role of the media in inßuencing trajectories of youth smoking · & gar y gio vino 4 centre for...
TRANSCRIPT
REVIEW
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
Blackwell Science, Ltd
Oxford, UK
ADDAddiction
1359-6357© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
98
Supplement 179103
Original Article
Media and youth smokingMelanie Wakefield
et al.
Correspondence to:
Melanie Wakefield PhDCentre for Behavioural Research in CancerThe Cancer Council Victoria1 Rathdowne StreetCarltonVictoriaAustralia 3053Tel:
+
61 3 9635 5046Fax:
+
61 3 9635 5380E-mail:
RESEARCH REPORT
Role of the media in influencing trajectories of youth smoking
Melanie Wakefield
1
, Brian Flay
2
, Mark Nichter
3
& Gary Giovino
4
Centre for Behavioural Research in Cancer, The Cancer Council Victoria, Carlton, Victoria, Australia
1
, Health Research and Policy Centers, University of Illinois at Chicago, Chicago, IL, USA,
2
Department of Anthropology, University of Arizona, Tucson, AZ, USA
3
and Department of Cancer Control, Epidemiology and
Biostatistics, Roswell Park Cancer Institute, Buffalo, New York, USA
4
ABSTRACT
This paper summarizes results of empirical studies on cigarette advertising andpromotions, antismoking advertising, product placement in movies, on televi-sion and in music media and news coverage about smoking. In addition, we pro-vide an overview of some of the theoretical literature relevant to the study ofmedia uses and effects. Finally, we discuss empirical findings in the context ofthese theories to draw some conclusions about media influences on smokingand identify issues for further research. We conclude that (a) the media bothshape and reflect social values about smoking; (b) the media provide new infor-mation about smoking directly to audiences; (c) the media act as a source ofobservational learning by providing models which teenagers may seek to emu-late; (d) exposure to media messages about smoking also provides direct rein-forcement for smoking or not smoking; (e) the media promote interpersonaldiscussion about smoking; (f) the media can influence ‘intervening’ behaviorsthat may make teenage smoking less likely; and (g) antismoking media mes-sages can also set the agenda for other change at the community, state ornational level. We outline priorities for further research which emphasize theneed for longitudinal studies, multi-level studies, an awareness of the probablydynamic relationship between tobacco advertising and antismoking advertis-ing, the importance of determining appraisal of tobacco industry youth smok-ing prevention efforts and the dearth of research on news coverage aboutsmoking.
KEYWORDS
Adolescence, anti-smoking advertising, mass media, media
advocacy, review, smoking, tobacco advertising.
INTRODUCTION
Media messages are endemic in our society through expo-sure to television and radio, movies, outdoor and point ofsale advertising, via newspapers and magazines, on theinternet and through books, brochures and posters.Given a conservative estimate of 2.5 hours of watchingTV each day over a life-time, and assuming 8 hours ofsleep per night, the average American would spend7 years of the approximately 47 waking years we have byage 70 watching TV [1]. Young people average 16–18hours of television watching per week, commencing atage 2 years [2]. Adolescents can spend nearly equal
amounts of time listening to the radio, although music isgenerally listened to while engaged in other activities [3].Over half of all 15–16-year-olds have seen the majority ofthe most popular recent R-rated movies [4] and almost allhave seen a copy of
Playboy
or
Playgirl
by the age of 15[5]. In addition, the internet continues to increase inaccessibility and popularity, exposing users to a widerange of information, previously not so accessible.
This paper is concerned with media messages aboutsmoking and their impact on youth smoking. Many ofthese messages come in the form of paid advertisingfrom tobacco companies, through promotions whichoffer accessories and clothing with cigarette brands
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
80
Melanie Wakefield
et al.
emblazoned on them, and other communications aboutsmoking which appear in the context of movies and intelevision programs and through sponsorships. Paid anti-smoking advertising is becoming more common on tele-vision and in a range of other media, and is broadcastmost often as one component of a comprehensive pack-age of strategies aimed at influencing existing smokers toquit, non-smokers not to start, and advocating for protec-tion from secondhand smoke. In addition, the public isfrequently exposed to news messages about smokingthrough television and radio news and current affairsprograms, and articles in newspapers and magazines.
This paper aims to provide a review of the literaturepertaining to media influences on tobacco, especially asthese relate to youth smoking, and to summarize some ofthe conceptual thinking about the pathways by whichmedia might influence the uptake of smoking amongyouth. The review has been undertaken as a first steptoward gaining a better understanding of a range of con-textual factors, including media, in influencing traject-ories of youth smoking.
The review summarizes first the results of empiricalstudies on cigarette advertising and promotions; anti-smoking advertising; product placement in movies, ontelevision and in music media; and news coverage aboutsmoking. The paper then gives an overview of some of thetheoretical literature relevant to the study of media usesand effects. The final section of the paper provides a furtherdiscussion of empirical findings in the context of these the-ories to draw some conclusions about media influences onsmoking and identify issues for further research.
LITERATURE REVIEW: EMPIRICAL STUDIES
Cigarette advertising
In 1997, the most recent year for which tobacco adver-tising and promotional expenditures have been pub-lished, the tobacco industry spent just under $9 billion onadvertising in newspapers, magazines, outdoor, transitand at point of sale to promote cigarettes [6]. A further$0.7 billion was spent on promotional items, such asgive-away Marlboro T-shirts and event sponsorships pro-moting cigarette brands, which are arguably also forms ofadvertising. Consumers use or wear these branded acces-sories or items of clothing, providing free walking promo-tions for cigarettes, with no health warning attached.Tobacco companies may also pay to place their productsin other media, such as movies and television programs,or there may be unpaid use of cigarettes in these enter-tainment fora. In the United States, tobacco advertisingon billboards ended in April 1999 under the terms of theMaster Settlement Agreement, but there is evidence thatcigarette advertising has increased in magazines since
this time, including magazines with high youthreadership [7]. Although Philip Morris recentlyannounced its intention to refrain from advertising insuch magazines, there is evidence that point of purchaseadvertising and promotions have increased since bill-board advertising was banned [8], so that tobacco adver-tising is still highly prevalent and visible to teenagers.
The purpose of tobacco advertising is to imbue theproduct with an image that is sufficiently attractive tomake people want to use it. There are a number of path-ways through which advertising can increase cigaretteconsumption. Four direct mechanisms include (a)encouraging youth to experiment with cigarettes and ini-tiate regular use; (b) increasing smoker’s level of dailyconsumption by acting as a cue to smoke; (c) reducingsmoker’s motivation to quit; and (d) encouraging formersmokers to resume smoking [9]. In addition, an indirectpathway which might lead to change in cigarette con-sumption is through the ubiquity and familiarity oftobacco advertising contributing to an environmentwhere tobacco use is perceived to be more socially accept-able, more normative and less hazardous than it in fact is[9].
All forms of cigarette advertising and marketing worktogether to promote cigarettes to potential and existingconsumers, so in a sense considering only that part of therange of marketing practices that are media-related issomewhat arbitrary. As explained by Kaufman & Nichter[10]:
modern marketing strives to attach symbolic meaning to specific tobacco brands by carefully manipulating the components of marketing: brand name, packag-ing, advertising, promotion, sponsorship, and place-ment in popular culture. The purpose of tobacco marketing is to associate its product with psychologi-cal and social needs that the consumer wants to fulfill, some of which emanate from the restructuring of social reality that advertising itself provides. Market-ing is more successful when these components work in a synchronized fashion, surrounding the target con-sumers with stimuli from multiple sources.
The tobacco industry has been extremely adept at cre-ating images for cigarette brands over a long period oftime, targeting advertising for different brands to women,African Americans and youth [11,12,13–16]. In addi-tion, it has used the concern about the health risks ofsmoking to its advantage by promoting ‘light’ and ‘lowtar’ cigarettes, leading consumers to believe they aresmoking a ‘safer’ cigarette, whereas in fact there is no evi-dence that these types of cigarettes deliver less tar whensmoked [17,18].
An early econometric study using disaggregated mea-sures of youth exposure to advertising provided support
Media and youth smoking
81
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
for the contention that prosmoking advertisingsignificantly increases youth smoking [19]. Studies ofsmoking initiation using population samples demon-strate convincingly that increases in teenage smoking,but not adult smoking, correspond with specific cigaretteadvertising campaigns [20–22]. Pollay and colleagues,using data from several large surveys of youth and adultsmoking over a 20-year period, found that whenever theadvertising of a brand increased, teenage smoking of thatbrand was increased three times more than adult smok-ing of the same brand [23].
The empirical evidence on the impact of tobaccoadvertising bans on smoking indicates that completeadvertising bans do influence aggregate cigarette con-sumption [24]. Many econometric studies of partialadvertising bans conclude that there is little or no effecton aggregate cigarette sales [24]. Because the advertisingbans examined in these studies were partial bans, andbecause advertising expenditures are so high, a marginalchange in expenditure will be unlikely to have any impacton aggregate sales. In addition, the tobacco industry isable to compensate for inability to advertise in onemedium, by transferring advertising dollars to othermedia outlets, so that there is, in effect, little overallchange in expenditure [9,23,25,26]. A recent analysis ofdata on 22 OECD countries from 1970 to 1992 con-cluded that
comprehensive
bans on advertising/promotionreduce smoking significantly, while limited bans have lit-tle or no effect [27].
There is substantial empirical evidence that childrenare exposed to, and remember cigarette advertising[28,29]. A series of cross-sectional studies have demon-strated that awareness, exposure and liking of cigaretteadvertising is associated with smoking status and smok-ing initiation among teenagers, and are independent ofother predictors of smoking such as peer, sibling andparental smoking [15,29–32]. Brand recall and recogni-tion is higher among teenage smokers than non-smokersin Australia [33], the United States [34–36], Britain [37–40] and Hong Kong [41].
Longitudinal studies provide more convincing evi-dence about the temporal relationship between cigaretteadvertising and smoking behavior. Alexander
et al
. [42],in following a cohort of schoolchildren aged 10–12 yearsover a 1-year period, found approval of cigarette advertis-ing to be associated among baseline non-smokers withtaking up smoking, and disapproval among baselinesmokers to be associated with quitting smoking.Armstrong
et al
. [43] found self-reported advertising-induced urges to try smoking to be associated positivelywith taking up smoking at a 1–2-year follow-up. In a 1-year follow-up survey of the Scottish study mentionedpreviously, prior awareness of cigarette advertising wasrelated to a stronger intention to smoke at follow-up after
controlling for potential confounders [44]. Using datafrom a cohort of 12–15-year-olds in Massachusetts,baseline brand-specific exposure to cigarette advertisingin magazines was found to be correlated highly withbrand of initiation among new smokers, brand smoked bycurrent smokers and brand whose advertisementsattracted the most attention [45].
Exposure to and approval of advertising also seems towork in concert with ownership of or willingness to usecigarette promotional items in influencing smoking.Teenagers have a high level of participation in ownershipof cigarette promotional items [46–48]. A series of cross-sectional studies have shown associations between own-ership of cigarette promotional items and susceptibility tosmoking [49–51], smoking behavior [15,52], age atbeginning to smoke [53] and cigarette consumptionamong youth [48]. Three longitudinal studies have dem-onstrated that youth who own promotional items aremore likely to go on to be smokers [46–48], with signifi-cant concordance between ownership of branded tobaccopromotional items among teenage non-smokers and thebrand that teenagers say they would choose if they didsmoke [48]. Redmond [54] has shown that in years ofhigh tobacco industry promotional expenditure, the rateof smoking initiation among US 9th graders was higherthan expected. These studies imply that choosing to ownor wear a cigarette promotional item may lead to a con-solidation of an identity that includes being a smoker. Thesymbolic function of owning a cigarette promotional itemmay reinforce social and group identification, acting tomaintain identification of the self as a smoker.
Taken together, these studies suggest that cigaretteadvertising and promotion probably has both predispos-ing and reinforcing effects on youth smoking, acting asan inducement to experiment with smoking, and rein-forcing continued progression towards regular smokingamong those who already have tried it. While theseeffects generally apply after holding constant the estab-lished influence of parental, sibling and friend’s smoking,further investigation of differential effects of cigaretteadvertising by baseline characteristics is rare. For exam-ple, there is little study of the influence of advertising onprogression across different stages of the uptake con-tinuum—for example, whether it is linked more stronglyto initial experimentation rather than progression fromexperimentation to regular use.
Antismoking advertising
Research concerned with the effects of antismokingadvertising on teenagers has spanned a number of disci-plines and methodologies. There are five basic types ofstudies: field experiments of persuasive mass-media inter-ventions, some of which have occurred in concert with
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
82
Melanie Wakefield
et al.
school-based prevention programs; population-basedstudies as part of the evaluation of government fundedantismoking campaigns; ecologic studies using aggre-gate data on antismoking advertising and/or youthsmoking; qualitative studies of youth appraisal of anti-smoking advertisements; and controlled laboratoryexperiments.
Field experiments
Table 1 summarizes the design and results of nine con-trolled experimental studies whereby youth have beenexposed to persuasive antismoking mass media messages,sometimes in combination with school-based smokingprevention programs, and followed over time to comparedifferences in smoking. In some cases, due to study design,it was not possible to determine whether effects were dueto the media intervention or school intervention, or acombination [55–58]. In the Midwestern PreventionProject, however [59], it was observed that effect sizes forthe combined media and school intervention were nearlydouble those of studies of the effects of school programalone [60]. Furthermore, in two trials where both strate-gies were compared, effects were significantly greaterwhere antismoking advertising occurred [61–67].
Overall, these studies suggest that antismoking adver-tising can have a beneficial effect on teenage smoking,and may be strengthened by concomitant exposure toschool-based smoking prevention programs. The patternof findings from these studies suggest that antismokingadvertising may play a greater role in preventing theuptake of smoking among teenagers rather than promot-ing cessation among teenagers who already smoke, as inseveral studies, most effects were observed for baselinenon-smokers. Consistent with this, effects seem more reli-able when exposure occurs early than later in adoles-cence, with greater effects observed in trials involvingpreteenagers and younger teenagers. However, other fac-tors aside from smoking experience may mediate theeffects of antismoking advertising.
In the Vermont study, for example, greater effects wereobserved in preventing uptake of smoking for high-risk(two or more smoking influences in their immediate envi-ronment) than low-risk girls. The authors suggest thatthe higher risk sample were faced with making decisivechoices about smoking earlier in life because of the strongsmoking models in their environment and that the mediamessages supporting non-smoking decisions may haveprovided protective effects at this critical juncture. Bycontrast, lower risk students may not have faced thesedecisive choices until later in adolescence, after the inter-vention period had ended.
In addition, because smoking among high-risk girlswas relatively more greatly reduced than among high-
risk boys, gender may be an additionally important vari-able which mediates the effect of media messages. Thismay be because girls saw more of the advertisementsthan boys, or because the message content of the adver-tisements (how to manage social relationships withoutbeing a smoker) may have been of greater utility to girlsthan boys. Alternatively, girls age into puberty earlierthan boys, so these effects could have been due largely tothe advertisements portraying information during a crit-ical period of their adolescent development, which wasonly later to be experienced by boys, at a time when theintervention was coming to completion.
Thus, gender and immediate social influences maymediate potential effects of antismoking advertising, aswell as age and previous smoking experience. Resultsfrom the controlled study in Norway also point to themediating role of the social environment, leading theauthors to conclude that emotional reactions to cam-paigns are mediated through interpersonal discussionsby the attitudes and opinions of significant others and bysocial norms regarding smoking.
Evaluation of antismoking campaigns as part of government-funded tobacco control programs
Evaluation of state and national government-fundedantismoking advertising campaigns provide a rich sourceof information about the effects of such advertising onteenagers. However, it is acknowledged that the mediacampaign is only one part of the entire antismokingeffort, which additionally encompasses community-based initiatives such as school smoking prevention pro-grams, worksite cessation programs, local policy enact-ment and enforcement and greater access to cessationservices. Furthermore, many of the state tobacco controlprograms have been funded by a percentage of the ciga-rette tax, leading to an increase in the price of cigarettesand a necessity to interpret changes accordingly.
Early studies from Australia suggested that mass-media antismoking campaigns could reduce populationsmoking among adults [68–70]. The early evidence wasgathered from an evaluation of a campaign broadcast inSydney, Australia, which encouraged smokers to quit andto set the agenda for health professionals to reinforce theimportance of not smoking. Melbourne, Australia, wasused as a non-campaign comparison community. In theintervention community, a significant decline in preva-lence of 2.8% was observed, compared with no change inthe control community [68]. When a similar campaignwas commenced later in the control community, smokingprevalence trend data from 1981 showed that there wasan immediate drop in smoking prevalence in both com-munities associated with the start of the respective cam-paigns, and a lesser decline thereafter [69].
