HAPPY GAMBLERS

1
HAPPY GAMBLERS?

• David Forrest
• and
• Ian McHale
• The Salford Seminar February, 2009

2

• We all know that consumers in all parts of the
  world are restricted in the choices they make
  about how, when and where to gamble
• What can explain this highly protective stance
  towards consumers?
• Government will cite problem gambling concerns in
  justifying their policies

3

• However, problem gamblers account for a small
  part of the population
• Problem drinking rates are higher than problem
  gambling rates but there are relatively few
  restrictions on the choices of responsible users
• The harsher treatment of gambling suggests that
  government believes that those who would gamble
  comprise a vulnerable population that needs
  protection from itself
• Presumably it believes that gambling causes
  problems to individuals that are not picked up by
  formal screens for problem gambling behaviour

4

• In any case, it is clear that gambling policy
  should be driven by more than concerns over
  clinically diagnosed problem gambling
• Attention needs to be paid to the majority of the
  population which gambles without the signs of
  dysfunctional behaviour tested for in problem
  gambling screens
• We take a direct approach to evaluating the
  wellbeing and mental state of the population who
  gamble
• In doing so we follow the tradition of the fast
  growing new discipline of happiness studies

5

• Happiness studies has grown because social
  surveys around the world now routinely ask people
  how happy they are
• It started in the 70s with the US General Social
  Survey which asked the question
• Taken all together how would you say things are
  these days? Would you say you are very happy,
  pretty happy or not too happy?
• very happy3, pretty happy2, not too happy1

6

• The scale has tended to change over time, so that
  there is now usually a ten point rather than a
  three point scale
• But the essence of the question has remained the
  same
• It asks how people rate their life these days-
  ie it attempts not to measure mood today but some
  long-run concept of well being
• Sets of answers exist for many countries, rich
  and poor, for long periods of time with thousands
  of respondents

7
ARE PEOPLES ANSWERS CREDIBLE?

• It seems so
• Statistical models reveal intuitively plausible
  patterns in answers that are stable over time and
  space
• For example, marriage always raises predicted
  happiness score by about 0.6-0.8 points on a ten
  point scale, very good rather than very bad
  health always raises happiness score by about two
  full points
• This suggests that peoples answers are
  considered and sensible and we can with
  confidence ask questions like what difference
  does gambling behaviour make?
• Moreover, psychologists validation studies find
  high correlation between individuals happiness
  scores and other indicators of mental wellbeing
  (eg how often the subject smiles) and other
  peoples assessment of the subjects state of mind

8
The General Approach

• Researchers build a statistical model to account
  for differences between happiness scores across
  individuals
• Key items to be taken into account in any study
  include age, living in a partnership, children,
  health and income
• The researcher then adds the focus of interest
  (here engagement with gambling)
• This is the first study to use a data set where
  there are happiness scores and information on
  gambling behaviour

9
OUR DATA

• From Taking Part, a survey on the use of
  leisure time carried out for DCMS
• Over 28,000 adults in England interviewed during
  2005 and 2006
• Very ambitious survey- median interview was 32
  minutes long

10

• THE HAPPINESS QUESTION
• Taking all things together, how happy would you
  say you are? On a scale of 1-10, 10extremely
  happy, 1 extremely unhappy
• THE GAMBLING QUESTION
• Participation in the preceding seven days in
  thirteen named modes of gambling (scratchcards,
  horse betting, table games, etc, etc)

11

• We built statistical models to account for (i)
  male and (ii) female happiness scores
• We used only white British respondents
• Gambling has a different cultural
  significance in different communities and cleaner
  results are likely if the sample is more
  homogenous
• This gave us sample sizes of 10,128 men and 12,
  772 women
• Few of these will be problem gamblers- the BGPS
  implies about 100 of the males and 13 of the
  females
• These small numbers are not likely to affect
  results very much- our results could therefore
  reasonably be interpreted as showing the
  relationship between happiness and gambling for
  non-problem gamblers

12
DISTRIBUTION OF HAPPINESS SCORES
13
HAPPINESS RESULTS

• male female
• age -0.028 -.020
• age squared 0.0003 0.0003
• married 0.698 0.580
• cohabiting 0.559 0.404
• (compared with single)

