigarette industry:
"...my own data, which includes younger teenagers, shows even higher market penetration [of Marlboro] among 15-17 year-olds. The teenage years are the most important because those are the years during which most smokers begin to smoke, the years in which initial brand selections are made and the period in the life-cycle in which conformity to peer-group norms is greatest."
He then describes how youth smoking has caused a "phenomenal" growth rate for Marlboro:
"Marlboro's phenomenal growth rate in the past has been attributable in large part to our high market penetration among our younger smokers...I pointed out that the number of 15-19 year olds is now increasing more slowly and will peak in 1976, and then begin to decline..."
Other parts of the document show PM's disappointment at lagging sales of Marlboros, particularly to youth, for example the discussion of "why things fell apart" in 1974, when sales of Marlboro fell below what they predicted based on the number of 15-19 year-olds.
Fields
- Notes
This document was found in the area of Thomas Osdene of Philip Morris.
- Quotes
To: Dr. R.B. Seligman
From: Myron Johnston
Date: May 21, 1975
Subject: The Decline in the Rate of Growth of Marlboro Red
I think Dr. Dunn's memo has very effectively dispelled the notion that nicotine reductions have been the cause of the slackening rate of growth of Marlboro Red...
In view of my fields of interest...Marlboro's phenomenal growth rate in the past has been attributable in large part to our high market penetration among our younger smokers and the rapid growth in that population segment. I pointed out that the number of 15-19 year olds is now increasing more slowly and will peak in 1976, and then begin to decline...
In my opinion, the delcine in Marlboro's growth rate is due to four factors:
1. Slower growth in the number of 15-19 year olds
2. The recession.
3. Price increases in 1974.
4. Changing brand preferences of younger smokers.
Obviously, we can do nothing about factors 1 and 2 and have ony partial control over factor 3....
DEMOGRAPHICS
It has been well established by the National Tracking Study and other studies that Marlboro has for many hears had its highest market penetration among younger smokers. Most of these studies have been restricted to people age 18 and over, but my own data, which includes younger teenagers, shows even higher market penetration among 15-17 year-olds. The teenage years are the most important because those are the years during which most smokers begin to smoke, the years in which initial brand selections are made and the period in the life-cycle in which conformity to peer-group norms is greatest.
It seemed reasonable to believe, therefore, that there should be a relationship between the number of 15-19 year-olds and Marlboro sales during the period of Marlboro's rapid growth. I started running some correlations and found some interesting relationships...
1. The correlation between Marlboro sales and 15-19 year olds as a proportion of the total population is .993. Projected into the future, this shows Marlboro sales peaking in 1976 and then beginning a decline.
2. The correlation between Marlboro sales and the number of 15-19 year olds is .996. Projected, this shows Marlboro sales peaking in 1976 and then declining...
...While these correlations are certainly more favorable to Marlboro's future than the correlations of time-series data, they also show that heavy reliance for sales on an age group that is declining in number is dangerous. Another interesting observation is that in all of these correlations 1974 fell far below the trend line, i.e., below what was predicted for 1974 by the effect of 15-19 year olds. The question, then, is why things fell apart in 1974.
The Recession
...In my economic forecast, you may recall I said that Marlboro was particularly vulnerable to the effects of the recession because the highest unemployment rates are among the younger age groups, precisely the groups in which Marlboro's penetration is highest. IN the first quarter of 1975, the unemployment rate of persons 16-19 years old was 20.4 percent, the highest it has been since the Bureau of Labor Statistics began compiling rates by age in 1948...
Lower income people...feel the decline in real income more more acutely, and...they are more liekly to quit smoking or reduce consumption. Marlboro smokers, being on the average considerably younger than the total smoking population, tend to have lower than average incomes. Thus, I would expect a disproportionately large number of Marlboro smokers to quit smoking or reduce daily consumption. In addition, young smokers are less habituated than older smokers, and can therefore probably quit or cut down more easily than an older smoker. Furthermore, many teenagers who might otherwise have begun to smoke may have decided against it because of the adverse conditions...
Changing Brand Preference of Young Smokers
A much more serious factor is the apparent decline in Marlboro's popularity among the young...
...[T]here is cause for concern, partly because college students do constitute a significant share of the Marlboro market, and partly because college students' behavioral and attitudinal patterns tend to carry over to non-colleg eyouth with a one-to-three year lag...Thus, the decline in the popularity of Marlboro Red among younger smokers will probably continue and, thus, further reduce its rate of growth.