Media and youth smoking
83
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
Tabl
e 1
Sum
mar
y of
exp
erim
enta
l tria
ls in
volv
ing
antis
mok
ing
adve
rtisi
ng a
nd t
eena
gers
.
Nam
e of
pro
ject
Nat
ure
of in
terv
entio
nTa
rget
gro
upSt
udy
desig
nM
ain
outc
omes
Nor
th K
arel
ia, F
inla
nd
(Var
tiain
en
et a
l
., 19
86;
1990
; 199
8)
C
=
No
inte
rven
tion;
11
=
pee
r-led
soc
ial i
nflue
nce
scho
ol
prog
ram
+
adul
t-fo
cuse
d m
ass m
edia
cam
paig
n
+
com
mun
ity
activ
ities
; I2
=
tea
cher
-led
soci
al in
fluen
ce s
choo
l pro
gram
+
adul
t-fo
cuse
d m
ass
med
ia c
ampa
ign
+
com
mun
ity a
ctiv
ities
7th
grad
ers
aged
12
–13
year
sQ
uasi-
expe
rimen
tal d
esig
n w
ith
mat
ched
sch
ools
in t
wo
coun
ties,
star
ting
1978
4 ye
ars
fu: s
mok
ing
prev
alen
ce s
ig. l
ower
in I
grou
ps; 8
yea
rs fu
: sm
okin
g in
itiat
ion
low
er in
I gr
oup
for
base
line
non-
smok
ers,
no d
iffer
ence
in q
uit r
ates
for b
asel
ine
smok
ers;
15 y
ears
fu: a
s ab
ove
Min
neso
ta H
ealth
Hea
lthPr
ogra
m (
Perr
y
et a
l
., 19
92)
C
=
mul
ti-co
mpo
nent
sch
ool p
rogr
am I
=
as
abov
e, p
lus
mas
s m
edia
on
hear
t he
alth
, inc
ludi
ng m
okin
g ce
ssat
ion
6th
grad
ers
Qua
si-ex
perim
enta
l des
ign
with
m
atch
ed s
choo
ls in
ano
ther
st
ate
3 ye
ars
fu: I
stu
dent
s ha
d sig
. low
er s
mok
ing
prev
alen
ce, b
ut
unab
le t
o se
para
te e
ffect
s of
sch
ool p
rogr
am a
nd m
edia
Stan
ford
Five
City
Pro
ject
(F
ortm
ann
et a
l
., 19
94;
Win
kelb
y
et a
l
., 19
93)
C
=
No
inte
rven
tion
I
=
med
ia a
dvoc
acy
and
som
e an
tism
okin
g ad
vert
ising
aim
ed m
ainl
y at
adu
lts d
urin
g 19
80–8
512
–14-
year
-old
sQ
uasi-
expe
rimen
tal d
esig
n w
ith
mat
ched
com
mun
ities
, cro
ss-
sect
iona
l sur
veys
From
197
9 to
199
0, p
reva
lenc
e of
dai
ly s
mok
ing
for
12–2
4-ye
ar-
olds
did
not
diff
er in
I an
d C
com
mun
ities
Verm
ont
Stud
y (F
lynn
et a
l
., 19
92; 1
994;
1997
; Sec
ker-
Wal
ker
et a
l
., 19
97;
Wor
den
et a
l
., 199
6; 19
98;
Wor
den
& F
lynn
, 200
0)
C
=
sch
ool p
rogr
am o
nly
I
=
sch
ool p
rogr
am
+
TV
and
radi
o an
tism
okin
g ad
vert
isem
ents
bro
adca
st fo
r 5 m
onth
s in
year
1
and
for
1 m
onth
in y
ears
2-4
, mor
e m
essa
ges
targ
eted
at
high
-risk
girl
s
5th
-
7th
grad
ers
Qua
si-ex
perim
enta
l des
ign
with
tw
o m
atch
ed p
airs
of s
choo
l co
mm
uniti
es
4 an
d 6
year
s fu
: I s
tude
nts
had
sig. l
ower
sm
okin
g st
atus
, pa
st
wee
k sm
okin
g an
d da
ily s
mok
ing;
Sig.
rela
tive
redu
ctio
n in
ou
tcom
es fo
r hi
gh-r
isk t
eens
, esp
ecia
lly g
irls
Sout
heas
tern
US
(Bau
man
et
al
., 19
88; 1
991)
I1
=
rad
io a
ntism
okin
g ad
vert
isem
ents
I2
=
as
abov
e, p
lus
adve
rtise
men
t to
enc
oura
ge p
artic
ipat
ion
in n
on-s
mok
ing
swee
psta
kes
I3
=
as
abov
e, p
lus T
V ad
vert
isem
ent
to
enco
urag
e pa
rtic
ipat
ion
in s
wee
psta
kes
12–1
4-ye
ar-o
lds
Hou
seho
ld s
urve
ys o
f tee
ns in
th
ree
non-
over
lapp
ing
med
ia
mar
kets
in S
E U
SA, fo
llow
ed-u
p at
2 y
ears
No
diffe
renc
es in
sm
okin
g be
havi
or a
t fo
llow
-up
Sout
hern
Cal
iforn
ia (
Flay
et
al
., 19
88; 1
995)
C1
=
no
inte
rven
tion
C2
=
hea
lth in
form
atio
n on
ly c
ontr
ol I
=
so
cial
influ
ence
s sc
hool
pro
gram
+
tel
evisi
on in
terv
entio
n7t
h gr
ader
sQ
uasi-
expe
rimen
tal d
esig
n w
ith
stud
ent
follo
w-u
p tw
ice
durin
g 7t
h gr
ade
and
once
dur
ing
each
of
8th
, 9th
, and
10t
h gr
ade
Stro
ng im
med
iate
pos
itive
effe
cts
on m
edia
tor
varia
bles
suc
h as
kn
owle
dge
of s
mok
ing
cons
eque
nces
, per
ceive
d sm
okin
g pr
eval
ence
and
effo
rts
to re
sist t
ryin
g ci
gare
ttes,
part
ial d
ecay
at
2 y
ears
fu, b
ut s
till s
ig. d
iffer
ent.
No
sig. e
ffect
s on
sm
okin
g be
havi
orM
idw
este
rn P
reve
ntio
n Pr
ojec
t (Jo
hnso
n
et a
l
., 19
90; M
acKi
nnon
et a
l
., 19
91; P
entz
et a
l
., 19
89a;
1989
b)
I
=
mul
ti-co
mpo
nent
sch
ool a
nd c
omm
unity
pro
gram
to
resis
t illi
cit d
rug
use,
incl
udin
g re
duci
ng s
mok
ing,
med
ia c
ompo
nent
in
volve
d ne
ws
cove
rage
Mid
dle
and
juni
or
high
sch
ool
stud
ents
Sing
le g
roup
des
ign
usin
g
long
itudi
nal fo
llow
-up
at 1
, 2 an
d
3 ye
ars
1, 2
and
3 ye
ars
fu: r
educ
ed r
ates
of p
ast
wee
k an
d pa
st m
onth
sm
okin
g. N
o co
ntro
l gro
up, s
o di
fficu
lt to
attr
ibut
e ca
usat
ion
to m
edia
, alth
ough
effe
ct si
zes a
lmos
t dou
ble
effe
cts o
f sch
ool
prog
ram
s al
one
Proj
ect
Sixt
een
(Big
lan
et a
l
., 20
00)
C
=
sch
ool i
nter
vent
ion
I
=
sch
ool
+
com
mun
ity in
terv
entio
n, in
clud
ing
paid
ant
ismok
ing
adve
rtise
men
ts o
n ra
dio,
ne
wsp
aper
art
icle
s an
d po
ster
s
7th
and
9th
grad
ers
Mat
ched
pai
rs o
f sch
ools
assig
ned
to t
he t
wo
cond
ition
s an
d fo
llow
ed-
up fi
ve t
imes
unt
il 1
year
afte
r in
terv
entio
n
At c
ompl
etio
n of
I an
d at
1 y
ear f
u: I s
tude
nts h
ad si
g. lo
wer
rat
es
of p
ast
wee
k sm
okin
g
Nor
way
(H
afst
ad
et a
l
., 199
6; 19
97; H
afst
ad &
Aar
o,
1997
)
C
=
no
inte
rven
tion
in c
ount
y I
=
cou
nty
has
antis
mok
ing
TV
adve
rtise
men
t, ci
nem
a ad
vert
isem
ent,
new
spap
er a
dver
tise-
men
ts a
nd p
oste
rs m
aile
d to
scho
ols.
Cam
paig
n ai
med
mai
nly
at g
irls,
emph
asizi
ng m
ismat
ch b
etw
een
bein
g a
smok
er a
nd
inde
pend
ence
, app
eara
nce
and
conc
ern
for
envi
ronm
ent
14–1
5-ye
ar-o
lds
Two
mat
ched
cou
ntie
s, us
ing
coho
rt o
f tee
nage
rs s
ampl
ed
from
hou
seho
ld s
urve
y, fo
l-lo
wed
-up
at 6
–12
mon
ths,
3 ye
ars
3 ye
ars
fu: s
ig. r
educ
tion
in o
dds
of b
eing
a s
mok
er fo
r I g
roup
co
mp
with
C g
roup
for
boys
and
girl
s. Re
duct
ion
in o
dds
of
smok
ing
for
base
line
smok
ers
appl
ied
equa
lly t
o m
ales
and
fe
mal
es. R
educ
tion
in o
dds
of s
mok
ing
for b
asel
ine
non-
smok
-er
s ap
plie
d on
ly t
o gi
rls
C
=
cont
rol;
I
=
inte
rven
tion.
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
84
Melanie Wakefield
et al.
In the United States, the first state-wide antismok-ing mass media campaign was conducted in Minnesotafrom 1986, as a result of government funding ofapproximately $2 million per year. Advertising designedto increase youth awareness of the negative social con-sequences of smoking and to correct normative expecta-tions for smoking among adolescents were broadcast ontelevision and radio and displayed in newspapers andon billboards. From 1986 to 1990, reported exposure toantismoking advertising was significantly higher among9th graders in Minnesota than the control state of Wis-consin, but there were no changes in smoking-relatedbeliefs [71] or behavior [72]. The investigators suggestthat the lack of effect may have been due to the lack ofongoing and substantive school-based smoking preven-tion programs, and speculated that both media andschool programs may be required to influence youthsmoking.
Table 2 summarizes information about ongoing state-wide comprehensive tobacco control programs in theUnited States that had media campaigns as an importantelement up to 1999. The California Tobacco Control Pro-gram (CTCP), funded by Proposition 99 from 1989, wasthe first ongoing comprehensive state-wide tobacco con-trol program in the United States.
An early evaluation of the media campaign involvinga series of cross-sectional surveys of school students ingrades 4–12 in California demonstrated a significantincrease in recall of the media campaign, attitudes moreunfavorable towards smoking, a decrease in intention tosmoke and a decline in 30-day smoking prevalence [73].However, there was no unexposed control group for com-parison purposes. As indicated in Table 2, the CTCP hasbeen associated with greater declines in per capita ciga-rette consumption and adult smoking prevalence thanthe rest of the United States, especially during the earlyprogram period when funding was highest.
From cross-sectional surveys conducted within Cali-fornia, standardized 30-day smoking prevalence did notchange among 12–17-year-olds from 1990 to 1993(9.2%), but from 1993 to 1996 increased significantlyfrom 9.2% to 12.0%, coincident with the reducedamount of tobacco control funding and the increasedratio of tobacco industry to tobacco control funding [74].In addition, there was an increase in the percentage of12–14-year-olds who were susceptible to becomingsmokers (from 34.5% in 1993 to 42.0% in 1996). Com-parison of data from the school-based Monitoring theFuture Surveys shows that although smoking increasedin California between 1993 and 1996 in both 8th (rela-tive increase of 16%) and 10th graders (relative increaseof 6%), this was less than was observed for 8th (increaseof 29%) and 10th graders (increase of 23%) in the rest ofthe United States [75].
The Massachusetts Tobacco Control Program (MTCP)was established in October 1993 and has the highest levelof per capita funding of all US states. The mass mediacomponent of the campaign has emphasized the healtheffects of smoking and passive smoking, as well as preven-tion messages aimed at youth. Per capita cigarette con-sumption declined with the start of the program to anextent greater than that expected from a short-term priceincrease, suggesting that the media campaign was animportant component in this change [76]. Furthermore,adult smoking prevalence has declined significantly sincethe start of the program, compared with other US states[76,77].
From 1993 to 1996, the relative change in smokingprevalence in Massachusetts teens compared with theircounterparts in the rest of the US was in the oppositedirection for 8th graders (1.9%
decrease
in Massachusettscompared with a 25.7% increase for rest of the UnitedStates), and has been minimized for 10th graders (16.3%increase for Massachusetts compared with 23.1%increase for the rest of the US) and 12th graders (7.4%increase for Massachusetts compared with 13.7%increase for US) [78]. More recent data confirm thistrend. From 1995 through 1997, smoking prevalenceamong all Massachusetts students changed from 35.7 to34.4% (a relative
decrease
of 3.6%), compared to achange from 34.8 to 36.4% nationally (a relative
increase
of 4.5%) (CDC, 1999). Surveys also show a decline in life-time cigarette use among 7th to 9th graders in Massa-chusetts [79,80]. Of most interest, comparisons withnational data from the MTF surveys show a relativedecline of 4.6% in life-time cigarette use for 8th gradersin Massachusetts (from 52.2 to 49.8%), against anational relative
increase
of 9.5% (from 45.3 to 49.3%)[78].
A recent study used data from a 4-year longitudinalpopulation sample of Massachusetts youth aged 12–15 in1993 to specifically examine the impact of antismokingadvertising on smoking behavior [81]. In the 4-yearperiod of the campaign, Massachusetts spent more than$12.5 million/year or $2 per capita per year on its mediacampaign. Among younger adolescents (aged 12–13 atbaseline), recall of antismoking advertising on televisionin the past 30 days was associated significantly with alower rate of progression to established smoking (self-report of smoking 100 cigarettes) at a 4-year follow-up,after controlling for age, gender, race, baseline smokingstatus, smoking by parents, friends and siblings, televi-sion viewing frequency and exposure to-non-media cam-paign related antismoking messages. However, exposureto television antismoking advertisements had no effect onprogression to established smoking among older adoles-cents (aged 14–15 at baseline), and there were no effectsof exposure to radio or billboard advertising.
Media and youth smoking
85
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
Baseline exposure to antismoking advertising was notassociated with subsequent differences in seven of eightsmoking-related knowledge and attitude variables. How-ever, youths exposed to antismoking advertising were 2.3times more likely to report at follow-up that less than halfthe youths at their high school were smokers. In thiscohort study, the relationship between exposure to anti-smoking advertising and this variable, denoting an accu-rate as opposed to inflated perception of youth smokingprevalence, was stronger for those aged 12–13 at base-line than for older adolescents [81]. This pattern of resultssuggests that the protective effect of antismoking adver-tising may be mediated in part by reducing perceivedyouth smoking prevalence (at least two of the advertise-ments aimed to show that smoking was not the norm),which itself is known to have a strong influence on youthsmoking initiation [11,82,83]. These results are consis-tent with indications from earlier research that antismok-ing advertising may have more demonstrable impact onyounger than older teens.
Unlike the other state-wide comprehensive programs,the Florida Tobacco Pilot Program (FTPP) funding wasprovided through the provisions of the settlementbetween the state and tobacco companies [84]. The so-called ‘Truth’ campaign began in April 1998, aimed atteenagers aged 12–17 years, and placed particularemphasis on engendering unfavorable attitudes towardsthe tobacco industry. The program also fostered commu-nity partnerships with all 67 Florida counties, school-based initiatives, an education and training initiative,enhanced enforcement of youth tobacco access laws anda law that penalized youth for possession of tobacco.
Media tracking surveys of teenagers demonstratedhigh rates of campaign awareness, and specific advertise-ment awareness in the first 6 weeks of the campaign,which persisted to 1 year [85]. In addition, over the firstyear of the campaign, there was a change in attitudesconsistent with the intention of campaign messages, anddecreases in adolescent intentions to smoke and smokingbehavior among Florida youth, compared with youth inother states with low levels of antismoking activity [85].In addition, in a follow-up study of Florida youth aged12–17 years, those scoring at intermediate and high lev-els on an index of ‘Truth’ advertising impact, were lesslikely to initiate smoking than youth who could not con-firm awareness of the television advertisements [86].
