What Proportion of GamblingRevenue is Derived fromProblem Gamblers?Dr. Robert Williams & Dr. Robert WoodUniversity of LethbridgeAlberta, Canada6 th European Conference on Gambling Studiesand Policy IssuesMalmo, SwedenJune 29, 2005 – July 2, 2005
• The source of gaming revenue hasimportant philosophical, sociological andgovernment policy implications
Recipients of the $13 Billion inannual Canadian gambling revenue7%2%3% 1% Provincial GovernmentsPrivate Operators34%53%Charities & CommunityOrganizationsHorse RacingAssociationsFirst NationsFederal Government
Source of this gamingrevenue relevant to:• The legitimacy of government sponsoredgambling and the continued expansion ofgambling• The amount of money devoted toprevention and treatment
• There have been several studies that haveindirectly investigated this issue as part ofjurisdiction-wide prevalence studies• A consistent and important finding of thesestudies is that problem gamblers do in factcontribute a disproportionate amount oftotal gaming revenue
Proportion of Revenue fromProblem GamblersAustralia(Productivity Commission, 1999)4 U.S. states & 3 Canadianprovinces (Lesieur,, 1998)Canada(Williams & Wood, 2004)New Zealand(Abbott & Volberg, , 2000)United States(Gerstein et al., 1998)33%30%23%(each province weighted equally)32%(weighted by population)19%15%
• However, these proportions are inconsistentbetween studies and inconsistent withactual gaming revenues
Reported Expenditure/Actual Revenue6 U.S. States(Volberget al., 2001)New Zealand(Abbott & Volberg, , 1999)Canada(Williams & Wood, 2004)Australia(Productivity Commission, 1999)United States(Gerstein et al., 1999)4.5 pari-mutuel betting 2.4 lottery4.1 casino table games 1.1 EGM3.1 bingoRatio much higher than actual for lotteries~1.0 horse & dog bettingRatio much lower than actual for casinos & EGMs2.1 overall1.4 lotteriesRatio much lower than actual for wagering andEGMs0.3 lotteries0.0 casinos (reported winning $3 billion)0.0 racetracks (reported winning $2 billion)
Reasons for inconsistencies• Problems with determining the prevalence rate of problem gambling False positives with SOGS False negatives due to direct questioning Under sampling of problem gamblers in telephone surveys• Methodological problems in assessing self-reported expenditures Ambiguous question wording Question wording that biases the response Fallible memory Seasonality of gambling• Difficulties in tabulating revenues from jurisdictional residents Out of jurisdiction revenue and expenditures Not reporting all the revenue
Using improved methodologyto reinvestigate this issue inAlberta & Ontario• Better methods to obtain self-reported expenditures– prospective 4 week diaries of gambling expenditures– clear, non-biasing questions explaining what is meant by ‘net expenditure’• Better methods to establish true problem gambling prevalence rate– better instrument (CPGI)– more exhaustive RDD sampling to achieve a better response rate (ONT)(– adjustments for populations not available for sampling (ONT)• Exclusion of out-ofof-province expenditures as well as revenues fromnon-jurisdictional residents.
Both Studies• $50 incentive to keep 1 month diary of gambling expenditures• Each diary included detailed instructions and examples toclarify we are interested in NET expenditures• Reliability and validity questions incorporated into the surveyAlberta Study (AGRI funding)• Participants recruited though posters at casinos, racetracks,lounges, malls, and bingo halls throughout Alberta (March2002 – October 2003)• Total sample size of 211 (33 Non Problem; 64 Low Risk; 63Moderate Problem; 51 Severe Problem)Ontario Study (OPGRC funding)• Participants recruited though random digit dialing of 6654Ontario residents (March 2003 – November 2003)• Total sample size of 364 (156 Non Problem; 116 Low Risk; 60Moderate Problem; 32 Severe Problem)
Results• Prospective diaries offer good estimates of gamblingexpenditures based on their fairly close match withactual provincial revenues.• Alberta– Estimated 39% of Alberta gaming revenuederived from moderate and severe problemgamblers• Ontario– Estimated 35% of Ontario gaming revenuederived from moderate and severe problemgamblers
ConclusionsThe exact proportion depends on several things:• The specific jurisdiction• Jurisdictions differ in• Their availability of certain types of gambling (e.g., gaming machines)• The strength of their policy and educational initiatives to prevent ent problemgambling• The specific time period studied• Gambling availability and government policies change fairly rapidly.• Places that have had gambling available for a longer period of time tmayhave different rates of problem gambling compared to places that havemore recently introduced it.• How you define and measure problem gambling.• The present study has calculated the proportion of revenue derived ed frompeople who report being negatively impacted by their gambling (problemand pathological gamblers combined). However, other people mightconsider that the proportion derived from ‘addicts’ (i.e., pathologicalgamblers) to be the more relevant figure.
Public Policy Implications• In Canada, the money spent on prevention/treatment/research isvery small compared to the amount contributed by problemgamblers– Ontario derives 30-40% of its gambling revenue from problemgamblers and spends only 1.8% of net revenue onprevention/treatment/research (Canada West Foundation, 2005).– Alberta derives 30-40% of its gambling revenue from problemgamblers and spends only 0.4% of net revenue onprevention/treatment/research (Canada West Foundation, 2005).• It is questionable whether governments, whose mandate is toserve the people, should be in the business of owning andoperating an enterprise where a substantial proportion of therevenue comes from problem gamblers
Public Policy Implications• Governments that do undertake the ownership and operation ofgambling need to aggressively minimize both the prevalence ofproblem gambling and the amount of revenue derived from thispopulation.• Education/Awareness efforts are only one pillar of an effectiveharm minimization approach, and are never successful on theirown. What is required are concomitant policy initiatives, manyof which are easily implemented:– Requiring ID to enter a gaming facility (as is often done in Europeto bar self-excluders).excluders).– Making gaming facilities legally liable for not enforcing their ownself-exclusion exclusion contracts.– Eliminating credit.– Eliminating ATMs from gaming facilities.– Eliminating or severely restricting the most dangerous forms ofgambling (i.e., gaming machines).