Gambling in young adolescents
Gambling in Young Adolescents
Susan G. Ide-Smith and Stephen E.G. Lea
University of Exeter, Department of Psychology
A questionnaire was used to investigate gambling in British adolescents. Re-
sponses from fifty 13- to 14-year-olds were analyzed. Gambling was found to be very
pervasive (90 % of subjects reported at least some gambling activity). Males gambled
more than females, both in absolute terms and as a proportion of income. Income was
found to have some influence on gambling behavior, but the effects of intelligence and
social class were nonsignificant. Slot machines were the commonest form of gambling
in both sexes.
In his review of the literature on gambling, Cornish (1978) stated
that almost 80% of the adult population of Britain gamble in some
way. Cornish classified “adults” as being over sixteen. Society defines
some forms of gambling as adult, by the legal restrictions on the
minimum age for participation, and such legislation presumably has
some influence on the popularity of these types of gambling with
children. However, given that such a large proportion of the adult
Reprint requests and all correspondence should be addressed to: S.E.G. Lea, University of
Exeter, Department of Psychology, Washington Singer Laboratories, Exeter EX4 4QG, Eng-
land.
The willing cooperation of the teacher and students concerned is gratefully acknowledged.
This paper is based on a report submitted by the first author in partial fulfillment of the
requirements for the BSc degree at the University of Exeter. Some of the data were presented to
the conference of the British Psychological Society Developmental Section, held in Exeter in
September 1986, and to the colloquium of the International Association for Research in Eco-
nomic Psychology, held at Aarhus, Denmark, in September 1987.
Journal of Gambling Behavior Vol. 4(2), Summer 1988
110 © 1988 Human Sciences Press
SUSAN G. IDE-SMITH AND STEPHEN E.G. LEA 111
population gambles, it would seem highly probable that most individ-
uals are first introduced to some form of gambling when they are still
under sixteen. Yet there seems to have been almost no formal research
into the way in whicli children begin to gamble (Lea, Tarpy & Webley,
1987, Chapter 14).
A priori, there are two routes through which children might ac-
quire gambling habits. One is through direct exposure to adult gam-
bling, presumably in the company of parents, or older friends and
relatives. As a result of legislative changes in the past 25 years, adult
gambling is now a highly visible activity in Britain (Cornish, 1978),
and children will frequently see betting shops, advertisements for
commercial gambling, and low-stake coin-in-the-slot gaming ma-
chines. However, gambling or its precursors may also be present
within the autonomous culture of children’s games. Direct gambling
for money stakes is only rarely recorded by those who have studied
children’s games (e.g. Sutton-Smith, 1972), and Lea, et al. argue that
chance, as distinct from risk, is relatively uncommon in the games
which are maintained by children themselves, as distinct from propri-
etary games. On the other hand, Moody (1985) demonstrates with a
vivid anecdote that chance can certainly be highly attractive to chil-
dren. Risk may well be more common, and risk taking may bring the
kind of social reward that might encourage a child towards a gambling
habit.
The present investigation was designed to gather some systematic
data on the acquisition of gambling habits. Pilot work suggested that
several forms of gambling are widespread among adolescents, whereas
a substantial majority of nine-year-old children reported no gambling
behavior at all. Accordingly, the present study concentrated on young
people in their early teens, which seemed likely to be a period of rapid
transition. The aims were to discover how prevalent gambling is in this
group, and to see whether exposure to adult gambling, or the culture of
children themselves, first introduces children to gambling.
The extent and nature of gambling varies substantially in the
adult population as a function of sex, age, social class, and income
(Cornish, 1978; Downes, Davies, & Stone, 1976). Accordingly, the
sample was selected so as to be broadly representative of the teenage
population on these variables, and attempts were made to measure all
of them. It was felt that the teenager’s own income, rather than family
income, was likely to be the most relevant economic variable.
112 JOURNAL OF GAMBLING BEHAVIOR
M E T H O D
Subjects
The subjects were 51 pupils from a neighborhood comprehensive high school in
Exeter, Devon; they comprised two of the three classes in one school year group. There
were 31 male and 20 female pupils in the sample. Their mean age was 13.7 years. One
class consisted of 24 pupils, the other of 27. The study was carried out during the
(compulsory) lessons on Personal, Social and Moral education.
The pupils were in their third year of secondary education. The classes were not
streamed, so the subjects were of mixed abilities. The high schools in this city receive
approximately 85% of the 13- to 16-year-old population, so the sample could be
expected to be representative of the teenagers in the city, except for some underrepre-
sentation of the extremes of the ability and parental income ranges. Exeter is a typical
English provincial city, with a population of about 100,000, serving as a commercial
and administrative center for an extensive agricultural and tourist hinterland.