Furthermore, in surveys undertaken by the FloridaDepartment of Health, the prevalence of 30-day cigaretteuse among middle school students declined significantlybetween 1998 and 1999 from 18.5% to 15.0% (declineof 18.9%) and among high school students from 27.4%to 25.2% (decline of 8.0%) [87]. The trends observed inFlorida are substantially larger than the modest relativedeclines observed (of 6.9% and 1.4% for middle and high
school students, respectively) between 1998 and 1999from the Monitoring the Future surveys [88]. A recentstudy indicated that these declines among Florida youthcontinued throughout 2000, and additionally that thepercentage of committed non-smokers increased signifi-cantly [89].
Two other states have been running state-wide tobaccocontrol programs for several years—Arizona has beenin the field since 1996 and Oregon since 1997. Oregonhas elected to run antismoking ads already developed inother states that pretest well, thereby allocating thebulk of media funding to broadcasting, rather thanadvertisement development. However, Arizona specifi-cally developed and directed its media advertising to pre-teens and teens, as well as pregnant women, and only in1999 began including adults as a target group for media[84]. For both states, recall of media elements of the pro-gram by teenagers has been high [90,91]. A rating eval-uation of the Arizona advertisements by 1831 6th to 12thgraders in Arizona indicated that many of the advertise-ments were perceived as being relevant to them andreduced their intention to smoke, especially among exist-ing non-smokers [92]. Trends in tobacco use by adoles-cents in Oregon mirrored national trends for the first 2years following commencement of the program [90] andmost recent state figures indicated a substantial and sig-nificant decline in 8th grade smoking between 1998 and1999 [93], which was greater than national trends,although different survey methods were used, so these fig-ures are still inconclusive. The extent of decline in per cap-ita consumption following the introduction of the OregonTobacco Education and Prevention Program [94] is highlyconsistent with what has been observed in California andMassachusetts. However, in Arizona, follow-up data arenot yet available to track change in teenage smoking.
With few exceptions, this body of research has beenconcerned with determining whether a state-wide com-prehensive tobacco control program is associated withreduced cigarette consumption and lower smoking prev-alence among adults and teenagers. To a great extent, theoverriding aim of determining the impact of the overallprogram on population smoking has been pursued at theexpense of more fine-grained research, which might havebeen focused on the advertising itself and upon individualdifferences in responsiveness to advertising messages.The study by Siegel & Biener [81] is one of the few to havegone beyond these overarching aims, in taking a closerlook at individual differences.
Overall, the conclusion from the studies of govern-ment funded antismoking campaigns is that they doreduce adult and youth smoking [84,95]. Aggregate percapita consumption declines in response to the onset ofcampaign activities and does so in a magnitude greaterthan that expected on the basis of a price increase alone.
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
86
Melanie Wakefield
et al.
Tabl
e 2
Sum
mar
y of
sta
te t
obac
co c
ontr
ol c
ampa
ign
appr
oach
es a
nd r
epor
ted
effe
cts
on y
outh
sm
okin
g.
Cal
iforn
ia: 1
989–
Mas
sach
uset
ts: 1
993–
Ariz
ona
Ore
gon
Flor
ida
Per
capi
ta p
rogr
am
spen
ding
(in
US
1999
dol
lars
)
1989
–93:
$3.2
7 19
93–9
6: $1
.78
1994
–97:
aver
age
$7.0
9 (A
bt
Ass
ocia
tes,
1998
) 19
96–9
8: av
erag
e $3
.89
(Bia
lous
&
Gla
ntz,
1999
)19
97–9
9: $2
.59
(Ore
gon
Hea
lth
Div
ision
, 199
9)19
97: $
2.61
199
8: $4
.73
(Give
l &
Gla
ntz,
1999
)
Prog
ram
com
pone
nts
1989
–96
(ave
rage
): M
ass
med
ia
17%
Loc
al le
ad a
genc
y gr
ants
26
% C
ompe
titiv
e gr
ants
22%
Sc
hool
-bas
ed p
rogr
ams 3
1%
Adm
inist
ratio
n an
d ev
alua
tion
5% (
Pier
ce
et a
l
., 19
98)
1996
–97:
Mas
s m
edia
33%
Loc
al
lead
age
ncy
gran
ts fo
r ce
ssat
ion,
educ
atio
n, ad
voca
cy
43%
Sch
ool p
rogr
ams
15%
St
ate-
wid
e se
rvic
es, e
.g.
trai
ning
and
qui
tline
5%
Re
sear
ch/e
valu
atio
n 4%
(A
bt
Ass
ocia
tes,
1998
)
1997
–98:
Mas
s m
edia
and
spo
nsor
ship
s 54
% L
ocal
lead
age
ncy
gran
ts fo
r sc
hool
edu
catio
n, ce
ssat
ion,
prot
ectio
n fro
m E
TS*
25%
Info
cl
earin
ghou
se a
nd q
uitli
ne 5
% S
tate
-w
ide
proj
ects
, adm
in. a
nd e
valu
atio
n 16
% (
Bial
ous
& G
lant
z, 19
99)
1997
–99:
Publ
ic a
war
enes
s an
d ed
ucat
ion
27%
Loc
al le
ad a
genc
y gr
ants
38%
Sta
te-w
ide/
regi
onal
pr
ojec
ts S
choo
l pro
gram
s 12
% C
o-or
d/ev
alua
tion
7% (
Ore
gon
Hea
lth
Div
ision
, 199
9)
1998
–99:
Mas
s m
edia
37%
Ed
ucat
ion/
trai
ning
23%
You
th
and
com
mun
ity p
rogr
ams
for
quitl
ine,
trib
al p
rogr
ams
16%
Ev
alua
tion
6% (
Giv
el &
Gla
ntz,
1999
)
Prog
ram
focu
sA
dults
Tee
nage
rs P
rote
ctio
n of
no
n-sm
oker
s fro
m E
TS*
Adu
lts T
eena
gers
Pro
tect
ion
of
non-
smok
ers
from
ET
S*Pr
e-te
enag
ers T
eena
gers
Pre
gnan
t w
omen
Adu
lts fr
om m
id-1
998
Adu
lts T
eena
gers
Pro
tect
ion
of n
on-
smok
ers
from
ETS
*Te
enag
ers
Eval
uatio
n el
emen
tsO
ngoi
ng c
ross
-sec
tiona
l po
pula
tion
surv
eys
of a
dults
an
d te
ens;
Coh
ort
stud
y of
te
ens
Ong
oing
pop
ulat
ion
surv
eys
of
adul
ts a
nd t
eens
; Coh
ort
stud
ies
of t
eens
and
adu
lts;
Trac
king
of p
er c
apita
co
nsum
ptio
n
Surv
eys
of r
ecal
l and
app
raisa
l of
cam
paig
ns; T
rack
ing
of p
er c
apita
co
nsum
ptio
n Po
pula
tion
surv
eys
of
teen
s an
d a
dults
Stan
dard
ized
repo
rts
on p
lace
men
t of
m
ass
med
ia, q
uitli
ne c
alls;
Tra
ckin
g of
pe
r ca
pita
con
sum
ptio
n; Su
rvey
s of
ad
ult
and
teen
sm
okin
g
Teen
and
adu
lt su
rvey
s to
ass
ess
reca
ll of
cam
paig
n an
d be
liefs
and
attit
udes
; sch
ool s
urve
ys t
o as
sess
sm
okin
g be
havi
or
Mas
s med
ia ca
mpa
ign
reca
ll an
d re
cogn
ition
Hig
h le
vels
of c
ampa
ign
awar
enes
s am
ong
adul
ts a
nd
teen
ager
s (P
ierc
e
et a
l
., 19
98c;
Poph
am
et a
l
., 19
94;
How
ard-
Pitn
ey
et a
l
., 19
98)
Incr
easin
g m
ajor
ity o
f ado
lesc
ents
ha
ve se
en a
nd h
eard
cam
paig
n ad
vert
ising
and
rec
ogni
ze
cam
paig
n th
eme
(Brit
on
et a
l
., 19
97)
1998
: 2/3
tee
ns, p
regn
ant
wom
en a
nd
adul
ts r
epor
ted
seei
ng a
dver
tisin
g in
la
st 3
0 da
ys
(Eise
nber
g
et a
l
., 19
98)
74%
of a
dults
and
84%
of t
eens
reca
ll at
le
ast
one
cam
paig
n ad
vert
isem
ent
(Ore
gon
Hea
lth D
ivisi
on, 1
999)
Sept
199
8: 28
% t
eens
rep
orte
d se
eing
one
adv
ertis
emen
t ea
ch
day
and
66%
, at
leas
t on
e ea
ch
wee
k (C
ente
r fo
r St
udy
of
Popu
latio
n, 19
98).
Jan.
1999
: 48
%
*ETS
=
Env
ironm
enta
l Tob
acco
Sm
oke
Media and youth smoking
87
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
Per
capi
ta
cons
umpt
ion
Sign
ifica
nt d
eclin
e co
mpa
red
with
bas
elin
e co
nsum
ptio
n an
d by
com
paris
on w
ith re
st
of U
SA a
nd g
reat
er t
han
expe
cted
from
pric
e in
crea
se
alon
e (P
ierc
e
et a
l
., 19
98c;
Gla
ntz
1993
; Hu
et a
l
., 199
5a;
1995
b)
Sign
ifica
nt d
eclin
e du
ring
1993
–96
com
pare
d w
ith b
asel
ine
perio
d 19
90–9
2 an
d fo
r res
t of
USA
, gr
eate
r th
an e
xpec
ted
for
pric
e in
crea
se a
lone
(H
arris
et a
l
., 19
96; B
iene
r
et
al
., 20
00)
Dec
line
of 5
.4%
in 1
995
afte
r adj
ustm
ent
for
stoc
kpilin
g of
low
er-p
riced
ci
gare
ttes:
due
to p
rice
incr
ease
onl
y, as
pro
gram
did
not
sta
rt u
ntil
1996
(H
ogan
et a
l
., 19
96)
Sign
ifica
nt d
eclin
e co
mpa
red
with
ba
selin
e co
nsum
ptio
n an
d by
co
mpa
rison
with
rest
of U
SA (P
izica
ni
et a
l
., 19
99)
Unk
now
n
Adu
lt pr
eval
ence
Rate
of d
eclin
e ex
ceed
ed th
at o
f re
st o
f USA
from
198
9 to
19
93, b
ut w
as le
ss t
han
for
rest
of U
SA in
199
3–96
(P
ierc
e
et a
l
., 19
98b)
Rela
tive
decl
ine
of 9
% fr
om
3 ye
ars
befo
re p
rogr
am to
firs
t 3
year
s of p
rogr
am, w
hich
was
gr
eate
r th
an 3
% d
eclin
e fo
r re
st
of U
SA (
Har
ris
et a
l
., 19
96;
Bien
er
et a
l
., 20
00)
Cha
nge
data
not
yet
rep
orte
dRe
lativ
e de
clin
e of
6.4
–21.
9% in
199
8,
but n
o na
tiona
l com
paris
on (O
rego
n H
ealth
Div
ision
, 199
9)
Cha
nge
data
not
yet
rep
orte
d
Teen
sm
okin
gW
ithin
-sta
te s
urve
ys s
how
no
chan
ge
in 1
2–17
-yea
r-old
pr
eval
ence
from
19
90 t
o 19
93 a
nd in
crea
se fr
om
1993
to
199
6, an
d in
crea
se
in n
on-s
mok
er s
usce
ptib
ility
(Pie
rce
et a
l
., 199
8b) A
mon
g 8t
h an
d 10
th
grad
ers,
rela
-tiv
e in
crea
se in
sm
okin
g
prev
alen
ce fr
om 1
993
to
1996
was
les
s tha
n ot
her U
S st
ates
(U
nger
et a
l
., 19
98)
Rela
tive
incr
ease
in 3
0-da
y p
rev-
alen
ce le
ss th
an fo
r res
t of U
S
for 8
th a
nd 1
0th
grad
ers f
rom
19
93 t
o 19
96 (
Brito
n
et a
l
., 19
97);
Rela
tive
incr
ease
for
9th
to 1
2th
grad
ers
less
tha
n fo
r re
st o
f US
from
199
3 to
19
97 (
Div
ision
of A
dole
scen
t Sc
hool
Hea
lth, 1
999)
; Rel
ative
de
clin
e in
life
-tim
e us
e fo
r 8t
h gr
ader
s com
pare
d to
incr
ease
fo
r re
st o
f USA
(Br
iton
et a
l
., 19
97)
Cha
nge
data
not
yet
rep
orte
dA
mon
g 8t
h an
d 11
th g
rade
rs, s
ame
as
natio
nal t
rend
s fo
r fir
st 2
yea
rs o
f ca
mpa
ign
(Ore
gon
Hea
lth D
ivisi
on,
1999
)
From
Feb
ruar
y 19
98 t
o 19
99, r
ela-
tive
decl
ines
in 3
0-da
y p
reva
-le
nce
for m
iddl
e an
d hi
gh s
choo
l st
uden
ts (
Baue
r
et a
l
., 19
99;
2000
) wer
e gr
eate
r tha
n na
tiona
l tr
ends
(U
nive
rsity
of M
ichi
gan
New
s, 19
99)
Cal
iforn
ia: 1
989–
Mas
sach
uset
ts: 1
993–
Ariz
ona
Ore
gon
Flor
ida
*ETS
=
Env
ironm
enta
l Tob
acco
Sm
oke
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
88
Melanie Wakefield
et al.
However, because adolescent smoking comprises only asmall percentage of overall aggregate consumption, theseresults are generally interpreted as being more indicativeof adult change rather than, although not excluding,change in teenage smoking behavior. However, cross-sectional and longitudinal population survey researchadds additional evidence that these types of campaignscan have enduring effects on reducing teenage smoking.Importantly, a salient finding from this literature is thatcampaigns seem to have more influence on younger, thanolder, teens.
Comparative effectiveness of antismoking themes
Plainly, not all antismoking advertising is equal. A grow-ing body of research has begun to explore the types ofantismoking message themes that might best resonatewith youth [96–104]. This research has produced mixedfindings, due probably to the different methodologiesinvolved, but some consistencies emerge.
Advertisements which elicit strong emotional arousal,typically those that portray graphically the adverse con-sequences of smoking, often rate highly among teens andadults, and are associated with increased intention not tosmoke [101–104]. Advertisements highlighting thedeceptive and misleading conduct of the tobacco industrytypically require a more ‘sophisticated’ target audiencewith additional experience in understanding these mes-sages [99,101]. Advertisements with a theme emphasiz-ing that teens need to make a choice about whether or notto smoke generally have lowest ratings among youth[101]. Thus, the Philip Morris ‘youth smoking preventioncampaign’, which exclusively uses the latter types ofadvertisements, are probably ineffective in motivatingyouth to ‘stop and think’ about smoking.
Another reason for inconsistent research findings isthat advertising effectiveness is also influenced by the wayin which the message is executed—for example, factorssuch as casting, lighting, sound, voice-overs, the numberof frames, the setting and the wardrobe may all influencethe ‘take-home’ message of the advertisement, its memo-rability, personal relevance and persuasiveness. Further-more, in the real world setting, advertisements are viewedin the context of television or radio programs, and this caninfluence how they are perceived [105]. Also, advertisingis received and interpreted in a context and this will differaccording to the novelty of the message. Newly com-menced antismoking advertising campaigns may haverelatively high salience in some states investing for thefirst time in antismoking advertisements, but for statesthat have had antismoking advertising for years, rotationand careful scheduling of messages may be necessary tomaintain message salience and avoid advertising ‘wear-out’. In addition, some messages may be more complex
than others and may require a longer duration or higherfrequency of exposure to engage the target audience asintended. Thus, research on the effectiveness of differentmessages themes in attempting to partition out one aspectof advertising as if all other factors that may influenceimpact are constant, which is clearly not the case. Furtherresearch in this field is needed.
