What Are The Social Costs of Gambling? - European Association for ...

easg.org
  • No tags were found...

What Are The Social Costs of Gambling? - European Association for ...

What Are The SocialCosts of Gambling?Douglas M. Walker, Ph.D.College of CharlestonU.S.A.7 th European Conference on Gambling Studies and Policy IssuesNova Gorica, , Slovenia, July 1-4, 12008


Background• The social and economic effects of gamblingis one of the primary areas of gamblingresearch– Pathological/problem gambling diagnosis,prevention, and treatment is another• Conferences in Whistler, CA (2000) andBanff, CA (2006) were dedicated toeconomic/social effects– Particular focus on costsWalker, What Are the Social Costs? 2


Background, cont.• There have been 3 generalperspectives on social effects– Economic perspective• Welfare economics– Cost of Illness• Substance/alcohol abuse– Public health• Focus on cost minimizationWalker, What Are the Social Costs? 3


Different Perspectives• There are some similarities in the variousapproaches– Economic and cost-ofof-illness perspectives aresimilar in many ways – measure the same things– Most people are sympathetic to the public healthfocus on harm minimization• Differences still exist in terms of– Definition of “social cost”– How to deal with transfers, lost workproductivity, and other cost items– Monetary measurement of costs and benefitsWalker, What Are the Social Costs? 4


No Perfect Metodology• There does not appear to be a singlemethodology that will satisfyresearchers in all different disciplines• The focus on monetary estimates ofcosts and benefits is not warranted– Monetary estimates are mostly arbitrary– Reliance on such data is a recipe forpoorly informed policyWalker, What Are the Social Costs? 5


Examination of TwoImportant Perspectives• This paper focuses on a few critical aspects of twoapproaches to the social/economic costs ofgambling– The definition of social cost provided by Markandya andPearce (1989) is the foundation of cost-ofof-illness studiesthat have been adapted to gambling and some economicstudies– The new Canadian SEIG (2008) is hoping to be the newstandard for cost-benefit studies – at least in Canada• The “economic” perspective, of which I am aproponent, is not discussed in detail here– It has its own serious problemsWalker, What Are the Social Costs? 6


[1] Markandya andPearce (1989)• This paper serves as the foundation for avariety of social cost studies, including…– Australian Productivity Commission (1999)– Single et al., “International Guidelines forEstimating the Costs of Substance Abuse, 2e”(2003, see pp. 9-10) 9– Collins and Lapsley (2003, see pp. 125-127)127)– Subsequent studies?• Markandya and Pearce (1989) extendsanalysis by Atkinson and Meade (1974)Walker, What Are the Social Costs? 7


Markandya and Pearce(1989), cont.• For a cost to be “private” the actor musthave full knowledge about the potentialcosts of consuming the good.• For smoking (the subject of their paper), itimplies that if the consumer is not “fullyinformed” about the harm of smoking, heunderestimates the harms and chooses tosmoke too much.• The result is a social cost, even if the cost isborne by the smoker himself.Walker, What Are the Social Costs? 8


Markandya and Pearce(1989), cont.“To the extent that the costs are knowinglyand freely borne by the consumer orproducer himself, they are referred to asPRIVATE COSTS but to the extent that theyare not so borne but fall on the rest ofsociety, they are referred to as SOCIALCOSTS. Hence, the total cost of any activityis the sum of the private and social costs.”(p. 1139)Walker, What Are the Social Costs? 9


Markandya and Pearce(1989), cont.• “If his actions are determined by aperceived cost that is in fact less thanhis actual cost, then the differencebetween the two can be viewed as asocial cost.” (p. 1140)• Some costs have been ignored, andthe smoker over-consumesWalker, What Are the Social Costs? 10


Markandya and Pearce(1989), cont.• This definition of social cost, whether applied tosmoking or gambling, may be biased towardoverestimating costs– Viscusi and Hakes (2008) find that people tend tooverestimate potential risks from smoking• This suggests that they smoke less than is socially optimal– The Markandya and Pearce (1989) framework assumesthat people will only under-estimate, estimate, and not over-estimate risks• All decisions are uncertain, to an extent– No one perfectly estimates costs and benefits of an actionWalker, What Are the Social Costs? 11


Markandya and Pearce(1989), cont.• Further investigation should examinethe extent to which this aspect of thesocial cost definition contributes tomonetary estimates• This example shows how importantthe definition of “social cost” may beWalker, What Are the Social Costs? 12


[2] Socio-Economic Impactof Gambling (SEIG)Framework• This Canadian report was released inFebruary 2008• Funded by a number of Canadianresearch/government organizations• Primary author Mark Anielski• SEIG is promoted as being a possible gold-standard for social/economic research• Provides a detailed list of costs and benefitsto be measuredWalker, What Are the Social Costs? 13


SEIG, cont.• While a standardized methodologywould be useful, SEIG may not be thebest hope– But it is already being used as afoundation for several Canadian studies• At least 8 issues should be consideredbefore researchers adopt this as astandard tool for analysisWalker, What Are the Social Costs? 14