Questionnaire Design
The questionnaire consisted of nine pages. The first two requested background
information. The following five pages asked virtually identical questions about five
kinds of gambling which pilot work had identified as likely to occur among teenagers.
The next page asked the same questions about any other kind of gambling the subject
could identify as something he or she did, while the final page allowed for a free
response to the questionnaire and towards gambling in general. The language used
was as simple as possible throughout.
The background information sought concerned the subjects’ sex, age, social
status, income, attitudes towards gambling, perceived attitudes towards gambling of
significant others, and intelligence. The last four of these were measured as follows:
Social Status. Participants were asked to state their parents’ current or most recent
job. A value for family social status was subsequently derived from Goldthorpe and
Hope’s (1974) Social Grading of Occupations Scale, using father’s occupation where a
sufficiently precise description of it was given, or mother’s occupation if not.
Income. The participants were asked to state the amount of their income per week
from two sources: pocket money, and “a paper round or a job like that.”
Attitudes. The participants were asked to reply with one of “approve,” “disap-
prove,” or “don’t mind” to questions about what they thought of gambling, and what
they thought their parents and teacher thought of gambling.
Intelligence. The second page of the questionnaire asked the participants to write
one-line definitions of 12 words from the Mill Hill Vocabulary Scale (Raven, 1958).
SUSAN G. I D E – S M I T H A N D S T E P H E N E.G. LEA 113
The definitions were subsequently marked as correct or incorrect by one of the
authors, to give a score for verbal intelligence, on a scale from 0 to 12.
The specific types of gambling investigated were card games, coin games, domi-
noes, slot machines, and “betting with your friends,” i.e. wagering. “Coin games” was
intended to include betting on the toss of a coin, and a locally popular game that
involved throwing pennies against a wall, the winner (usually the player whose penny
lands closest to the wall) taking all the pennies. At the top of the page about each kind
of gambling, the participant was asked, for example, “Do you play slot machines?” and
instructed to turn to the next page if not.
The questions asked about each of these, and the answer categories made avail-
able, were:
How often do you [play]? Often (more than once a week)/Occasionally (one or
two times a month)/Rarely (one or two times a year).
How much money do you [bet at one time]? 1-10p/11-50p/more than 50p.
Where do you play? At home/At school/Elsewhere
Who do you [play] with? Friends/Brothers or sisters/Parents
Who taught you to play this game? Friends/Brothers or sisters/Parents/Other
relations
How old were you when you started playing this game?
When you win what do you do with the money?
For each kind of gambling, a rough measure of expenditure per year for each
participant was calculated by multiplying the amount wagered (scoring 1-10p as 10p,
10-50p as 50p, and more than 50p as £1) by the frequency of playing (scoring often as
52, occasionally as 12 and rarely as one). Clearly this procedure cannot give a very
accurate quantitative estimate of gambling expenditure, but it should produce a
reasonably valid ordinal scale.
Procedure
The questionnaires were completed by the two classes on one afternoon, so that
there was no opportunity for discussion between the classes. The pupils sat at double
desks with a boy and a girl at each desk; since young teenagers tend to have same sex
friends, it was hoped that these seating arrangements would dissuade them from
collaborating. The teacher of the classes was present, as an observer only, for most of
the time.
Before the questionnaires were given out, the first author introduced them. She
stressed the definition of gambling as a voluntary activity involving risk, chance, and
sometimes a degree of skill, gave examples of gambling (football pools, betting on
horse races and greyhound races, card games, lotteries, dominoes, bingo, casino
games, wagering friends, coin games and slot machines) and emphasized that video
games were not a form of gambling. She also stressed that she was not taking a
particular moral stance, but was merely interested in obtaining some information, and
that only she would see the replies. Finally, she told the pupils not to put their names
114 JOURNAL OF GAMBLING BEHAVIOR
on the questionnaire (this instruction was repeated on the first page), and gave general
instructions about how to fill it in.
The questionnaires were then handed out, and collected in from each pupil as
soon as he or she had finished. When the whole class had returned their question-
naires, there was a brief discussion about the study and about gambling in general.
R E S U L T S A N D D I S C U S S I O N
O f the 51 completed quest ionnaires , one had to be discarded
because vir tual ly no informat ion was given. In a few other cases, no
socioeconomic status or weekly income could be de te rmined; data f rom
these subjects are excluded f rom analyses involving these variables.
Tab le 1 summar izes the main characteristics of the sample, b roken
down by sex.