Comparative analyses of antismoking advertising and cigarette advertising
From 1967 to 1970, prior to the ban on broadcasttobacco advertising in the United States, television net-works were required by the Federal Communication Com-mission’s Fairness Doctrine to broadcast one antismokingadvertisement for approximately every three tobaccoadvertisements, and an exposure of one antismokingadvertisement to every four tobacco advertisements wasachieved. Over this period, nearly $200 million in com-mercial advertising time (in 1970 dollars) was donatedfor this purpose [106] equivalent to approximately $292million per annum in 2000 dollars. Several empiricalstudies have found that, over the period these antismok-ing advertisements were broadcast, per capita cigaretteconsumption declined by over 10%, a trend that hadoccurred only once before in the century, when thehealth hazards of smoking were first publicized by theSurgeon-General in 1964 [25,106–108]. Furthermore,analysis showed that these changes involved teenagers[18]. When cigarette advertising was removed in 1971and the antismoking advertisements were removed withthem, consumption resumed an upward trend. In the late1990s, the advertising and marketing expenditure of thetobacco industry has increased markedly in real terms, sothat an antismoking expenditure equivalent to $1 billionper year would probably be required today to replicate thecomparative exposure ratio of antismoking advertising tocigarette advertising achieved during the late 1960s.
Apart from this ‘natural experiment’, relatively fewstudies have sought to evaluate the comparative effects ofdifferent levels of exposure to antismoking advertisingand cigarette advertising.
Pechmann & Ratneshwar [109] exposed 7th gradersin California to antismoking advertising, cigarette adver-tising or control advertising embedded in a magazine foradolescents. Later, they participated in an ostensiblyunrelated study in which they read trait informationabout a peer who was either identified or not identified asa smoker. Youth’s evaluative judgements of the peer andtheir thoughts and inferences about the person were thenassessed. Antismoking advertisements were found tolower perceptions of a teenage smoker’s common sense,personal appeal, maturity and glamor. In contrast, ex-posure to cigarette advertisements resulted in more
Media and youth smoking
89
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
favorable thoughts about the teenage smoker. Pechmann[110] found that just one antismoking advertisementwas able to offset the impact of three cigarette advertise-ments, which would otherwise have enhanced percep-tions of a teenage smoker’s social stature, poise,popularity and vitality.
It seems conceptually plausible that antismokingadvertising and cigarette advertising might compete withone another for the attention of potential and existingsmokers. Saffer & Chaloupka [27] and Levy & Friend[111] discuss advertising response functions in attempt-ing to portray the critical points where advertising is mosteffective, where additional exposures are subject todiminishing marginal returns, and how these thresholdsmight be modified by other influences. Advertisingresponse functions suggest that advertising has very littleeffect until it reaches a certain critical level of exposure,or threshold, after which there are large pay-offs forincreasing investment in the campaign [112,113]. This isconsistent with the point made by Pechmann [114], thatantismoking ads must have ‘adequate “share of voice” tobreak through ad clutter, attract attention and persuade’.
Saffer & Chaloupka [27] and Levy & Friend [111] positthat the ratio of cigarette advertising to antismokingadvertising changes the relative effectiveness of each.Thus, cigarette advertising may increase the threshold ofantismoking messages required to effectively influencesmoking behavior. In addition, there is evidence thattobacco companies increase their marketing efforts, atleast at the point of sale, in states where comprehensivetobacco control programs exist [115], in an attempt tooffset the effects of the antismoking campaign. Thus, thetwo may be in a dynamic relationship, and are not inde-pendent from each other.
Portrayal of smoking in movies, television programs, magazines and music media
Teenagers are three times more likely than adults to befrequent movie-goers [116]. It is estimated that the aver-age teenager goes to the movies approximately once eachmonth and watches two feature movies per week on tele-vision or videotape [117]. In addition, adolescents areavid consumers of magazines, which carry both cigaretteadvertising and fashion images depicting smoking.
Portrayals of smoking in popular films occur withmuch more frequency than expected on the basis ofsmoking prevalence [118,119]. The rate at whichtobacco appears per minute of film reached a minimumduring the 1980s and increased during the 1990s torates comparable to those observed in the 1960s [119].Of a random sample of the top 20 highest grossing filmseach year from 1990 to 1996, an average of 5 minutesper film involved a tobacco-related incident (including
implied or actual consumption of tobacco, tobacco para-phernalia, talking about tobacco and discussion aboutthe dangers of tobacco use), of which only 43 secondsinvolved an antismoking message [119]. Tobacco useoccurred in two-thirds of children’s G-rated animatedfilms from 1937 through 1997, and in all such filmsreleased in 1996–97 [120]. Several studies have foundthat smokers are depicted more positively than non-smokers and that Hollywood’s portrayal of smoking tendsto ignore the negative consequences and correlates ofsmoking [119,121]. A recent study found that moviesaimed at young audiences were less likely to carry nega-tive messages associated with tobacco use than weremovies aimed at mature audiences [122].
Although portrayal of tobacco in fictional televisionprograms is less frequent, there is evidence that it wasincreasing during the early 1990s [118] after a period ofdecline during the 1980s [123,124]. Tobacco portrayalalso occurs in music lyrics and music videos. One study ofmusic videos broadcast on network television foundtobacco use to be present in 12% of country music videosand 26% of those broadcast on MTV [125]. Furthermore,style and fashion magazines with high youth readershipshow models and personalities smoking in their editorialpages, especially fashion spreads, and there is evidencethat this practice has increased [126].
Tobacco portrayal in films has often been a form of‘product placement’ used commonly by commercial mar-keters to promote brands [127,128]. This can occur intelevision programs, radio shows, music videos, maga-zines, video games, plays, songs and even novels. AsHadju [129] estimated, movies in the 1980s contained,on average, 30–40 minutes of screen time involving prod-uct promotions, translating into approximately one-thirdof a typical movie. Product placement can assist in offset-ting the costs of film-making and distribution and can bea highly efficient way for advertisers to associate a prod-uct with desired qualities and situations [127]. An addi-tional motivation for tobacco companies to use films hasoccurred as other avenues for positioning their productsbecome unavailable through tighter government controlson tobacco advertising and promotion [130,131].Although the Tobacco Institute claims that payments forspecific brand placements ended in 1990 [132], the fre-quency of smoking in movies has dramatically increasedsince then [119]. In late 1997, California Senator JohnBurton held hearings on the use of tobacco in films [132]and Vice President Al Gore convened a meeting of enter-tainment leaders to discuss smoking in the movies[132,133]. Soon after, the Entertainment IndustriesCouncil (EIC) launched an initiative aimed at reducingthe gratuitous use of tobacco in films [134]. Interviewsundertaken in the Spring of 1998 with a range offilm actors, directors, producers and writers showed
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction,
98
(Suppl 1), 79–103
90
Melanie Wakefield
et al.
surprisingly little awareness of product placement andpoor awareness of the EIC initiative, but highlighted theuse of smoking to elucidate aspects of a character and therole of actor’s own smoking status as a determinant ofsmoking on film [135]. Irrespective of whether sophisti-cated product placement of cigarettes continues, film-makers are likely to make autonomous decisions toinclude portrayal of smoking in movies for reasons of‘artistic licence’.
The depiction of smoking in movies has been arguedby some as a possible reason for increased smoking initi-ation among adolescents in the 1990s. Some argue thatpopular films should be considered as important as par-ents and teachers as models of values, beliefs and behav-iors [116]. Distefan and colleagues [136] argue thatmovie stars have the potential to be even more powerfulrole models than parents and teachers because of thehigh visibility of their behavior on and off screen andtheir larger-than-life status.
When smoking is portrayed as normative, teenagersare more likely to overestimate smoking prevalence, whichmay increase their predisposition for taking up smoking.A recent study found that having a favorite movie star whosmokes on and off screen is associated among teens withsusceptibility to become a smoker [136].
All the research undertaken so far about the way inwhich tobacco is portrayed in movies has focused onquantitative assessment, such as minutes of exposure totobacco images [119] and adult perceptions and ratingsof movie images and scenes involving tobacco [119–122]. An important step in understanding the influencethat smoking in movies may have on teenagers is toexplore how teenagers perceive and appraise portrayal oftobacco in movies. As Chapman & Davis [130] explain,the appearance of a cigarette in a movie ‘does not haveone or even a small number of fixed meanings that can beunproblematically lifted into any context to produce thesame meaning for any audience. Rather, depending oncontext and character, cigarettes can be used to signify awide range of meanings, some of which might actuallypromote
negative
associations with smoking.’ Thus, it isimportant to determine what kind of messages teens takeaway from movies about tobacco.
Pechmann & Shih [137] have undertaken whatappears to be the only experimental study to date that hasassessed how on-screen smoking might influence youngviewers. These investigators compared the responses of9th graders to movies with smoking scenes left intactwith those where the scenes were edited out profession-ally. Compared with non-smoking scenes, smokingscenes positively aroused young viewers (as measured ona seven-point scale from ‘boring’ to ‘exciting’), enhancedtheir perceptions of smoker’s social stature, and increasedtheir intent to smoke. However, showing teenagers an
antismoking advertisement before the movie nullifiedthese effects. Viewing a movie with the smoking scenesprofessionally edited out did not change the viewer’s lik-ing of the movie.
Amos and colleagues [126,138–140] have under-taken research examining the impact of cigarette imag-ery in fashion magazine pictures. They found thatinclusion of a cigarette influenced how pictures wererated by teenagers, and by smokers and non-smokers[140]. Smoking images were rated as more ‘druggy’, wildand depressed, and non-smoking images as more healthy,rich, nice, fashionable, slim and attractive. However,smokers and non-smokers rated themselves in the sameway as they differentiated between smokers and non-smokers in the photographs, so that for smokers, smok-ing-related attributes were considered desirable. Thiseffect was strongest for males aged 15–16 than forfemales or younger and older adolescents. The magazineimage studies emphasize that perception varies withobserver characteristics. In addition, what a cigarette sig-nifies was found to vary considerably in different images,depending on the prominence of the cigarette, presence ofother cues such as types of clothes and background, andthe ‘strength’ of the image [139,140].
Apart from this magazine research, there has been nowork to investigate the mediating role of viewer charac-teristics in determining differential viewer effects in elec-tronic and print media images. As pointed out bySolomon & Englis [141], as movie or television audiencesdo not identify these forms of entertainment as persuasivecommunications, product placements are not identifiedas commercial messages. This means that images aboutsmoking may not generate counter-arguments as theymight for advertising messages, so that the persuasiveimpact of the communication may be enhanced. Furtherresearch in this area is needed.
News coverage on smoking
Television and radio are the media to which most teen-agers are exposed, but newspapers frequently set theagenda for what is reported later in the day on television,so that the synergy between print and electronic news isimportant, and newspapers serve as an accessible proxyfor all news reporting.
While news coverage would be unlikely to influenceteenage tobacco use directly there is good reason toexpect that it may have an indirect influence, by settingthe agenda for discussions about tobacco among parents,older siblings and friends, in schools and in other com-munity activities. The news media has enormous influ-ence over the way in which issues are presented forconsideration by the public. If it were possible to quantifyall news media coverage of tobacco, it would almost
Media and youth smoking
91
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs
Addiction, 98 (Suppl 1), 79–103
certainly be the case that in aggregate, this coveragewould eclipse routinely even the most intensive coveragegained through formal counteradvertising campaigns[142–143]. Some of the most potent examples of newsmedia reporting have been highly memorable—for exam-ple, presidential candidate Bob Dole’s foot-in-mouth sagaover tobacco policy [144], the tobacco industry testifyingbefore a congressional subcommittee that tobacco is notaddictive, and Sixty Minutes’ coverage of Jeffrey Wigand’sstory (which was eventually made into the film TheInsider). Some of the news coverage material is planned aspart of a media advocacy strategy to support antismokingmedia campaigns and/or the passage of tobacco controllegislation [142,145]. News reports that bring attentionto or complement an antismoking campaign may act toincrease interpersonal discussion about relevant issues,which can provide additional benefit.
There have been two community trials involving ele-ments of media advocacy about tobacco within the con-text of a broader set of strategies designed to influenceyouth smoking. Both of these, the Midwestern PreventionProject [59,146–148] and Project Sixteen [67], are sum-marized in Table 2. The design of these intervention stud-ies did not permit disaggregation of the effects of themedia component from other intervention elements.
News coverage may have an influence not onlybecause of the frequency of reporting about tobaccoissues, but through framing stories in particular ways.‘Framing’ refers to the context in which a story is pre-sented—elements such as the background, tone andslant of the article. This is important because it is knownthat people are highly sensitive to contextual cues whenthey make decisions, formulate judgements or expressopinions [149]. Framing of the way in which news storiesare crafted to define the problem of tobacco not only sug-gests to policy-makers and the public why tobacco isimportant, but suggests appropriate solutions to thetobacco problem. In this way, the media tells people notonly what issues to think about in relation to tobaccocontrol, but how to think about them [150]. For example,an analysis of newspaper coverage of the tobacco settle-ment in the front section of the Washington Post foundthat tobacco was framed as a ‘youth issue’ in 55% of arti-cles, but as a deadly product in none of the articles [151].
There is some evidence from other fields that mediaportrayal influences public perceptions. Using time-seriesanalysis, Fan [152] showed that change in public percep-tion of drugs being the most important problem facingthe United States (from 5% in 1985 to 60% in 1994) wasexplained by increases in the news media describingdrugs as ‘a crisis’, whereas other types of discussion aboutdrugs contributed negligibly. In the reverse direction,public opinion about drugs was a weak but significantcontributor to press coverage. In addition, there has been
other work that has linked news coverage about crimeand violence to inaccurate beliefs about violence,attitudes about criminal sentencing, fears of personalsafety and satisfaction with law enforcement [153–155].
The way in which the media frame issues for publicconsumption may also influence behavior. At least short-term changes in self-reported crack and cocaine use by12th graders was observed following extensive news cov-erage of two star athletes who died after using crackcocaine in the mid-1980s [156]. News coverage aboutbinge drinking was found to be related to teenage disap-proval of binge drinking and decline in binge drinkingbehavior from 1978 to 1996 [157]. Laugesen & Meads[158] in New Zealand showed that cigarette consumption,measured by weekly purchases of tobacco from a selectednumber of tobacco outlets, was related significantly to theweekly number of news stories about tobacco, althoughthe effect was short-lived. They estimated that a doublingof news coverage had approximately the same impact as a10% increase in price. To date, apart from the workinvolved in the ASSIST evaluation, there has been rela-tively little work which has sought to systematically mon-itor the amount and type of news stories about tobaccodespite this being called for [143,159] and no studies haveassessed how news coverage might influence teenagesmoking-related beliefs, attitudes and behavior. This is anarea that is ripe for further research.
The tobacco industry is highly successful in creating‘controversy’ about tobacco issues through effective use,and perhaps manipulation, of news media. News cover-age about tobacco has been found to differ in Australiannewspapers [160,161] and in newspapers for schools inthe United States [162,163] according to whether thenewspaper is owned by individuals or groups associatedwith tobacco companies, with fewer tobacco control sto-ries likely to be published. Similarly, magazine acceptanceof cigarette advertising appears to influence the likelihoodof whether magazines will run articles about the risks ofsmoking [164,165] and whether articles about passivesmoking label the research as ‘controversial’ [166].Tobacco company internal documents reveal well-orchestrated disinformation campaigns, often usingindustry ‘front groups’ to cast doubt on the findings of sci-entific reports about passive smoking [167]. There is rel-atively little research that has attempted to link to whatextent youth perceive that there is ‘controversy’ aboutsuch issues, and explores what this might do to their per-ceptions about risk and, ultimately, their likelihood of tak-ing up smoking. However, the literature is clear that bothyouth and adults do not really understand the risks, andunderestimate the addictiveness, of smoking [168–171]so that misinformation will be likely to further confoundrisk perceptions. Additional study of information thatcould potentially cause youth to maintain, or further
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
92 Melanie Wakefield et al.
underestimate the risk of smoking and passive smoking,and how this might translate into changes in the likeli-hood of smoking uptake, is needed.
THEORETICAL APPROACHES TO UNDERSTANDING MEDIA INFLUENCE
The consequences of media dissemination of images,ideas, themes and stories are commonly discussed underthe rubric of media effects. Media effects research encom-passes the study of how the media influences knowledge,opinion, attitudes and behavior among individuals,groups, institutions and communities, and also examineseffects flowing in the opposite direction—from audiencesto media. This research framework regards audiences asactive seekers and users of information, rather than pas-sive recipients. Because of their relevance to publichealth efforts to guide social and behavioral change, thissection first considers four media effects perspectives:knowledge gap, agenda setting, cultivation studies, andrisk communication.
Four general theoretical perspectives on media effects
The knowledge gap hypothesis proposes that an increas-ing flow of information into a social system (from a cam-paign, for example) is more likely to benefit groups ofhigher socio-economic status (SES) than those of lowerSES [172]. This conceptual framework was importantbecause it emphasized that the media have a differentialinfluence on audiences that may be moderated by socialand environmental factors. While early work providedsupport for the general contention, subsequent studiesfound that knowledge gaps were not intractable andcould be minimized by using different channels of com-munication via groups and institutions, and tailoringmessage content [173].