CONCLUSION
...If we consider futher the reported decline in Marlboro's popularity among the young, we are left with the unpleasant but very real likelihood that Marlboro's market share will peak in the very near future, probably in 1976 or 1977. The company will then have to rely on other brands for growth...
- Rank
- 1
- Company
- Philip Morris Cos., Inc.
- Author
- Johnston, M. (Philip Morris)
- Recipient
- Seligman, R.B. (Philip Morris)
- Region
- United States
- Litigation
- This document was used as a trial exhibit in Florida, Minnesota, Texas and Missouri.
- Type
- Memorandum
- Subject
- market forecast
- target market
- youth
- children
- consumer market
Document Images
Page 1: qfb12a00
Log in for more options!
Ir
.,,
.
. nicotine reductions have been the cause of the slackening in the rate of
PHJLIP'MORRIS U. S. A.
INTER-OFf I C E CORRESPONDENCE
RICHMOND, YIRGINIA.
Dr. R. B. Seligman - Date: May 21, 1975
Myron Johnston
The Decline in the Rate of Growth of Marlboro Red
I think Dr. Dunn's memo has very effectively dispelled the notion that
recession.
also hypothesized that Marlboro would be particularly vulnerable to the
in my 1975-1980 Economic Forecast. It was my contention that Marlborof s
phenomenal growth rate in the past has been attributable in large part to our
high market penetration among younger smokers and the rapid growth in that
population segment. I pointed out that the number of 15-19 year-olds is now
increasing more slowly and will peak in 1976, and then begin to decline. I
for the decline in Marlboro's rate of growth. Indeed, I treated these factors
to investigate the economic and demographic factors that could be responsible
In view of my fields of interest, it should come as no surprise that I chose .
growth of Marlboro Red.
r, -
In my opinion, the decline in Marlboro's growth rate is due to four factors:
1. Slower growth in the number of 15-19 year-olds
2. The recession
3. Price increases.in 1974
4. Changing brand preferences of younger smokers
Obviously, we can do nothing about factors 1 and 2 and have only partial control
over factor 3. (State taxes are beyond our control, for example.) Let us look
at each of these factors individually.
Demographics
made, and the period in the life-cycle in which conformity to peer-grou
smokers begin to smoke, the years in which initial brand selections are
years are also important because those are the years during which most .-
higher Marlboro market penetration among 15-17 year-olds. The teenage ~
It has been well established by the National Tracking Study and other studies
that Marlboro has for many years had its highest market penetration among
younger smokers. Most of these studies have been restricted to people age
18 and over, but my own data, which includes younger teenagers, shows even
norms is greatest.
Page 2: qfb12a00
Log in for more options!
Page 2
t
some interesting relationships, all significant at better-than the .005 level:
of Marlboro's rapid growth. I started running some correlations and found
between the number of 15-19 year-olds and Marlboro sales during the period
It seemed reasonable to believer therefore, that there should be a relationship
'
... i . . ' . 1~=.~.. . :
f ' 1 1 ::;' .. ' .r.. ~ ' .. . . . . . .. ~.:. .:.1~.... ~ a ...~, ..
gating year- to-year changes in the relationship of Marlboro sales and teenagers.
The correlation between Marlboro sales and 15-19 year-olds
as a proportion of total population is .T93. Projected into
the fu re, his shows Marlboro sales peaking in 1976 and
then beginning a decline.
The correlation between Marlboro sales and the number of
15-19 year-olds is .996. Projecte~s shows Marl.~~oro
sales peaking in 1976 and then declining.
. for 1975, a peak of 12.21% in 1976, and then a decline.
15-19 year-olds as a proportion of o al popula Tion is
.9997. Projecte , this s ows a 12.0 o share o market
The correlation between Marlboro share of market and
The correlation between Marlboro market share and the
number of 15-19 year-olds is .99985. Projec led, this
s iows Marlboro market share peaking at 12.92% in 1976
and then declining.
It should be noted, however, that these were all correlations of time-series
data, and correlations of time-series data frequently yield correlations that,
while statistically significant, may be spurious. Accordingly, I began investi-
Again, I found some interesting results:
1. The correlation between percent change in the number of
15-19 year-olds and the cHange in Marlboro maref siare
is .963. This shows Marlboro market share peaking a
13.34% in 1978 befoye beginning a decline.