SEIG 1: Flexibility-Comparability Tradeoff• A primary benefit, according to its authors, is itsflexibility• Researchers in different jurisdictions may wish tofocus on different items to measure– SEIG lists many, many different items that may bemeasured– No single study could do all of them• Such flexibility implies limited comparability acrossjurisdictions, through time, and in evaluatingdifferent treatment tools• Researcher bias has been a serious problem inmonetary estimates of costs/benefitsWalker, What Are the Social Costs? 15


SEIG 2: Bias AgainstBenefits of Gambling• One serious limitation, for economists, isthat the SEIG ignores “consumer’s s surplus”(CS)– CS is the difference between what people arewilling to pay, and what they actually pay• Like profit to producers, except for consumers• CS is perhaps the largest benefit from legalizedgambling• The result of this is a likely under-estimateestimateof the benefits of gamblingWalker, What Are the Social Costs? 16


SEIG 3: MeasurementProblems• The SEIG authors are upfront about datameasurement problems in most of theindicators they identify– This is a very important positive aspect of SEIG• Many of the costs identified are simply un-measurable• Even if researchers developed ameasurement tool, it’s s not clear thatmonetary estimates are useful– This view may be shared by Eric Single, MarkKleiman, , and others (COI)Walker, What Are the Social Costs? 17


SEIG 4: Attribution ofCosts to Gambling• Over half of problem/pathological gamblershave comorbid disorders• SEIG authors acknowledge that it is notpossible to dole out measured costs inshares corresponding to causality• Without some reasonable mechanism, costestimates are not much more than arbitraryWalker, What Are the Social Costs? 18


SEIG 5: Means to an End?• Goals of SEIG are– Help show where investment should be made to reducepreventable negative impacts of gambling– Help to evaluate how well such investments are achievingtheir goals over time• These goals are very reasonable, and the SEIG canhelp in these regards• Yet, it is not clear that identification andmeasurement of all costs and benefits is necessaryto achieve the above goals• Researchers have come to some agreement abouttypes of potential harms of problem/pathologicalgambling behaviorsWalker, What Are the Social Costs? 19


SEIG 6: Reliance on GPI• The SEIG uses the Genuine Progress Index, whichis a relatively new alternative to basingmeasurement on GDP (gross domestic product)• Indeed, GPD has numerous problems– Ignores non-market activity– Many “bad” expenditures add to GDP• Medical expenses/treatment• Repairs from natural disasters– Many expenditures that add into GDP can hardly be seenas “good”,, even though they may be rational• GPI attempts to measure in a way that good thingsincrease the index, “bad” things do notWalker, What Are the Social Costs? 20


GPI, cont.• I agree that GDP is flawed, and I amsympathetic to the GPI principle• But GPI also has flaws…– Ignores CS, which is really a benefit toconsumers– Treats inequality as a cost– Includes wealth transfers in its computations– Subjective• As an economist, these things trouble me• The likely result is a bias against gamblingWalker, What Are the Social Costs? 21


SEIG 7: Advantages overOther Methodologies?• The SEIG report is very detailed about differentitems to be identified/measured• Suggests that it will help researchers come to anagreement• But the SEIG framework raises a whole variety ofnew issues to be debated among researchers• A flawed methodology that is well known may bepreferred to an unknown (and also flawed)methodology• What benefit will a shift to an entirely newmethodology provide?Walker, What Are the Social Costs? 22


SEIG 8: Funded ResearchMay Avoid Peer-Review• The SEIG report emphasizes the need foradditional research funding• It provides a long menu for potentialresearch projects• However, if such research is funded and isnot aimed for academic journals, the lack ofpeer review could be counter-productive• To truly become an accepted standard forsocial/economic research in gambling, suchstudies must be able to survive peer reviewWalker, What Are the Social Costs? 23


SEIG: Summary• Overall the SEIG report does highlight manyof the areas that have not yet beenexamined by researchers• It is important to bring attention to theseissues• But the usefulness of detailed research assuggested by the SEIG report has not beendemonstrated• Given the problems with accurate costattribution, monetary measurement, etc., itis not clear such research is worthwhileWalker, What Are the Social Costs? 24


The Economic Perspective• As an economist, I still lean toward aperspective based on welfare economics• However, there are controversies– transfers of wealth– lost work productivity– costs that are “internalized”– many of the negative effects of problemgambling that have been identified aren’tconsidered “social costs”• Has some of the same measurementproblems as other methodsWalker, What Are the Social Costs? 25


Conclusions• Specific monetary estimates for the costs ofproblem gambling may not be worthdeveloping• There exist major disagreements amongresearchers about the definition of costs/benefits, and how to measure them• Other limitations– Researcher bias (discipline-related; pro- or anti-gambling)– Data availability/qualityWalker, What Are the Social Costs? 26


Contact InformationDoug WalkerDept of Economics and FinanceCollege of Charleston5 Liberty StreetCharleston, SC 29401USAhttp://www.cofc.edu/~walkerdWalkerD@cofc.eduWalker, What Are the Social Costs? 27

More magazines by this user
Similar magazines