Another field of media effects research has been con-cerned with the mass media’s role and influence in set-ting the public agenda of important problems and issues.An axiom of this approach is that mass media may not beso successful in telling us what to think, but they are veryeffective in telling us what to think about. Kosicki [174]suggests that there are three streams of agenda-settingresearch: public agenda-setting, which examines the linkbetween media portrayal of issues and priorities asassigned by the public; policy agenda-setting, whichexplores the connection between media coverage and thelegislative agenda of policy-making bodies; and mediaagenda-setting, which focuses on factors that influencethe media to cover certain issues. More recent refine-ments within agenda-setting theory have encompassedthe notion that mass media do have a significant impact
on what we think. This line of enquiry stems from the fieldof semiotics, or the science of symbols, signs and codesand the meanings they convey in a variety of social con-texts. As early as 1959, Levy [175] proposed that peoplebuy products for what they mean as well as for what theydo. According to this more recently incorporated view[173] the media provide ways of thinking about specificissues by the signs and symbols used to define the issues—thus, health and social problems are social constructions,defined symbolically through media images and mes-sages, and built by various community groups, institu-tions and advocates. The context of a television programor movie, for example, can provide a social context for aparticular product (such as a cigarette) and the product,in turn, adds to the social and cultural atmosphere of theprogram. Much of the linkage between the movie or tele-vision program and the product is non-conscious andaffective transfer from the program to the product ishypothesized to occur [128].
Russell [128] hypothesizes that the construct of ‘audi-ence connectedness’ defines intense relationshipsbetween the audience and a television program thatextend beyond the television watching experience intoindividuals’ personal and social lives. Highly connectedaudiences are more susceptible to the consumptionimages presented in the television programs, so that thedegree of connectedness moderates the effectiveness ofproduct placement efforts [128]. Although highly con-nected audiences are influenced by the products por-trayed in ‘their’ show, they do not necessarily perceive thecommercial intent of the technique. Russell [128,176]suggests further that each individual–program relation-ship can be conceived as being on a continuum from noinvolvement to fanaticism. Understanding and evaluat-ing these differences may be crucial in assessing the trueimpact of advertising and other media-based marketingefforts, such as product placements [128].
A third perspective, cultivation theory, is concernedwith the impact that mass media may have on percep-tions of ‘reality’. Gerbner’s cultivation theory is based onthe premise that television images cultivate dominanttendencies of a culture’s beliefs and ideologies. Cultiva-tion theory suggests that cumulative exposure to partic-ular media messages is likely to encourage the audienceto develop a consonant world view [153]. In cultivationtheory, the amount of viewing is considered to be a pri-mary variable in television’s impact on behavior. Heavytelevision viewers are less selective in their viewing,engage in habitual viewing and experience a good deal ofsameness of content. According to cultivation theory,general concepts of social reality develop in heavy televi-sion viewers. Thus, for example, heavy television viewers,because they are exposed to more television violence,significantly overestimate the prevalence of violence in
Media and youth smoking 93
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
society. Light viewers are more likely to have manyother sources of information that take up time and dis-place television viewing time.
One of the criticisms of cultivation theory is that itconceptualizes television exposure at the general leveland does not discriminate between different types of pro-grams [177,178]. Another criticism is that the directionof causality is unclear—although heavy viewers maydevelop a particular perception of the world, this percep-tion may also determine their viewing habits and behav-iors. Finally, there has been only limited empiricalapplication of the theory, with mixed results [179].
In contrast to cultivation theory which places empha-sis on television content, uses and gratifications theory isconcerned with the actual motivations of viewers, theuses they make of television and the needs they have thatmight be satisfied by the media [180]. Thus, it looks atwhat people do with the media, rather than what themedia do to them. According to uses and gratificationstheory, there are four main uses of the media: to obtaininformation and knowledge; for diversion (includingstimulation, relaxation or emotional release); for socialintegration (including overcoming loneliness, or allow-ing parasocial relationships with television characters);and for withdrawal (to provide a barrier or avoid chores)[173]. In this model, the audience selects and uses con-tent that will best meet their needs, and the same pro-gram may gratify needs in different audience members.There is distinction between instrumental and ritualisticviewing. Instrumental viewing refers to a goal directeduse of the media, where there is greater involvement. Rit-ualized viewing is more frequent and non-selective.
Uses and gratifications theory has been criticized bysome for its assumption that use of the media is voluntaryand it does not recognize that users do not always have achoice about what is viewed [1]. Furthermore, viewersmay have many different needs at the same time, whichmay be satisfied by a wide range of programs [179].
Developmental level is also important in a uses andgratifications approach—it influences not only the mediachoices that are made, but also how people interpretmedia content. According to primary socialization theory,television can be instrumental in helping with resolutionof the developmental tasks of adolescence by providingdirect learning, as a coping mechanism, for stimulatingfantasies or interpersonal discussion of options [181].Arnett [182] proposes that media is a secondary factor inthe socialization process of adolescents, but may be a pri-mary factor in cases where the traditional primary factorsare lacking, such as parents, siblings and peers. In gen-eral, teenagers have much greater control over theirmedia choices than over socialization from parents, theirschool and their community. This results in more self-socialization and potentially a lack of integration in ado-
lescent socialization as they receive conflicting messagesfrom different sources. The theory suggests that the pri-mary socialization process influences the effects of mediathrough selection (choices of media and program types),selective perception (how media messages are inter-preted) and exposure norming (interaction with primarysocialization sources after exposure can change interpre-tation of media messages) [181]. It may well be that themedia is more influential now than even 10 years ago,due to lower parental involvement with children.
Van Evra’s [179] integrative model of media effectssuggests different media effects depending upon use ofmedia (information versus diversion); perceived reality(real versus unreal); amount of viewing (heavy versuslight) and information alternatives (few versus many).Thus, a maximum effect might be expected when viewingfor information, where the content is perceived as real,there is a heavy viewing pattern, and there are few infor-mation alternatives. Van Evra [179] also posits thatdevelopmental level, race, gender and other factors deter-mine use made of television, reality perceived, amountviewed and information alternatives. Thus, there is acomplex interaction of viewer characteristics, use andamount of viewing, perceived reality and informationalternatives that are proposed to mediate effects.
A fourth media effects research perspective is that ofrisk communication. At the individual level, researchersemphasize the cognitive mechanisms by which individu-als attend to and interpret information about risk, andwhether and how this influences their behavior[183,184]. Theoretical bases for this research stem fromthe Health Belief Model, the Theory of Reasoned Action(for example, perception of personal risk susceptibility toand severity of smoking-related illness) and social cogni-tive theory (for example, self-efficacy for quitting smok-ing) [173]. At the community level, risk communicationstudies focus upon the interaction of populations andsocial institutions in the formulation and management ofpublic opinion and policy making about risk. The theoret-ical basis for this field of study is from borrowed theagenda-setting perspective and also from research intothe definition and framing of public issues.
In the public health literature, it is accepted that con-ceptually, the media does more than directly educate itsaudience. Mass media can also support (reinforce oldmessages, support health changes, encourage mainte-nance of health changes or keep public health issues onthe agenda), promote (publicize products and services)and play a supplementary role by being part of a largerand complementary community-based program oftobacco control activity [185]. Theory suggests that massmedia are more effective if they stimulate interpersonalcommunication [186–188]. Thus mass-media cam-paigns that stimulate interpersonal communication and
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
94 Melanie Wakefield et al.
contribute to mobilizing social support for behaviorchange may improve the likelihood of behavioral effects.In mass media research, opinion leaders have beenassumed to play an important role in receiving, interpret-ing and communicating messages to those around them[189]. Thus, although media may assist the dissemina-tion of new information, interpersonal discussion isinstrumental in persuading people to make changes tobehavior. Mass media campaigns can also stimulatepurposive information seeking.
In summary, these differing perspectives on mediaeffects emphasize that individuals vary in the way theywatch television and use other forms of media. Because ofthis, and due to other differences between audiences, dif-ferent individuals perceive different messages as beingsalient within the same communication or set of commu-nications. In addition, interpersonal factors can mediatethe messages received and interpreted from mediasources.
Theoretical perspectives relevant to persuasive communication
The study of advertising has been influenced predomi-nantly by theories from social psychology [190]. Earlymodels, which focused on processes within the individ-ual, were labeled hierarchies of effect [191,192] and sug-gested that consumers process information fromadvertisements through an ordered series of stages. Forexample, McGuire [192] suggests that, in order to beinfluenced by a message the audience must be exposed tothe message, pay attention to and understand the mes-sage, and develop a cognitive or affective response to themessage. These types of models assume that failure at anyof the steps will lead to no response to the advertising, andthat attitudes and behaviors in response to advertisingare developed consciously and rationally [190].
In the 1970s, an influential model of advertisementprocessing was the Fishbein & Ajzen [193] multi-attribute model. This attitude-based model suggestedthat a consumer’s attitude toward any brand or service isdetermined by summing the consumer’s evaluativeresponse toward each individual product attribute, mul-tiplied by a subjective estimate of the probability that thebrand in question actually possesses the attribute.According to this model, an advertisement changesbrand attitude either by changing a person’s perceptionof the probability that a brand has some attribute, or bychanging a person’s evaluative beliefs about theattribute. Another group of models influential at thistime focused on the influence of advertising on memory[194,195] laying a foundation for the use of memory byadvertising practitioners as an index of the effectivenessof advertising.
Since the early 1980s however, the advertisingresearch literature has been influenced by thinking aboutthe level of involvement with advertising. Various modelsconceptualize involvement in slightly different ways, ascentral rather than peripheral processing [196], system-atic rather than peripheral processing [197] or brandrather than non-brand processing [198]. A dominantmodel, the elaboration likelihood model (ELM), is a dual-process model of attitude change [196] which has impli-cations for persuasive communications, such as advertis-ing. The ELM suggests that attitude change can be eithercentral route (based on effortful information processingactivity aimed at uncovering the central merits of anissue) or peripheral route (based on low effort attitudechange processes). The route used depends upon level ofmotivation and ability to assess central merits of a mes-sage. Thus, when motivation or ability to process a mes-sage is low attitudes are more likely to be changed byrelatively simple associations, such as classical condition-ing or heuristics retrieved from memory. Attitudes formedvia this route are hypothesized to be less enduring andless likely to lead to long-term behavior change.
As shown by Petty et al. [199], people exposed underlow involvement conditions agree with a message more ifthere are more arguments, while people under highinvolvement conditions agree with a message more if thearguments are more compelling. Thus, at the low end ofthe elaboration continuum, it is the quantity of cues thataffects the degree of persuasion, while at the high end ofthe continuum, it is the quality of the argument thataffects the amount of persuasion.
There are also affective models of advertising recep-tion. Some of these assume that emotion operates indi-rectly on brand attitudes via attitude toward theadvertisement [200,201]. Others suggest that emotionbecomes associated through conditioning with other ele-ments in the advertisement, including memory for thewhole advertisement and attitude toward the brand. Srull[195] suggests that when consumers are asked to evalu-ate brands as they process advertisements, moodsinduced by the advertisements affect brand attitudes viamood state association. Previous research has shown aconfident memory advantage exists for negative messages[202] and that inducing negative mood in the targetgroup increases the likelihood of central processing ofinformation [203]. This framework has implications forthe line of research that seeks to investigate appraisal of‘types’ of antismoking ads, as discussed later.
CONCLUSIONS
Overall, it can be seen that in the public health literaturethe alignment between theory and empirical studies has
Media and youth smoking 95
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
generally been poor. This is due probably to the fact thatmuch of the empirical literature is sourced from research-ers who do not have a background that spans consumerprocessing, psychology and advertising, and are unfamil-iar with the literature. However, much of the progress inimplementing planned state-wide antismoking cam-paigns has been made only in the past decade, following aprevious decade of progress focused on school and com-munity interventions. At least in the case of government-funded antismoking campaigns, those who would evalu-ate campaign messages have only a small influence overthe selection, development and placement of antismokingadvertising, and theory testing takes a firm second placeto determining whether the campaign is having animpact on population smoking prevalence, by way ofmeeting the demands of public accountability for cam-paign expenditure. Nevertheless, population monitoringresearchers in California and Massachusetts have madeexcellent use of their data to investigate media influences,such as tobacco marketing and antismoking advertising,on youth. However, understanding antismoking mes-sages that are more or less effective with youth hasreceived mixed informed attention and more research isrequired, although high emotional arousal does appearto be an important ingredient. Economists have alsoplayed an important role in comparative studies of theinfluence of tobacco advertising across countries andover time. Analysis of news media about tobacco issues isin its infancy and much remains to be learned [143].Most of the published research that has studied place-ment of tobacco messages in television programs andmovies has been virtually theory-free, and this field is ripefor further investigation, perhaps along the lines pro-posed by Russell [128].
Pathways of media influence
Although this paper has for the most part discussed sep-arately different types of media in relation to youthsmoking, it is a truism that teenagers are exposed to allof these types of media and that the entire media envi-ronment needs to be considered. It is possible that con-sistent messages about tobacco from different mediachannels may have an amplified effect on youth smok-ing, rather than being purely additive. While there ismuch to be learnt, there is nevertheless much that wealready know about the role of the media in influencingyouth smoking. Considering the theoretical perspectiveson media effects and the empirical evidence base for theinfluence of the media on youth smoking, we draw thefollowing conclusions about pathways by which adver-tising (both pro- and antismoking), product placementin media, and news coverage about smoking mightinfluence youth smoking.
First, it seems that media both shape and reflect socialvalues about smoking. The emergence of antismokingcampaigns has occurred because governments believethat the public requires adequate information, additionalmotivation and accessible assistance to quit smoking ornot to start smoking. In turn, campaigns are a very visiblereminder to the public about the undesirability of smok-ing. The extent and framing of news reporting aboutsmoking issues can also be clearly seen as both a reflec-tion of, and reinforcer for, community beliefs and atti-tudes about smoking and tobacco companies.
Secondly, media provide information about smokingdirectly to audiences. Cigarette advertising alerts con-sumers to new brands on the market-place. Antismokingadvertising provides insights about previously little-known health risks of smoking and publicizes the avail-ability of smoking cessation services. Pharmaceuticalcompany advertising lets smokers know about the avail-ability of new medications, such as nicotine replacementtherapy and buproprion, to assist with cessation. Newsreports publicize new laws and policies applying tosmoking.
Thirdly, the media act as a source of observationallearning by providing models which teenagers may seekto emulate. Thus, cigarette advertisements use glamorouswomen (Virginia Slims), tough men (Marlboro) andfriendly fun-loving cartoon characters (Camel) to build abrand image that might appeal to potential customers.Television programs and movies portray particular life-styles and issues, which may be highly involving to teen-agers, so that product placements, or even incidental use,of tobacco in these contexts may be highly appealing.These factors may be important in mediating the per-ceived prevalence of smoking, a factor that is linkedstrongly to increased risk of smoking uptake among teen-agers. On the other side of the equation, antismokingadvertising may significantly reduce perceived smokingprevalence among teenagers and affect other normativeinfluences, such as the acceptability of smoking in indoorpublic places.
Exposure to media messages about smoking also pro-vides direct reinforcement for smoking or not smoking.Thus, exposure to cigarette advertising and promotionsmay act as a cue for a smoker to have a cigarette, or mayeven prompt a relapse among someone who is in the pro-cess of quitting. Exposure to antismoking advertising mayreinforce non-smoking in a person who has recently quit,or can reinforce an intention not to smoke in future.
A considerable body of theory and empirical researchsuggests that media promotes interpersonal discussionabout smoking, which in turn affects ultimate impact onattitudes and behavior. Thus, the views and behaviors ofpeers, parents and close friends may moderate mediamessages about smoking. It would be possible in principle
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
96 Melanie Wakefield et al.
to examine interaction effects between baseline charac-teristics and cigarette advertising recall and appraisal, todetermine whether, for example, very high levels of peerinfluence increase the likelihood of cigarette advertisingprecipitating experimentation with smoking and/or morerapid progression towards regular smoking. Alterna-tively, there may be protective effects offered by clearparental disapproval of smoking, especially for teenagersat particular ages, so that cigarette advertising is lessappealing—in effect, so that there is interruption of thepositive imagery the ads attempt to portray. In contrast,having peers who are smokers may ‘undo’ any potentialprotective effects of antismoking advertising.