2. The correlation between the change in the number of 15-19
year-olds and the change in Marlboro market share is .971
and shows Marlboro peaking in 1978 a a T0 o s are of
market and then declining.
of 15-19 year-olds and the percen c ange in Marl oro
sales is .987. A projection o this rela tionship shows
Marlboro sales peaking at 94 billion units in 1979 before
.
3. The correlation between the percent change in the number
declining.
Page 3: qfb12a00
Log in for more options!
Page 3
4. The correlation between the change in the number of 15-19
year-olds and the percent change in Marlboro sales is .988
and also shows Marl oro pe ng at 94 billion uni~in 1979.
These correlations all show Marlboro sales and market share peaking in 1978
or 1979 and then declining.
year-olds. The question, then, is why things fell apart in 1974.
interesting observation is that in all of these correlations 1974 fell far below
the trend line, i.e., below what was predicted for 1974 by the effect of 15-19
for sales on an age group that is declining in number is dangerous. Another
than the correlations of time-series data, they also show that heavy reliance
While these correlations are certainly more favorable for Marlboro's future
5.2 percent from the fourth quarter of 1973 to the fourth quarter of 1974 and
another 0.8 percent in the first quarter of 1975.
of 1973 (at an average annual rate of 3.50 percent), real per capita DPI tumbled
After a very smooth and virtually uninterrupted increase from 1960 to the end
1946 has there been such a sharp and prolonged decline in real disposable
personal income (DPI) per capita as we have witnessed in the past 18 months.
This must, of necessity, be somewhat speculative, since at no time since
Many cigarette sales forecasting models have shown some income elasticity
in the demand for cigarettes (i. e. , expenditures on cigarettes decline with a
decline in incomes). As I noted in my economic forecast, the Conference
Board calculated the price elasticity of cigarettes to be 0.29. Given the 5. 2
percent decline in real per capita DPI, an income elasticity of 0.29 would
account for a shortfall of 9 billion units in 1974. Marlboro's share of this
would have been about 1.1 billion units and its sales increase in 1974 would
have been 8.0 percent instead of 6.4 percent. This would have closed about
40% of the gap between Mariboro's actual 1974 sales increase and the sales
increase it should theoretically have achieved on the basis of my regression
equations.
This still leaves some of the gap to be accounted for, and does not explain
why Marlboro market share is below the regression line. In my economic
forecast, you may recall 1. said that Marlboro was particularly vulnerable
to the effects of the recesion because the highest unemployment rates are
among the younger age groups, precisely the groups in which Marlboro's
market penetration is highest. In the first quarter of 1975, the unemploy-
ment rate of persons 16-19 years old was 20.4 percent, the highest it has
-been since the Bureau of Labor Statistics began compiling rates by age in
1948.
Page 4: qfb12a00
Log in for more options!
Page 4
older smoker. Furthermore, many teenagers who might otherwise have ,,,
begun to smoke may have decided against it because of the adverse economic
conditions.
consumption. In addition, young smokers are less habituated than older
smokers, and can therefore probably quit or cut down more easily than an
being on the average considerably younger than the total smoking population,
tend to have lower than average incomes. Thus, I would expect a dispropor-
tionately large number of Marlboro smokers to quit smoking or reduce daily
in real income much more acutely, and since expenditures for cigarettes
constitute a larger proportion of their discretionary purchases, they are
more likely to quit smoking or reduce consumption. Marlboro smokers,
they do make are likely to take the form of buying by the carton rather than
from a machine. Lower income people, on the other hand, feel the decline
will cut back very little on expenditures on cigarettes, and what cutbacks
cigarettes are a relatively small proportion of their total expenditures, they
pinch of declining real income and may cut back on expenditures, but since
level. That is, the lower the income of a specific population group, the
greater will be the depressing effect on cigarette sales of a decline in real
income. Families whose incomes are well above the median will feel the
It is my contention that income elasticity is inversely proportional to income
These things, I think, help explain Marlboro's unusually poor showing in 1974.
It is perhaps significant that Marlboro's growth rate departed from the trend
line in the firs quar er o 1974, the very qua er in w c real per capi a DPI
s owed i s sarpes decline.