14

• male female
• young child -0.067 -0.130
• pre-teen -0.089 -0.039
• teenager -0.169 -0.158
• (compared with no children)
• unemployed -0.557 -0.426
• retired 0.229 0.291

15

• male female
• very good health 0.989 0.928
• good health 0.476 0.504
• bad health -0.454 -0.588
• very bad health -1.061 -0.932
• (compared with fair health)
• Model also included information on income, level
  of education, access to a car

16
Participation by mode (with standard controls in
place)

• Male Female
• National Lottery 0.116 0.096
• other lotteries
  0.142 0.018
• scratchcards 0.188 0.043
• Football pools 0.113
  -0.199
• horse betting 0.106 -0.024
• dog betting -0.747 0.151
• sports betting 0.078 0.922
• table games 0.080 -0.550
• bingo 0.285 0.164
• slots -0.082 0.172
• private betting -0.096
  0.010
• raffles -0.041
  -0.574
• e-gaming
  0.189 1.228

17

• There is a limit to what can be learned here as
  number of participants for some modes is too low
  for statistically significant findings to emerge
• But most signs are positive
• National Lottery players and bingo players (of
  either gender) are strongly significantly happier
  than non-players with similar circumstances
• Female sports bettors appear to be particularly
  happy people
• The only negative and significant sign relates to
  men who bet on dogs they experience low
  happiness relative to their circumstances
• In this one case, results may be driven by the
  presence of (unknown) problem gamblers in the
  sample
• Thats because an exceptional proportion of
  males who bet on dogs are problem gamblers (16
  according to the BGPS)

18
to help reach a broader conclusion.

• we decided to measure the degree of an
  individuals engagement with gambling by the
  number of modes of gambling in which he or she
  had participated in the preceding week
• range was from 0 to 10 for men, 0 to 5 for women

19
Male Female number of modes 0.077
0.072 People who engage in gambling to a
greater degree are happier than similar people
who engage to a lesser degree
20
Of course..

• This finding, while it shows strong positive
  association between happiness and engagement in
  gambling, does not demonstrate CAUSATION
• Causation could go either way
• Gambling might make people happier
• Or it might just be that happier people might be
  more inclined to gamble
• For example, extroversion, risk loving nature,
  sociability are all personality traits linked
  with happiness and each of them may also
  predispose an individual to enjoy gambling

21
a technical note

• Sometimes data allow a test for causation using
  instrumental variables
• Instrumental variables must be correlated with
  the variable of interest (here number of modes)
  but exogenous to the individual
• We used as instruments information about gambling
  events in the week before an interview- size of
  Lotto rollover, Grand National (both important in
  determining female engagement) and World Cup (a
  negative influence on male number of modes)
• Preliminary results are that there is no CAUSAL
  effect running from gambling to happiness

22
RATHER

• Happier people are more likely to be gamblers
• Recall that happiness scores are recognised as
  good indicators of mental health and wellbeing
• Gamblers as a group do not appear to be
  especially vulnerable people
• They are choosing to gamble and there is no
  evidence here that their choices should not be
  respected
• Indeed it might be considered that, since
  gamblers are happy, it would be reckless to
  meddle in their choices
• Does not mean of course that we should forget
  problem gambling
• But it does imply that policies to mitigate harm
  should be targeted at the very vulnerable- they
  should not be too restrictive of the activities
  of the majority, responsible gamblers whom we now
  know also to be happy gamblers

23
POSTSCRIPT

• The instrumental variables model provides results
  from a sub-model of what determines the focus
  variable (here, number of modes)
• This model has strong predictive power and shows
  how family, health and employment status, age,
  education, income etc influence engagement with
  gambling
• Relative to single people, married and cohabiting
  men and women take part in more modes of gambling
• Divorce and widowhood increase engagement with
  gambling for men but not for women
• Less educated people choose more modes of
  gambling
• There is a very strong positive relation with
  income this should be of concern to the industry
  given current pressures on income
• To avoid ending on that negative note, let me say
  again that the main message of this research is
  that gamblers are not sad people. Generally
  they are happy people. And that in itself has
  implications for the stance government should
  take towards gambling.