Related to this, media can also influence other ‘inter-vening’ behaviors that may make teenage smoking lesslikely. For example antismoking advertising, by makingthe undesirability of smoking more salient, may reduceadults’ and older siblings’ or friends’ willingness to supplycigarettes to younger teenagers. In addition, such adver-tising may encourage more parental disapproval of smok-ing which, in turn, may reduce the likelihood of smokinguptake by teenagers.
At a more macro-level, antismoking media messagescan also set the agenda for other change at the commu-nity, state or national levels. Making the health effects ofsmoking and passive smoking more salient may garnersupport among the public and policy-makers for othertobacco control policies, such as cigarette price increases,restrictions on smoking in public places and on tobaccoadvertising and promotions, and more aggressiveenforcement of youth access laws. Each of these may haveindependent effects on smoking, and may interact withantismoking advertising to influence youth smoking.
Any influence exerted by the media on youth smokinginteracts with a range of other individual, family, peer andcommunity-level factors. Clearly the family is important,because this is the context in which rules about, andhabitual patterns of, media usage are mostly established.Parents may impose limits on how much and what type ofprograms on television their children might watch, theymay make rules about whether their children may viewmovies classified for older viewers, and they may disap-prove of their children owning cigarette promotionalitems. They, and perhaps older siblings, may be avid con-sumers of newspapers and magazines, thereby facilitat-ing access by youth to news stories about tobacco, and/ormagazine advertisements for cigarettes. Thus, parentsmay be active mediators of youth exposure to messagesabout smoking. In families with heavy television andmovie viewing habits, they may be more likely to perceivesmoking as normative [153].
The family is also likely to moderate the way in whichmedia messages are received and appraised. For example,some parents may actively seek to engage their children
in discussion about antismoking advertising, as well astelevision programs and movies, and newsworthy eventsand issues that may involve tobacco. Parental and siblingsmoking behavior and attitudes to smoking may influ-ence the processing of media messages about smoking. Aspreviously alluded to, strong parental disapproval ofsmoking may confer a protective effect on youth, by rais-ing the threshold of influence from exposure to cigaretteadvertising and promotions. Alternatively, permissiveparental and sibling attitudes about smoking may lowerthis threshold and/or may raise the threshold of influencefrom exposure to antismoking advertising. If parents orsiblings smoke, the availability of cigarettes in the house-hold may facilitate youth acting on temptations to try acigarette, perhaps after having watched a movie where afavorite move star smokes.
Peer influence is also likely to be an important moder-ator of media effects. To the extent that persuasive mediamessages may elicit discussion, those with whom a youthdiscusses what was seen will probably have an importanteffect on eventual appraisal of the message. Thus, if manyclose friends are experimenting with or are already regu-lar smokers and discount antismoking advertising, allother things being equal, the child may be also morelikely to discount the advertising message.
There are also likely to be cultural differences in con-sumption of different types of media. For example, teen-agers in Finland are much more likely to read thenewspaper than American teenagers and tend to usemedia in general for informational purposes rather thanentertainment purposes, the prime concern of Americanteenagers [204]. African Americans spend more timewatching television than whites, but patterns of use byHispanics have only recently become a subject of study[205]. Cultural differences in identification with differentsmoking-related media messages are also likely, and helpto explain the positioning of different cigarette brands fordifferent racial and ethnic groups.
School-based smoking prevention programs canimprove the likelihood that antismoking advertising orcommunity-based activities that include news mediaadvocacy about tobacco can reduce the uptake of smok-ing, and that effects seem more robust when exposureoccurs during preteen or early teenage years, as opposedto later ages. This may be due to a range of factors, such asdevelopmental age and likelihood of already havingexperimented with smoking, as discussed previously.
Tobacco policy is linked clearly to the extent to whichmedia messages about smoking may be influential. Thereis evidence that increases in tobacco tax and media anti-smoking campaigns may operate at least in an additive, ifnot synergistic way, to reduce tobacco consumption.Strong tobacco control laws may assist youth to perceivesmoking as non-normative, and may reduce the
Media and youth smoking 97
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
pervasiveness of tobacco advertising and promotionalmessages. The extent to which cigarette advertising ispermitted through state or local laws may also influencethe effectiveness of antismoking advertising.
Areas for further research
On the basis of the preceding discussion, the followingresearch areas are emphasized as priorities for furtherstudy.1 Is there an age or developmental stage window of opti-
mum influence for tobacco advertising and marketingon smoking uptake and does this vary by other impor-tant personal characteristics such as gender, race/eth-nicity, smoking intentions and past smokingexperience? To what extent do family and peer relation-ships and smoking habits mediate receptivity totobacco advertising and marketing strategies? Studiesthat use a longitudinal approach with regular follow-up from preadolescence to late adolescence arerequired to better assess the temporal relationshipbetween tobacco advertising and marketing exposureand smoking outcome, exploring whether there aredifferential effects by baseline characteristics. Thisresearch might, for example, be able to identify constel-lations of baseline characteristics that interact with acritical level of advertising and marketing exposure toresult in a rapid uptake trajectory. In contrast, it maybe possible to identify factors that may protect againstsuch a trajectory. This kind of research should alsoinclude a focus upon advertising for specific brandsand look for changes in uptake of those brands by gen-der and race/ethnicity.
2 A similar series of questions can be asked about the dif-ferential influence of antismoking advertising and alongitudinal design with multiple follow-up wouldagain be informative. This research could also examinethe effect of ‘competing’ exposures of antismokingadvertising and tobacco advertising and marketing onyouth smoking uptake, to determine under what cir-cumstances and for which individuals, antismokingadvertising could have most beneficial impact. Also,are there teenage smokers who may perceive that anti-smoking messages do not apply to them because theydo not label themselves as smokers, or consider theirlevel of smoking unlikely to be harmful?
3 What strategies are used by the tobacco industry to off-set the potential effects of antismoking campaigns andother tobacco control strategies? For example, is thereevidence of more intensive advertising and marketingefforts in states with ongoing antismoking campaigns?Is there evidence of industry involvement in dissemi-nating misinformation through news reports and soon? Monitoring of marketing efforts, news reports and
key informant interviews in a sample of communitiesin states with and without ongoing antismoking cam-paigns could assist in pursing this research agenda.
4 What characteristics of antismoking advertisementsmost lead teenagers to seriously stop and think orengage in discussion about smoking? What character-istics of advertisements are associated with low levelsof cognitive engagement or interpersonal communica-tion? What kind of influence do advertisements aimedprimarily at adults have on teenagers? For example, doadvertisements promoting nicotine replacement ther-apy have an unintended effect on teenagers in giving amessage that quitting is easier than it actually is? Addi-tional audience studies which test a wide range ofadvertising themes as well as executional styles arerequired.
5 What strength and configuration of antismokingadvertising efforts might best reduce youth smoking?To what extent do other tobacco control policies inter-act with antismoking efforts, once variation in tobaccoadvertising and marketing is controlled? Is there amultiplicative or additive relationship between anti-smoking advertising and other tobacco control policiesin effects on youth smoking, or are there thresholdsinvolved for optimum effect? Can teenage-directedantismoking advertising have an impact on youthsmoking in the absence of broader effects to change thenormative environment for smoking through advertis-ing aimed at reducing smoking among adults? Amulti-level analysis could be informative in answeringthese questions. Such a design could use a nation-widesample of teenage respondents, onto which archivalmedia market records of antismoking advertisingcould be mapped, as well as state and local tobaccocontrol policies and local observations of tobaccoadvertising and marketing activities.
6 How do teenagers appraise the range of youth smokingprevention efforts presently being undertaken bytobacco companies? When compared with state andnational funded antismoking advertisements, how dotobacco company youth prevention advertisementsrate? What effects do tobacco company promotionalefforts to portray themselves as good corporate citi-zens, such as those by the Philip Morris group of com-panies, have on teenage perceptions of the tobaccoindustry in general and their views about tobacco inparticular? The monitoring of youth awareness andappraisal of such advertising and other efforts could beinformative in this regard.
7 Given that restrictions on tobacco advertising arebecoming more common, and that it is known thatcompanies respond to restrictions by using othermarketing strategies, what new marketing strategiesare emerging which may influence teenagers? A
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
98 Melanie Wakefield et al.
monitoring system involving routine point-of-purchase observation and scanning of magazines andother media outlets, or one that used volunteer smok-ers to alert tobacco control advocates to newlyencountered marketing efforts could provide earlywarning of new strategies and raise the need for anassessment of their attractiveness to youth.
8 How do teenagers perceive the different types of smok-ing portrayals that appear in movies and what mes-sages do they take away from viewing? Importantly,what viewer characteristics might predispose teenag-ers to take away favorable impressions of smokingwhen it is portrayed as such in a movie setting? Arethere measures that can be taken reliably to nullifypotentially favorable impressions, such as broadcast-ing antismoking ads prior to the movie?
9 What is the relationship between the extent and tone ofnews coverage on tobacco issues, whether this bethrough newspaper reports, magazine stories or elec-tronic news, on teenager’s smoking perceptions, inten-tions and behavior? Does favorable tobacco controlnews coverage amplify any effects of antismokingadvertising on reducing youth smoking and con-versely, does coverage unfavorable to tobacco controlreduce effects?
ACKNOWLEDGEMENTS
The research and writing of this paper was supported byfunding from the Robert Wood Johnson Foundationthrough the Tobacco Etiology Research Network, theNational Cancer Institute through grant no. 1 R01CA86273-01 ‘Youth Smoking and the Media’.
REFERENCES
1. Kubey, R. W. & Csikszentmihalyi, M. (1998) Television asescape: subjective experience before and evening of heavyviewing. Communication Reports, 3, 92–100.
2. American Medical Association (1996) Physician’s Guide toMedia Violence. Chicago, IL: American Medical Association.
3. Klein, J. D., Brown, J. D. & Walsh, K. (1993) Adolescent’srisky behavior and mass media use. Pediatrics, 92, 24–31.
4. Greenberg, B. S., Linsangan, R. L. & Soderman, A. (1987)Adolescents and their exposure to television and moviesex. Project CAST report no. 4. East Lansing, MI: MichiganState University. Department of Telecommunications.
5. Brown, D. & Bryant, J. (1989) Uses of pornography. In:Zillmann, D. & Bryant, J., eds. Pornography: ResearchAdvances and Policy Considerations, pp. 3–24. Hillsdale, NJ:Lawrence Erlbaum.
6. Federal Trade Commission (1999) Tar, nicotine and car-bon monoxide of the smoke of 1252 varieties of domestic
cigarette for the year 1997. http://www.ftc.gov. Lastaccessed Sept 1999.
7. Turner-Bowker, D. & Hamilton, W. L. (2000) Cigaretteadvertising expenditures before and after the MasterSettlement Agreement: preliminary findings.http://www.state.ma.us/dph/mtcp/report/mag.htm. Lastaccessed Oct 2001.
8. Wakefield, M. A., Terry-McElrath, Y. M., Chaloupka, F. J.,Barner, D. C., Slater, S. J., Clark, P. I. & Giovino, G. A.(2002) Tobacco industry marketing at the point-of-purcahse: After the 1998 MSA billboard advertising ban.American Journal of Public Health, 92, i73–i80.
9. US Department of Health and Human Services (1989)Reducing the Health Consequences of Smoking: 25 Years ofProgress. A report of the US Surgeon General. Rockville,MD: US Department of Health and Human Services, PublicHealth Service, Centers for Disease Control, Centre forChronic Disease Prevention and Health Promotion, Officeof Smoking and Health.
10. Kaufman, N. J. & Nichter, M. (1999) The Culture of theBody: The Influence of Media, Fashion, and Marketing onWomen’s Tobacco Use. Proceedings of the World Health Orga-nization, July 1999. World Health Organization, Geneva.
11. US Department of Health and Human Services (1994) Pre-venting Tobacco Use Among Young People. A report of theSurgeon General. Atlanta, GA: US Department of Healthand Human Services, Public Health Service, Centers forDisease Control and Prevention. National Center forChronic Disease Prevention and Health Promotion.
12. Cummings, K. M., Giovino, G. & Mendicino, A. J.(1987) Cigarette advertising and black–white prefer-ences in brand preference. Public Health Reports, 102,689–701.
13. King, C., Siegel, M. & Pucci, L. G. (2000) Exposure of blackyouths to cigarette advertising in magazines. Tobacco Con-trology, 9, 64–70.
14. Hackbarth, D. P., Silvestri, B. & Cosper, W. (1995)Tobacco and alcohol billboards in 50 Chicago neighbor-hoods: market segmentation to sell dangerous products tothe poor. Journal of Health Policy, 16, 213–230.
15. Schooler, C., Feighery, E. & Flora, J. A. (1996) Seventhgraders’ self-reported exposure to cigarette marketing andits relationship to their smoking behaviour. American Jour-nal of Public Health, 86, 1216–1221.
16. Luke, D., Esmundo, E. & Bloom, Y. (2000) Smoke signs:patterns of tobacco billboard advertising in a metropolitanregion. Tobacco Control, 9, 16–23.
17. Jarvis, M. J. & Bates, C. (1999) Why Low Tar CigarettesDon’t Work and How the Tobacco Industry has Fooled theSmoking Public. Action on Smoking and Health UK.http://www.ash.org.uk. Last accessed 24 October 2001.
18. Pollay, R. W. & Dewhurst, T. (2001) Marketing cigaretteswith low machine-measured yields. In: Institute, N. C., ed.Risks Associated with Smoking Cigarettes with Low MachineMeasured Yields of Tar and Nicotine. Smoking and TobaccoControl Monograph no. 13, NIH publication no. 02–5074. Bethesda, MD: US Department of Health andHuman Services, National Institute of Health.
19. Lewit, E. M., Coate, D. & Grossman, M. (1981) The effectsof government regulation on teenage smoking. Journal ofLaw and Economics, 24, 545–569.
20. Pierce, J. P., Lee, L. & Gilpin, E. A. (1994) Smoking initia-tion by adolescent girls, 1944 through 1988. Journal of theAmerican Medical Association, 271, 608–611.
Media and youth smoking 99
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
21. Pierce, J. & Gilpin, E. (1995) A historical analysis oftobacco marketing and the uptake of smoking by youth inthe United States: 1890–1977. Health Psychology, 14,500–508.
22. Cummings, K. M., Shah, D. & Shopland, D. R. (1995)Trends in smoking initiation among adolescents andyoung adults—United States, 1980–89. Morbidity andMortality Weekly Reports, 44, 521–525.
23. Pollay, R., Siddarth, S., Siegel, M., Haddix, A., Merrit, R.,Giovino, G. & Erikson, M. (1996) The last straw? Cigaretteadvertising and realised market shares among youths andadults, 1979–93. AMA Journal of Marketing, 60, 1–16.
24. Chaloupka, F. J. & Warner, K. E. (2000) The economics ofsmoking, In: Newhouse, J. P. & Cuyler. A. J., eds. The Hand-book of Health Economics, pp. 1539–1627. New York:North Holland, Elsevier Science BV.
25. Warner, K. E. (1986) Selling Smoke: Cigarette Advertisingand Public Health. Washington, DC: American PublicHealth Association.
26. Saffer, H. (1998) Economic issues in cigarette and alcoholadvertising. Journal of Drug Issues, 28, 781–793.
27. Saffer, H. & Chaloupka, F. (2000) The effect of tobaccoadvertising bans on tobacco consumption. Journal ofHealth Economics, 19, 117–137.
28. Pierce, J. P., Farkas, A. J. & Evans, N. (1993) Tobacco usein California in 1992. A Focus on Preventing Uptake in Ado-lescents. Sacromento, CA: California Department ofHealth Services.
29. Difranza, J. R., Richards, J. W. & Paulman, P. M. (1991)RJR Nabisco’s cartoon camel promotes cigarettes to chil-dren. Journal of the American Medical Association, 226,3149–3153.
30. O’Connell, D. L., Alexander, H. M., Dobson, A. J., Lloyd,D. M., Hardes, G. R., Springthorpe, H. J. & Leeder, S. R.(1981) Cigarette smoking and drug use in schoolchildren,II. Factors associated with smoking. International Journal ofEpidemiology, 10, 223–231.
31. Potts, H., Gillies, P. & Herbert, M. (1986) Adolescentsmoking and opinion of cigarette advertisements. HealthEducation and Research, 1, 195–201.
32. Botvin, E. M. & Botvin, G. J. (1991) Adolescent smokingbehaviour and the recognition of cigarette advertisments.Journal of Applied Social Psychology, 21, 919–932.