J u d g i n g by the present sample, gambl ing is a pervasive activity in
young Brit ish adolescents of both sexes. Eighty-n ine percent of our
subjects (boys, 93 %; girls, 84 %) claimed to gamble in some way. Th e
most c o m m o n form of gambl ing was slot machines (81% part icipa-
tion), and their popular i ty may have been exaggera ted by the location
of the study: f rom Exeter there is easy access to nea rby seaside towns
where there is an abundance of a m u s e m e n t arcades, and Exeter itself
has one or two arcades. At the seaside, in part icular , chi ldren can get
unres t r ic ted access to slot machines . Some other forms of gambl ing
Table 1
Sample C h a r a c t e r i s t i c s
Sex
Boys Girls
Variable Mean Range Mean Range
N 30 20
Socioeconomic status 51 18-75 35 18-65
Income (£/wk) 5.07 0 .60-15 .00 4.21 1 .50-10.00
Vocabu la ry score 6.8 1-10 6.2 2-8
(/12)
Age (yrs :mos) 13:8 13:0-14:0 13:8 13:0-14:6
SUSAN G. IDE-SMITH AND STEPHEN E.G. LEA 115
were also quite prevalent, for example card games (53% participa-
tion), wagers (51N) and coin games (twenty-six percent). Wagers
reported involved an interesting variety of topics: there was one report
of a conventional gambling topic (greyhound racing), but more typical
were wagers about the subjects’ own activities (fights, races into town),
what would be on TV, the weather, and even earwig races and “a fly
and a moth crawling up a wall.”
The subjects reported starting gambling at surprisingly young
ages, the means being eight yrs three months for boys and eight yrs
nine months for girls. While these retrospective reports need to be
checked with a cross-sectional study, and the youngest ages reported
may well be taken with a pinch of salt, it does seem likely that a
substantial proportion of children, especially boys, are engaging in
some form of gambling by the age of ten.
Reported gambling expenditures were considerable. The highest
figure was £260 per year, which probably reflects the inadequacies of
our estimation procedure. But the mean estimates (£38.45 for boys and
£9.52 for girls), and the mean percentages of income (6.0% and
1.4%), are plausible. At the time of this survey, £1 exchanged for
approximately 1.85 US dollars.
The sex differences in gambling expenditure and the percentage
of income spent on gambling were highly significant. They emerged
both in simple t-tests (t = 3.31 for expenditure, 3.59 for percentage of
income, df = 48, p < 0.01) and in a multiple regression analysis of
gambling expenditure in which age, social status, and income were
also taken into account (the regression coefficients against the dummy
variable of sex was 21.71 £/yr, p < .005). They were found whether
nonparticipants were counted as having zero expenditure, or whether
they were excluded from the analysis.
Sex differences were found for every form of gambling. Downes et
al. (1976) report a similar effect for adults. It is possible, of course, that
it was inflated by a tendency for boys to overreport gambling expendi-
ture, and/or for girls to underreport it. However, even such reporting
biases would still reflect very different attitudes to gambling in the two
sexes. While the present study cannot explain the sex differences in
gambling behavior within our society, therefore, it does show that they
are of fairly early developmental origin. Gambling is not something
that girls and boys sample equally, but only boys persist with; it is
much more part of the culture of boys than of girls, from quite a young
age.
Income had a substantial effect on reported gambling behavior
(regression coefficient of 4.98 £/yr gambling expenditure per £/wk
116 JOURNAL OF GAMBLING BEHAVIOR
income, p < 0.001). It may have been increased because the subjects'
incomes were proportionately small; in effect, most of an adolescent's
income is "discretionary income" (Katona, 1975), i.e., not committed
to necessary expenditure on food, rent, and suchlike. Although not
significant, the effect of income was still positive when reported gam-
bling expenditure was expressed as a proportion of income; this means
that gambling has a positive income elasticity of demand among ado-
lescents, so that it would be classified econometrically as a luxury.
Intelligence and social status had no apparent effect. The mea-
sures of these variables were of course rough and ready, but they were
good enough to detect predicted relationships among the independent
variables; for example, the correlation between vocabulary score and
social status was positive (r = 0.53, p < 0.002; cf. Butcher, 1968,
Chapter 10), and that between social status and income was negative
(r = - 0 . 3 1 , p < 0.05; cf. Reid, 1977, though his result was not
replicated by Furnham & Thomas, 1984). So if these factors had been
strongly related to gambling, the effect would have been detected.
Table 2 summarizes the main qualitative findings about the con-
text of gambling. Sex differences were slight, so these data are not
reported separately for boys and girls. Because of the occurrence of
multiple and null responses, these data were not suitable for statistical
analysis. However, the Table does make it clear that the social contexts
of different forms of gambling are different. Card games and dominoes
are primarily family activities, most often learned and played at home,
and so children clearly acquire these forms of gambling from the adult
world. The same is evidently true of the small amount of "commercial"
(betting shop) gambling reported in this study (three boys said that
they betted on greyhound racing or did football pools, and because of
legal restraints they would have needed adult cooperation to do this).