Marlboro sales are probably more responsive to price changes than are ~
the sales of brands which appeal to older segments of the population. 0
Marlboro smokers, being younger, tend to have lower incomes. Thus,
income people than on those with higher incomes. As mentioned above, ..a}
elasticity in the demand for cigarettes, although we all disagree as to the
actual figure. In any event, the wholesale price increases of 1974, which
were, of course, passed on to the consumer, undoubtedly depressed sales.
.~.
I think price elasticity, like income elasticity, has a greater effect on lower
Price Increases
_ Virtually all investigators agree that there is a certain amount of price
- . ~
i
t
F
~
hi
d E
ac
conom
c
ors
c an
Combined Effect of Demograp
.
I am convinced that the recession and the price increases are responsible
sales and Marlboro share should increase. " Because of the decline in the
more than any other brand. As we recover from the recession both industry
.
for the slower industry growth, and that these factors have hurt Marlboro
Page 5: qfb12a00
Log in for more options!
Page 5
recession will almos t certainly resume eir upward trend, but at a
continue to decline. Other things being equal Marlboro sales after the
number of young people, however, I expect Marlboro's rate of increase to
progressively slower rate. Unfortunately, other things tend not to remain
equal, such as changes in brand preference.
the decline in incidences of smoking may well be true only of college students.
It may also be true that the decline in Marlboro's popularity among 18-22
year-olds is also only a college phenomenon.
likely to smoke than people of the same age who are not in college. Thus,
as. well as data from the National Panel, show that college students are less,
and a 39% increase for females. The data from the National Clearinghouse,
house for Smoking and Health in 1968 and 1974 show no differences for males
from 1968 to 1974, but data for 17-18 year-olds from the National Clearing-
projectable to the total population aged 18 to 22. One suggestion that it is
not projectable is the reported incidence of smoking: The College Student
Smoker Survey shows a decline in the incidence of smoking for both sexes
not enrolled in college, we do not know if the results of the survey are
this survey did not also sample the 70 percent of 18-22 year-olds who are :`
among the young, as reported in the College Student Smoker Survey. Since
A much more serious factor is the apparent decline in Marlboro's popularity
Changing Brand Preferences of Young Smokers
Nonetheless, there is cause for concern, partly because college students do
constitute a significant share of the Marlboro market, and partly because
college students' behavioral and attitudinal patterns tend to carry over to
non-college youth with a one-to-three year lag, as noted in various Yankelovich
studies. Thus, the decline in the popularity of Marlboro Red among younger
smokers will probably continue and, thus, further reduce its rate of growth.
Conclusion
I think it is clear that demographic factors have been important in the recent
decline in Marlboro's rate of growth and will cause a continuing decline. The
recession and 1974 price increases have been further temporary depressants
life cycles. Much as we might like to think otherwise, Marlboro Red is
tastes, like other tastes, do change over time, and products tend to have
The company will then have to rely on other brands for growth. Cigarette
market share will peak in the very near future, probably in 1976 or 1977.
young, wb are left with the unpleasant but very real likelihood that Marlboro
If we consider further the reported decline in Marlboro's popularity among the
on Marlboro sales. Taking only economic and demograp ic ac Fors into consid- -®,
eration, a reasonable conclusion would be that as economic conditions improve 0
Marlboro will resume its growth but at a declining rate and on a permanently ~
lower trend line, until it peaks in 1978 or 1979. 0
Page 6: qfb12a00
Log in for more options!
Page 6
Mr. W. McDowell
Mr. J. Morgan
Mr. S. Fountaine
Dr. H. Wakeham
Mr. F. Resnik
Mr. R. Thomson
-:Dr. T. Osdene
Dr. W. Gannon
Mr. H. Daniel
Page 7: qfb12a00
Log in for more options!
~;:`- ~,,~., t~:.
~ T ~ '~~..ln F~ ~ ' rl:d.,,
.~.k;a~M ~ii~.~7 ~ v~.~ .w~-3 L~_ .
~y 1 s
l i:~f.3i"? li.~.:~,~;~.lw e~a....,. ...
'!`F') f'i.~tF ~yy g ~V./y ~
_ . J :..1{.~ :,~."3. V ~ . .
.} ~tY'ss3 ~ - ;'~.' a =_ ~ ~.
."i`, :.';+,.
e
.. :~:'.. . .