33. Chapman, S. & Fitzgerald, B. (1982) Brand preference andadvertising recall in adolescent smokers: some implica-tions for health promotion. American Journal of PublicHealth, 72, 491–494.
34. Goldstein, A. O., Fischer, P. M., Richards, J. W. & Creten,D. (1987) Relationship between high school studentsmoking and recognition of cigarette advertisements. Jour-nal of Pediatrics, 110, 488–491.
35. Difranza, J. R. & Aisquith, B. F. (1995) Does the Joe Camelcampaign preferentially reach 18–24 year old adults?Tobacco Control, 4, 367–371.
36. Mccann, J. E. (1992) Tobacco logo recognition. Journal ofFamily Practice, 34, 681–684.
37. Aitken, P. P. & Leathar, D. S. (1985) Brand stretching’advertisements for cigarettes: the impact on children.Health Education Journal, 44, 201–202.
38. Aitken, P. P., Leathar, D. S. & O’Hagan, F. J. (1985) Chil-dren’s perceptions of advertisements for cigarettes. SocialScience and Medicine, 21, 785–797.
39. Aitken, P. P., Leathar, D. S., O’Hagan, F. J. & Squair, S. J.(1987) Children’s awareness of cigarette advertisements
and brand imagery. British Journal of Addiction, 82, 12–22.
40. Aitken, P. P. & Eadie, D. R. (1990) Reinforcing effects ofcigarette advertising on under-age smoking. British Jour-nal of Addiction, 85, 399–412.
41. Peters, J., Betson, C. L. & Hedley, A. J. (1995) Recognitionof cigarette brand names and logos by young children inHong Kong. Tobacco Control, 4, 150–155.
42. Alexander, H. M., Callcott, R. & Dobson, A. J. (1983) Cig-arette smoking and drug use in school children. Interna-tional Journal of Epidemiology, 12, 59–66.
43. Armstrong, B., De Klerk, N., Shean, R., Dunn, D. & Dolin,P. (1990) Influence of education and advertising on theuptake of smoking by children. Medical Journal of Austra-lia, 152, 117–124.
44. Aitken, P. P., Eadie, D. R., Hastings, G. B. & Haywood, A. J.(1991) Predisposing effects of cigarette advertising onchildren’s intentions to smoke when older. British Journalof Addiction, 86, 383–390.
45. Pucci, L. G. & Siegel, M. (1999) Exposure to brand-specificcigarette advertising in magazines and its impact on youthsmoking. Preventive Medicine, 29, 313–320.
46. Pierce, J., Choi, W., Gilpin, E., Farkas, A. & Berry, C.(1998) Tobacco industry promotion of cigarettes and ado-lescent smoking. Journal of the American Medical Associa-tion, 279, 511–515.
47. Biener, L. & Siegel, M. (2000) Tobacco marketing and ado-lescent smoking: more support for a causal inference.American Journal of Public Health, 90, 407–411.
48. Sargent, J. D., Dalton, M. A. & Beach, M. L. (2000) Expo-sure to cigarette promotions and smoking uptake in ado-lescents: evidence of a dose–response relationship. TobaccoControl, 9, 163–168.
49. Evans, N., Farkas, F., Gilpin, E., Berry, C. & Pierce, J.(1995) Influence of tobacco marketing and exposure tosmokers on adolescent susceptibility to smoking. Journal ofthe National Cancer Institute, 87, 1538–1545.
50. Altman, D. G., Levine, D. W., Coeytaux, R., Slade, J. & Jaf-fer, R. (1996) Tobacco promotion and susceptibility totobacco use among adolescents aged 12 through 17 yearsin a nationally representative sample. American Journal ofPublic Health, 86, 1590–1593.
51. Feighery, E., Borzekowski, D. L. G., Schooler, C. & Flora, J.(1998) Seeing, wanting, owning: the relationshipbetween receptivity to tobacco marketing and smokingsusceptibility in young people. Tobacco Control, 7, 123–128.
52. Gilpin, E. A., Pierce, J. P. & Rosbrook, B. (1997) Areadolescents receptive to current sales promotion prac-tices of the tobacco industry? Preventive Medicine, 26,14–21.
53. Unger, J. B. & Chen, X. (1999) The role of social networksand media receptivity in predicting age of smoker initia-tion: a proportional hazards model of risk and protectivefactors. Addictive Behavior, 24, 371–381.
54. Redmond, W. (1999) Effects of sales promotion on smok-ing among US ninth graders. Preventive Medicine, 28,243–250.
55. Vartiainen, E., Pallonen, U. A. M., Koskela, K. & Puska, P.(1986) Four-year follow-up results of the smoking preven-tion program in the North Karelia Youth Project. Preven-tive Medicine, 15, 692–698.
56. Vartiainen, E., Pallonen, U., Mcalister, A. L. & Puska, P.(1990) Eight year follow-up of an adolescent smoking
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
100 Melanie Wakefield et al.
prevention program: the North Karelia Youth Project.American Journal of Public Health, 80, 78–79.
57. Vartiainen, E., Paavola, M., Mcalister, A. & Puska, P.(1998) Fifteen-year follow-up of smoking preventioneffects in the North Karelia Youth Project. American Jour-nal of Public Health, 88, 81–85.
58. Perry, C. L., Kelder, S. H., Murray, D. M. & Klepp, K. I.(1992) Community-wide smoking prevention: long-termoutcomes of the Minnesota Health Program and the Classof 1989 Study. American Journal of Public Health, 82,1210–1216.
59. Johnson, C. A., Pentz, M. A. & Weber, M. D. (1990) Rela-tive effectiveness of comprehansive community program-ming for drug abuse prevention with high-risk and low-risk adolescents. Journal of Consulting Clinical Psychology,58, 447–456.
60. Tobler, N. & Stratton, H. (1997) Effectiveness of school-based drug prevention programs: a meta-analysis of theresearch. Journal of Primary Prevention, 18, 71–128.
61. Flynn, B. S., Worden, J. K. & Secker-Walker, R. H. (1992)Prevention of cigarette smoking through mass mediaintervention and school programs. American Journal ofPublic Health, 82, 827–834.
62. Flynn, B. S., Worden, J. K., Secker-Walker, R. H. & Pirie, P.L. (1994) Mass media and school intervention for ciga-rette smoking prevention: effects 2 years after completion.American Journal of Public Health, 84, 1148–1150.
63. Flynn, B. S., Worden, J. K. & Secker-Walker, R. H. (1997)Long-term responses of higher and lower risk youths tosmoking prevention interventions. Preventive Medicine,26, 389–394.
64. Secker-Walker, R. H., Worden, J. K., Holland, R. R., Flynn,B. S. & Detsky, A. S. (1997) A mass media programme toprevent smoking among adolescents: costs and cost-effec-tiveness. Tobacco Control, 6, 207–212.
65. Worden, J. K., Flynn, B. S. & Geller, B. M. (1988) Develop-ment of a smoking prevention mass media program usingdiagnostic and formative research. Preventive Medicine, 17,531–558.
66. Worden, J. K. & Flynn, B. S. (2000) Effective use of massmedia to prevent smoking. Journal of Public Health Manage-ment and Pract, 6, vii–viii.
67. Biglan, A., Ary, D., Smolkowski, K., Duncan, T. & Black, C.(2000) A randomised controlled trial of a communityintervention to prevent adolescent tobacco use. TobaccoControl, 9, 24–32.
68. Pierce, J. P., Dwyer, T. & Frape, G. (1986) Evaluation ofthe Sydney ‘Quit For Life’ antismoking campaign. MedicalJournal of Australia, 144, 341–347.
69. Pierce, J., Macaskill, A. & Hill, D. (1990) Long-term effec-tiveness of mass media led antismoking campaigns inAustralia. Australian Journal of Public Health, 80, 565–569.
70. Macaskill, P., Pierce, J. P., Simpson, J. M. & Lyle, D. M.(1992) Mass media-led antismoking campaign canremove the education gap in quitting behavior. AmericanJournal of Public Health, 82, 96–98.
71. Murray, D. M., Prokhorov, A. V. & Harty, K. C. (1994)Effects of a state-wide antismokig campaign on massmedia messages and smoking beliefs. Preventive Medicine,23, 54–60.
72. Murray, D. M., Perry, C. L. & Griffin, G. (1992) Resultsfrom a state-wide approach to adolescent tobacco use pre-vention. Preventive Medicine, 82, 449–472.
73. Popham, W. J., Potter, L. D. & Hetrick, M. A. (1994) Effec-tiveness of the California 1990–91 tobacco educationmedia campaign. American Journal of Preventive Medicine,10, 319–326.
74. Pierce, J., Gilpin, E. & Emery, S. (1998) Has the CaliforniaTobacco Control Program reduced smoking? Journal of theAmerican Medical Association, 280, 893–899.
75. Unger, J., Johnson, C. A. & Rohrbach, L. (1998) Integrationof Tobacco Control Efforts. Independent Evaluation Consor-tium. Final report of the independent evaluation of the Cal-ifornia Tobacco Prevention and Education Program: Wave1 Data, 1996–97. Rockville, MD: Gallup Organization.
76. Harris, J., Connolly, G., Brooks, D. & Davis, B. (1996) Ciga-rette Smoking Before and After an Excise Tax Increase andan Antismoking Campaign – Massachusetts, 1990–1996.Morbity and Mortality Weekly Report, 45, 966–970.
77. Biener, L., Harris, J. E. & Hamilton, W. (2000) Impact of theMassachusetts tobacco control program: population basedtrend analysis. British Medical Journal, 321, 351–354.
78. Briton, N. J., Clark, T. W. & Baker, A. K. (1997) Adoles-cent Tobacco Use in Massachusetts: Trends Among PublicSchool Students, 1984–96. Boston: Health and AddictionsResearch, Inc.
79. Massachusetts Department of Education (1998) TobaccoUse Among Massachusetts High School Students. 1997Massachusetts Youth Risk Behavior Survey results.http://www.doe.mass.edu/doedocs/health/tobrpt98.html. Last accessed July 2001
80. Soldz, S., Kreiner, P., Clark, T. & Krakow, M. (2000)Tobacco use among Massachusetts youth: is tobacco con-trol working? Preventive Medicine, 31, 287–295.
81. Siegel, M. & Biener, L. (2000) The impact of an antismok-ing media campaign on progression to established smok-ing: results of a longitudinal youth study. American Journalof Public Health, 90, 380–386.
82. Sussman, S., Dent, C. W., Stacy, A. W. & Craig, S. (1998)One-year outcomes of Project Towards No Drug Abuse.Preventive Medicine, 27, 632–642.
83. Chassin, L., Presson, C. C., Sherman, S. J., Corty, E. &Olshavsky, R. W. (1984) Predicting the onset of cigarettesmoking in adolescence: a longitudinal study. Journal ofApplied Social Psychology, 14, 224–243.
84. Wakefield, M. & Chaloupka, F. (2000) Effectiveness ofcomprehensive tobacco control programmes in reducingteenage smoking in the USA. Tobacco Control, 9, 177–186.
85. Sly, D., Heald, G. R. & Ray, S. (2001) The Florida ‘Truth’anti-tobacco media evaluation: design, first year resultsand implications for planning future media evaluations.Tobacco Control, 10, 9–15.
86. Sly, D., Hopkins, R., Trapido, E. & Ray, S. (2001) Influenceof a counteradvertising media campaign on initiation ofsmoking: the Florida ‘truth’ campaign. American Journal ofPublic Health, 91, 233–238.
87. Bauer, U., Johnson, T., Pallentino, J. & R. H. (1999)Tobacco use among middle and high school students—Florida, 1998 and 1999. Morbidity and Mortality WeeklyReport, 48, 248–253.
88. Johnston, L. D., O’Malley, P. M. & Bachman, J. G. (1999)Cigarette Smoking among American teens continuesgradual decline (Monitoring the Future). Available at:http://www.monitoringthefuture.org/pressreleases/99cigpr.html.
89. Bauer, U., Johnson, T., Hopkins, R. & Brooks, R. (2000)Changes in youth cigarette use and intentions following
Media and youth smoking 101
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
implementation of a tobacco control program: findingsfrom the Florida Youth Tobacco Survey, 1998–2000. Jour-nal of the American Medical Association, 284, 723–728.
90. Oregon Health Division (1999) Tobacco Prevention and Edu-cation Program Report 1999. Portland, OR: Department ofHuman Resources.
91. Eisenberg, M., Lee, H. & Burgoon, M. (1998) HistoricalImpact of the TEPP Media Campaign. Report no. 1. Tuscon,AZ: University of Arizona, Arizona Cancer Center.
92. Burgoon, M. (1999) School survey report: Impact ofthe TEPP Media Campaign on the Youth Population ofArizona. Tuscon, AZ: Arizona Cancer Center, Univer-sity of Arizona.
93. Oregon Health Division, Department of Human Services(2000) Delivering Results: Saving Lives and Saving Money,Tobacco Education and Prevention in Orego. Program Report2000. http://www.ohr.hr.state.or.us/tobacco/arpt2000/welcome.html. Last accessed July 2001.
94. Pizacani, B., Mosbeck, C. & Hedberg, K. (1999) Decline incigarette consumption following implementation of acomprehensive tobacco prevention and education pro-gram—Oregon, 1996–98. Mortality and Morbidity WeeklyReports, 48, 140–143.
95. Institute of Medicine and National Research Council(2000) State Programs Can Reduce Tobacco Use. Washing-ton DC: National Academy of Sciences, National CancerPolicy Board, Institute of Medicine, National ResearchCouncil and Board on Health Promotion and Disease Pre-vention, Institute of Medicine.
96. Goldman, L. K. & Glantz, S. A. (1998) Evaluation of anti-smoking advertising campaigns. Journal of the AmericanMedical Association, 279, 772–777.
97. Balch, G. & Rudman, G. (1998) Antismoking advertisingcampaigns for youth. Journal of the American Medical Asso-ciation, 280, 323–324.
98. Worden, J. K., Flynn, B. S. & Secker-Walker, R. H. (1998)Antismoking advertising campaigns for youth. Journal ofthe American Medical Association, 280, 323.
99. Pechmann, C. & Goldberg, M. (1998) Evaluation of AdStrategies for Preventing Youth Tobacco Use. Berkeley, CA:California Tobacco-Related Disease Research Program.
100. Peracchio, L. A. & Luna, D. (1998) The development of anedvertising campaign to discourage smoking initiationamong children and youth. Journal of Advertising, 27, 49–55.
101. Teenage Research Unlimited (1999) Counter-TobaccoAdvertising Exploratory. The states of Arizona, California,and Massachusetts Public Health Anti-Tobacco MediaCampaigns: Summary Report, January–March.Northbrook, IL, USA: Teenage Research Unlimited.
102. Hill, D., Chapman, S. & Donovan, R. (1998) The return ofscare tactics. Tobacco Control, 7, 5–8.
103. Wakefield, M., Miller, C. & Roberts, L. (1999) Comparisonof the National Tobacco Campaign with a youth targettedSouth Australian campaign. In: Hassard, K., ed. Australia’sNational Tobacco Campaign Evaluation Report, vol. 1. Can-berra: Commonwealth Department of Health and AgedCare.
104. Harrison Health Research (1998) ‘Stressing Out’ Preven-tion Campaign Evaluation. Adelaide, South Australia: Har-rison Health Research .
105. Aylesworth, A. B. & Mackenzie, S. B. (1998) Context is key:the effect of program-induced mood on thoughts about thead. Journal of Advertising, 27, 17–31.
106. Warner, K. E. (1979) Clearing the airwaves: the cigarettead ban revisited. Policy Analysis, 5, 435–450.
107. Schneider, L., Klein, B. & Murphy, K. (1981) Governmen-tal regulation of health information. Journal of Law andEconomics, 24, 575–612.
108. Baltagi, B. H. & Levin, D. (1986) Estimating dynamicdemand for cigarettes using panel data: the effects of boot-legging, taxation and advertising reconsidered. Review ofEconomics and Statistics, 68, 148–155.
109. Pechmann, C. & Ratneshwar, S. (1994) The effects of anti-smoking and cigarette advertising on young adolescents’perception of peers who smoke. Journal of ConsumerResearch, 21, 236–251.
110. Pechmann, C. & Knight, J. (2002) A experimental investi-gation of the joint effects of advertising and peers on ado-lescent beliefs and intentions about cigaretteconsumption. Journal of Consumer Research, 29, 5–19.
111. Levy, D. T. & Friend, K. (2000) Gauging the effect of massmedia policies: what do we need to know? Journal of PublicHealth Management and Practice, 6, 95–106.