On the other hand, coin games and wagers mostly occur out of the
home, often at school (despite school rules forbidding all games of
chance). These are clearly part of the autonomous culture of child-
hood; despite a combined residence in Exeter of 15 years, the authors
had never heard of the game of throwing pennies against the wall, in its
local form, before this study.
Slot machines are a category on their own; obviously they involve
the adult world, as suppliers of the machines, but most children
reported first playing them with friends. Several subjects reported that
playing them was something they did on their own; and in discussion
after the questionnaires had been filled in, it was only in relation to slot
machines that the children came forward with anecdotes of anything
SUSAN G. IDE-SMITH AND STEPHEN E.G. LEA 117
Table 2
The Context of Gambling: Percentages of Participants in Each
Kind of Gambling Who Responded as Indicated to the Questions
About Where They Gambled, Who They Gambled With, and Who
Taught Them to Gamble
Gambling Type
Card Coin Slot
Games Games Dominoes m/cs Wagers Other
Number of
participants 2 7 13 5 41 2 5 6
Where do you play?
At home 78% 38% 60% 0% 28% 33%
At School 4 31 0 0 60 0
Elsewhere 37 38 20 100 64 67
Who do you play with?
Friends 52% 92% 40% 59% 96% 100%
Siblings 37 31 20 29 20 17
Parents 44 8 40 20 12 0
Who got you into this?
Friends 22% 69% 40% 44% 68% 83%
Siblings 15 8 20 12 24 17
Parents 44 8 40 17 16 17
Other relatives 41 15 0 15 24 17
like compulsive behavior. Expenditure on slot machines was not well-
correlated with expenditure on other forms of gambling, and it was
very poorly predicted by the independent variables (only 12% of
variance accounted for in the regression analysis, compared with 26 %
for card games, the next most prevalent form of gambling). It may thus
be a somewhat isolated kind of behavior, and, despite its prevalence, it
remains to be seen whether it is this kind of behavior, or more
conventional gambling, that would be a better predictor of an adoles-
cent's subsequent gambling career.
With two boys' questionnaires discarded because of incomplete or
multiple responses, results from the attitude questions were as follows.
Two girls (10%) and eight boys (29%) positively approved of gam-
bling; one girl (5 %) and no boys disapproved; and the rest responded
118 J O U R N A L OF G A M B L I N G B E H A V I O R
"don't mind." One girl and one boy claimed that their parents ap-
proved, two girls and eight boys that they disapproved. One girl and
three boys claimed that their teacher approved of gambling, 14 girls
and 19 boys (70 % and 68 %) that she disapproved. This majority view
was true in the sense that she told the experimenters that she would not
gamble herself, but she did not in fact condemn others who partici-
pated. It is likely that the subjects' perceptions stemmed from their
stereotype image of a teacher as a disciplinarian, and of gambling as an
only one-half legitimate activity.
In several cases subjects attributed disapproval of gambling to
their parents, yet subsequently reported that their parents had intro-
duced them to one or more forms of gambling, and had participated
with them; these data must be contaminated by inconsistency of some
sort. The inconsistency need not, of course, be on the subjects' part; it
too may well be a reflection of the ambiguous status of gambling in the
society in which these adolescents are growing up.
REFERENCES
Butcher, H.J. (1968). Human intelligence. London: Methuen.
Cornish, D.B. (1978). Gambling." A review of the literature. London: Her Majesty's Stationery Office.
Downes, D.M., Davies, B.P., David, M.E., & Stone, P. (1976). Gambling, work and leisure: A study
across three areas. London: Routledge & Kegan Paul.
Furnham, A., & Thomas, P. (1984). Pocket money: A study of economic education. British
Journal of Developmental Psychology, 2, 205-212.
Goldthorpe, J.H., & Hope, K. (1974). Social Grading of Occupations. Oxford: Oxford University
Press.
Katona, G. (1975). Psychological economics. New York: Elsevier.
Lea, S.E.G., Tarpy, R.M., & Webley, P. (1987). The individual in the economy. New York:
Cambridge University Press.
Moody, G.E. (1985). Playing with chance. In Eadington, W.R. (Ed.) The gambling studies:
Proceedings of the sixth national conference on gambling and risk taking Vol. 3, (pp. 317-329). Reno,
NV: University of Nevada.
Raven, J.C. (1958). Extended guide to using the Mill Hill vocabulary scale with the progressive matrices scale
(2nd revision). London: Lewis.
Reid, I. (1977). Social class differences in Britain. London: Open Books.
Sutton-Smith, B. (1972). The folk-games of children. Austin, TX: University of Texas Press.