112. Rao, R. & Miller, P. (1975) Advertising/sales responsefunctions. Journal of Advertising Research, 15, 7–15.
113. Ackoff, R. L. & Ernshoff, J. R. (1975) Advertising researchat Anheuser-Busch, Inc. (1963–68). Sloan ManagementReview, 16, 1–15.
114. Pechmann, C. (1997) Does antismoking advertising com-bat underage smoking? A review of past practices andresearch. In: Goldberg, M. E. Fishbein, M. & Middlestadt, S.E., eds. Social Marketing: Theoretical and Practical Perspec-tives, pp. 189–216. Mahwah, NJ: Lawrence Erlbaum Asso-ciates.
115. Slater, S. J., Chaloupka, F. & Wakefield, M. A. (2001) Statevariation in retail promotions and advertising forMarlboro cigarettes. Tobacco Control, 10, 337–339.
116. Terre, L., Drabman, R. S. & Speer, P. (1991) Health-relevant behaviors in media. Journal of Applied Social of Psy-chology, 21, 1303–1319.
117. Veronis, S. A. A. (1996) The Veronis, Shuler and AssociatesCommunications Industry Forecast. New York: Veronis,Suhler and Associates.
118. Hazan, A. R. & Glantz, S. A. (1995) Current trends intobacco use in prime-time fictional television. AmericanJournal of Public Health, 85, 116–117.
119. Stockwell, T. F. & Glantz, S. A. (1997) Tobacco use isincreasing in popular films. Tobacco Control, 6, 282–284.
120. Goldstein, A. O., Sobel, R. A. & Newman, G. R. (1999)Tobacco and alcohol use in g-rated children’s animatedfilms. Journal of the American Medical Association, 281,1131–1136.
121. McIntosh, W. D., Bazzini, D. G., Smith, S. M. & Wayne, S.M. (1998) Who smokes in Hollywood? Characteristics ofsmokers in popular films from 1940 to 1989. AddictiveBehaviors, 23, 395–398.
122. Escamilla, G., Crdock, A. L. & Kawachi, I. (2000) Womenand smoking in Hollywood movies: a content analysis.American Journal of Public Health, 90, 412–414.
123. Cruz, J. & Wallack, L. (1986) Trends in tobacco use on tele-vision. American Journal of Public Health, 76, 698–699.
124. Diener, B. J. (1993) The frequency and context of alcoholand tobacco cues in daytime soap opera programs: Fall1986 and Fall 1991. Journal of Public Policy Marketing, 12,252–257.
125. Durant, R. H., Romes, E. S., Rich, M. & Allred, E. (1997)Tobacco and alcohol use behaviors protrayed in music vid-
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
102 Melanie Wakefield et al.
eos: a content analysis. American Journal of Public Health,87, 1131–1135.
126. Amos, A. (1992) Style and Image—Tobacco and AlcoholImages in Selected Style and Fashion Magazines. London, UK:Health Education Authority.
127. Wasko, J., Phillips, M. & Purdie, C. (1993) Hollywoodmeets Madison avenue: the commercialization of US films.Media Culture and Society, 15, 271–293.
128. Russell, C. A. (1998) Towards a framework of productplacement: theoretical propositions. Advances in ConsumerResearch, 25, 357–362.
129. Hadju, D. (1988) Commercials on cassette. Video Review,December, 33–35.
130. Chapman, S. & Davis, R. (1997) Smoking in movies: is it aproblem? Tobacco Control, 6, 269–271.
131. Colford, S. W. (1990) Tobacco group ends paid place-ments. Advertising Age, 61, 31.
132. Vanzi, M. (1997) Lawmaker targets on-screen smoking:movies influence youngsters to light up, says Sen. JohnBurton, who will hold a hearing in effort to ‘hightenawareness’ among filmmakers. Los Angeles Times, October26, p. A–3.
133. US Newswire (1997) Gore statement of entertainmentindustry and tobacco. OTC COMTEX Newswire/US News-wire, 3 December 1997.
134. Madigan, N. (1997) Hollywood alliance denouncestobacco use. RTve (Reuters/ Variety Entertainment) NewsService, December 5.
135. Shields, D. L. L., Carol, J., Balbach, E. D. & McGee, S. (1999)Hollywood on tobacco: how the entertainment industryunderstands tobacco portrayal. Tobacco Control, 8, 378–386.
136. Distefan, J. M., Gilpin, E. A. & Sargent, J. D. (1999) Domovie stars encourage adolescents to start smoking? Evi-dence from California. Preventive Medicine, 28, 1–11.
137. Pechmann, C. & Shih, C. F. (1999) Smoking scenes inmovies and antismoking advertisements before movies:effects on youth. Journal of Marketing, 63, 1–13.
138. Gray, D., Amos, A. & Currie, C. (1996) Exploring youngpeople’s perceptions of smoking images in youth maga-zines. Health Education Research, 11, 215–230.
139. Gray, D., Amos, A. & Currie, C. (1997) Decoding theimage—consumption, young people, magazines andsmoking. An exploration of theoretical and methodologi-cal issues. Health Education Research, 12, 505–517.
140. Amos, A., Currie, C., Gray, D. & Elton, R. (1998) Percep-tions of fashion images from youth magazines: does a cig-arette make a difference? Health Education Research, 13,491–501.
141. Solomon, M. R. & Englis, B. G. (1994) The big picture:product complementarity and integrated communica-tions. Journal of Advertising Research, 34, 57–64.
142. Stevens, C. (1998) Designing an effective counteradvertis-ing campaign—California. Cancer, 83, 2736–2741.
143. Chapman, S. (1999) The news on tobacco control: time tobring the background into the foreground. American Jour-nal of Public Health, 79, 1419–1421.
144. Davis, R. M. (1996) Doling out soundbites in the presiden-tial campaign. Tobacco Control, 5, 317–339.
145. Wallack, L. (1994) Media advocacy: a strategy for empow-ering people and communities. Journal of Public Health Pol-icy, Winter, 15(4), 421–436.
146. Mackinnon, D. P., Johnson, C. A. & Pentz, M. A. (1991)Mediating mechanisms in a school-based drug prevention
program: first year effects of the Midwestern PreventionProject. Health Psychology, 10, 164–172.
147. Pentz, M., Dwyer, J. & Mackinnon, D. (1989) A multicom-munity trial for primary prevention of adolescent drugabuse. Journal of the American Medical Association, 261,3259–3266.
148. Pentz, M. A., Mackinnon, D. P. & Dwyer, J. H. (1989) Lon-gitudinal effects of the Midwestern Prevention Project onregular and experimental smoking in adolescents. Preven-tive Medicine, 18, 304–321.
149. Iyengar, S. (1991) Is Anyone Responsible? How TelevisionFames Political Issues. Chicago: University of ChicagoPress.
150. Menashe, C. L. & Siegel, M. (1998) The power of a frame:an analysis of newspaper coverage of tobacco issues—United States, 1985–96. Journal of Health Communication,3, 307–325.
151. Lima, J. C. & Siegel, M. (1999) The tobacco settlement: ananalysis of newspaper coverage of a national policy debate,1997–98. Tobacco Control, 8, 247–253.
152. Fan, D. P. (1996) News media framing sets public opinionthat drugs is the country’s most important problem. Sub-stance Use and Misuse, 31, 1413–1421.
153. Gerbner, G., Gross, L., Signorielli, N. & Morgan, M. (1980)Television violence, victimization and power. AmericanBehavioral Scientist, 23, 705–716.
154. Roberts, J. V. & Doob, A. N. (1990) News media influenceson public views of sentencing. Law and Human Behavior,14, 451–468.
155. Graber, D. A. (1980) Crime News and the Public. New York,NY: Praeger.
156. Fan, D. P. & Holway, W. B. (1994) Media coverage ofcocaine and its impact on usage patterns. InternationalJournal of Public Opinion Research, 6, 139–162.
157. Yanovitsky, I. & Stryker, J. (2001) Mass media, socialnorms, and health promotion efforts: a longitudinal studyof media effects on youth binge drinking. CommunicationResearch, 28, 208–239.
158. Laugesen, M. & Meades, C. (1991) Advertising, price,income and publicity effects on weekly cigarette sales inNew Zealand supermarkets. British Journal of Addiction,86, 83–89.
159. Wakefield, M. & Chaloupka, F. J. (1998) Improving themeasurement and use of tobacco control ‘inputs. TobaccoControl, 7, 333–335.
160. Chapman, S. (1984) Not biting the hand that feeds you:tobacco advertsising and editorial bias in Australian news-papers. Medical Journal of Australia, 140, 480–482.
161. Chapman, S. (1989) The news on smoking: editorial cov-erage of tobacco and health issues in Australian newspa-pers, 1987–88. American Journal of Public Health, 79,1419–1420.
162. Balbach, E. D. & Glantz, S. A. (1995) Tobacco informationin two grade school weeklies: a content analysis. AmericanJournal of Public Health, 85, 1650–1653.
163. Dejong, W. (1996) When the tobacco industry controlsthe news: KKR, RJR Nabisco, and the weekly reader cor-poration. Tobacco Control, 5, 142–148.
164. Amos, A., Jacobson, B. & White, P. (1991) Cigarette adver-tising policy and coverage of smoking and health in Britishwomen’s magazines. Lancet, 337, 93–96.
165. Warner, K. E. (1985) Cigarette advertising and media cov-erage of smoking and health. New England Journal of Med-icine, 312, 384–388.
Media and youth smoking 103
© 2003 Society for the Study of Addiction to Alcohol and Other Drugs Addiction, 98 (Suppl 1), 79–103
166. Kennedy, G. E. & Bero, L. A. (1999) Print media coverageof research on passive smoking. Tobacco Control, 8, 254–260.
167. Ong, E. K. & Glantz, S. A. (2000) Tobacco industry efortssubverting International Agency for Research on Cancer’ssecond-hand smoke study. Lancet, 355, 1253–1259.
168. Slovic, P. (1998) Do adolescent smokers know the risks?Duke Law Journal, 47, 1133–1141.
169. Weinstein, N. D. (1999) Accuracy of smokers’ risk percep-tions. Nicotine and Tobacco Research, 1, S123–S130.
170. Slovic, P. (2000) Commentary: the perils of Viscusi’s anal-ysis of smoking risk perceptions. Journal of Behavioral Deci-sion Making, 13(2).
171. Slovic, P. (2000) What does it mean to know a cumulativerisk? Adolescents’ perceptions of short-term and long-term consequences of smoking. Journal of Behavioral Deci-sion Making, 13, 259–266.
172. Tichenor, P. J., Donohue, G. A. & Olien, C. N. (1970) Massmedia flow and differential growth in knowledge. PublicOpinion Quarterly, 34, 159–170.
173. Finnegan, J. R. & Viswanath, K. (1996) Communicationtheory and health bevavior change: the media studiesframework. In: Glanz, K. Lewis, F. M. & Rimer, B. K., eds.Health Behavior and Health Education: Theory, Research andPractice, pp. 313–341. San Francisco: Jossey-BassPublishers.
174. Kosicki, G. M. (1993) Problems and opportunities inagenda-setting research. Journal of Communication, 43,100–127.
175. Levy, S. (1959) Symbols for sale. Harvard Business Review,37, 117–124.
176. Russell, C. A. & Puto, C. P. (1999) Rethinking televisionaudience measures: an exploration into the construct ofaudience connectedness. Marketing Letters, 10, 387–401.
177. Potter, W. J. (1993) Cultivation theory and research: aconceptual critique. Human Communications Research, 19,564–601.
178. Rubin, A. M., Perse, E. M. & Taylor, D. S. (1988) A meth-odological examination of cultivation. CommunicationResearch, 15, 107–134.
179. Vanevra, J. (1998) Theoretical perspectives. In: Vanevra,J., ed. Television and Child Development, pp. 133–152. Mah-wah, NJ: Lawrence Erlbaum.
180. Rubin, A. M. (1993) Audience activity and media use.Communication Monographs, 60, 98–105.
181. Kelly, K. & Donohew, L. (1999) Media and primary social-ization theory. Substance Use and Misuse, 34, 1033–1045.
182. Arnett, J. J. (1995) Adolescents’ uses of media for self-socialization. Journal of Youth and Adolescence, 24, 519–532.
183. Slovic, P., Fischhoff, B. & Lichtenstein, S. (1981) Inform-ing the public about the risks from ionizing radiation.Health Physics, 41, 589–598.
184. Weinstein, N. D. (1984) Why it won’t hapen to me: per-ceptions of risk factors and susceptibility. Health Psychol-ogy, 3, 431–457.
185. Flora, J. A., Maibach, E. W. & Maccoby, N. (1989) The roleof media across four levels of health promotion interven-tion. American Review of Public Health, 10, 181–201.
186. Flay, B. R. & Sobel, J. L. (1983) The role of mass media inpreventing adolescent substance abuse. National Instituteon Drug Abuse Research Monograph Series, 47, 5–35.
187. Flay, B. R. (1987) Mass media and smoking cessation: a crit-ical review. American Journal of Public Health, 77, 153–160.
188. Rogers, E. M. & Storey, J. D. (1987) Communication cam-paigns. In: Berger, C. R. & Chaffee, S. H., eds. Handbook ofCommunication Science, pp. 817–847. Thousand Oaks, CA:Sage.
189. Katz, E. (1957) The two step flow of communication: anupdate report on a hypothesis. Public Opinion Quarterly,21, 61–68.
190. Thorson, E. (1996) Advertising. In: Salwen, M. B. &Stacks, D. W., eds. An Integrated Approach to Communica-tion Theory and Research. Mahwah, NJ: LawrenceErlbaum.
191. Lavidge, R. J. & Steiner, G. A. (1961) A model for predictivemeasurement of advertising effectiveness, Journal of Mar-keting, 61, 59–62.
192. Mcguire, W. J. (1985) Attitudes and attitude change. In:Lindsay, G. & Aronson, E., eds. Handbook of Social Psychol-ogy, pp. 65–72. New York: Random House.
193. Fishbein, M. & Ajzen, I. (1975) Belief, Attitude, Intentionand Behavior: an Introduction to Theory and Research. Read-ing, MA: Addison-Wesley.
194. Srull, T. K. (1984) Affect and memory: the impact of affec-tive reactions in advertising on the representation of prod-uct information in memory. In: Bagozzi, R. P. & Tybout, A.M., eds. Advances in Consumer Research, pp. 572–576. SanFrancisco: Association for Consumer Research.
195. Srull, T. K. (1984) The effects of subjective affective stateson memory and judgement. In: Kinnear, T. C., ed.Advances in Consumer Research, pp. 530–533. SanFrancisco: Association for Consumer Research.
196. Petty, R. E. & Cacioppo, J. T. (1986) Attitude Change. Cen-tral and Peripheral Routes to Persuasion. Advances in Experi-mental Social Psychology. New York: Springer-Verlag.
197. Chaiken, S. (1980) Heuristic versus systematic informa-tion processing and the use of source versus message cuesin persuasion, Journal of Personality and Social Psychology,39.
198. Gardner, M. P., Mitchell, A. A. & Russo, J. E. (1985) Lowinvolvement strategies for processing advertisements.Journal of Advertising, 14, 4–12.
199. Petty, R. E., Cacioppo, J. T. & Schumann, D. W. (1986)Central and peripheral routes to advertising effectiveness:the moderating role of involvement. Journal of ConsumerResearch, 10, 135–146.
200. Edell, J. A. & Burke, M. C. (1987) The power of feelings inunderstading advertising effects. Journal of ConsumerResearch, 14, 421–433.
201. Mackenzie, S. B., Lutz, R. J. & Belch, G. E. (1986) The roleof attitude toward the ad as a mediator of advertising effec-tiveness: a test of competing explanations. Journal of Mar-ket Research, 23, 130–143.
202. Reeves, B. R., Newhagen, J., Maibach, E. & Basil, M. (1991)Negative and positive television messages: effect of mes-sage type and context on attention and memory. AmericanBehavioral Scientist, 34, 679–694.
203. Eagly, A. H. & Chaiken, S. (1993) The Psychology of Atti-tudes. Fort Worth, TX: Harcourt Brace Jovanovich CollegePublishers.
204. Burton, D., Johnson, C. A., Uutela, A. & Vartiainen, E.(1990) Media use patterns among Finnish and Americanyouth: implications for smoking intervention. Family andCommunity Health, 13, 73–81.
205. Romer, D. & Kim, S. (1995) Health interventions forAfrican American and Latino youth: the potential role ofmass media. Health Education Quarterly, 22, 172–189.