The theme of the 2006 Institute for Research on Pathological Gambling and Related Disorders annual conference on gambling addiction was Lost in Translation? The Challenge of Turning Good Research into Best Practice. During the next few weeks, The BASIS is pleased to present a series of editorials from some of the faculty members of that conference. In this week’s editorial, Douglas M. Walker discusses roadblocks to effectively estimating the social costs and benefits of gambling.
Douglas M. Walker
Associate Professor of Economics Georgia College
Faculty, The 2006 Institute for Research on Pathological Gambling and Related Disorders annual conference on gambling and addiction
During 2006 there were two majors conferences dedicated to the economic and social costs and benefits of legalized gambling. One was in Banff, Canada in April. The other was the NCRG/Divison on Addictions conference in Las Vegas in November. The fact that economic aspects of problem gambling are the focus of conferences is an indication of the importance and controversy of these issues. When state governments are considering the legalization of casinos, they often view the potential tax revenues as the primary benefit. This ignores the benefits to consumers who enjoy gambling. In any case, the benefits of gambling are often overshadowed by the potential social costs. Some individuals develop serious problems resulting from their gambling behavior. Politicians must consider this side of the equation because to the extent the social costs offset the tax revenues or other benefits, the political argument in favor of casino legalization becomes weaker. A monetary estimate of the magnitude of benefits and costs provides politicians and voters with concise and (seemingly) simple information about the effects of gambling. For example, politicians may vote to legalize gambling as long as the benefits outweigh the costs. This would be the right course of action from an economic perspective. The problem is that estimating the costs and benefits due to gambling is extremely complicated. In my new book, The Economics of Casino Gambling (Walker 2007), I discuss many of these measurement issues. In this article I will summarize some of the most serious roadblocks to effectively estimating the social costs and benefits of gambling.
These problems are so serious that readers should be very skeptical of any empirical study that does not address each of these issues. We will focus specifically on costs since these are much more often the focus in the literature.
The first problem with estimating the social costs of gambling is that most authors either fail to define what they are measuring, or they improperly define effects of gambling as “social costs.” Researchers from different disciplines and even within particular disciplines disagree on the proper way to define social cost. Obviously, this makes it unlikely that any particular cost estimate would be uncontroversial.
Even if we can clear this first hurdle and we can agree on a definition of social cost, there are a number of other difficulties in actually measuring the costs. Perhaps the most important is comorbidity. That is, pathological gamblers may have other problems that contribute to their socially costly behavior. A recent study (Petry, Stinson and Grant 2005) found that almost 75% of pathological gamblers also have alcohol use disorders and almost 40% have drug problems. Consider a problem gambler who is also an alcoholic. Suppose his behavior results in social costs of $1,000. Most gambling researchers will simply attribute the entire $1,000 cost to gambling, even though the drinking may be responsible for some (or even most) of the total cost. How should comorbidity be handled, in terms of estimating the costs of a particular affliction? This question has not been answered by researchers.
A second issue is the counterfactual scenario. What if casinos were not legal? Would pathological gambling and the associated social costs disappear? Probably not. The correct estimate of the costs of pathological gambling is not the total cost of pathological gambling behaviors. Rather, the relevant cost is the difference between the costs when casinos are legal and when they are not. Unfortunately, it is very difficult to know with accuracy the counterfactual scenario.
A third problem with estimating the social costs of gambling is that many of the published estimates have been based on unreliable survey data. In some studies authors have based their cost estimates on diagnostic tools like the DSM-IV or SOGS. Some papers use original surveys in which problem gamblers are asked about the extent of their gambling losses or the sources of their money used for gambling. A recent study (Blaszczynski et al. 2006) found that many survey respondents are unable to estimate their gambling losses, even if they are given instructions on how to do so. This evidence suggests that it would be difficult for the same individuals to reliably report the source of their gambling losses. This is because budgets are fungible (Walker 2007, p. 121). Yes, a person may gamble too much. But she may also have a very high car payment. How confident are we that this person (or the researcher, for that matter) could accurately identify what source of income—paycheck, bank loan, cash gifts, theft, etc.—was used to finance specific expenditures? A person taking a survey on problem gambling may be predisposed to blame all their problems on gambling even when there are other problems present.
The fourth problem I would like to mention relates to how government expenditures are handled. A large portion of the social costs of gambling may be related to government expenditures. For example, suppose government provided treatment is available, and many pathological gamblers commit crimes which create legal costs. Most social cost estimates simply take the value of these government expenditures and call them “social costs.” It would seem obvious that since government spending requires taxes that these expenditures should be considered social costs. Indeed, most people would agree that lower spending on these sorts of things would be preferred to higher spending. But the same is not necessarily true of, say education. People often vote for more public education spending. Government expenditures are not equivalent to social costs. If they were, then we could reduce the social costs of gambling by simply reducing spending on gambling-related problems. Unfortunately, this does not leave us with a clear and appropriate way to classify gambling related government expenditures. Yes, such expenditures may be a reflection of social costs, but they may also represent social costs of our policy decisions. This issue was pointed out by Browning Terms of Use (1999). As with the previous issues, there is no ideal way to deal with this one effectively. Finally, since the above problems (among others) make it very difficult to obtain credible data on the social costs of gambling, many researchers rely on a variety of wildly arbitrary assumptions in performing their analyses. The result is sometimes completely meaningless cost estimates. There are other problems in the gambling literature that make cost-benefit analyses unreliable. This brief article discusses what I see as the most serious ones. Until these research problems can be resolved, I am afraid that the quality of the economics of gambling research will remain deficient. In many ways, the problem gambling literature parallels the substance abuse literature. That work provides a possible path for gambling researchers to follow. But even the better-established substance abuse literature has its critics. Although the political debate over casino gambling often revolves around jobs, tax revenues, economic revitalization, pathological gambling, crime, etc., I do not expect a significant short-term improvement in the quality of research, at least on the economic side. Perhaps rather than focusing on questionable monetary estimates of costs and benefits of gambling policymakers and voters should put more emphasis on fundamental issues like freedom of choice, personal responsibility, and the role of government in a free society.
What do you think? Click here and let us know! Comments on this article can be addressed to Douglas M. Walker.
The views expressed in the Op-Ed/Editorials page are solely the views of the author(s) and do not necessarily reflect the views of the BASIS, its sponsors, or affiliated organizations.
References
Blaszczynski A, R Ladouceur, A Goulet, and C Savard. 2006. “How much do you spend gambling?”: Ambiguities in questionnaire items assessing expenditure. International Gambling Studies 6(2): 123-128.
Browning E. 1999. The myth of fiscal externalities. Public Finance Review 27: 3-18.
Petry N, F Stinson, and B Grant. 2005. Comorbidity of DSM-IV pathological gambling and other psychiatric disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Clinical Psychiatry 66(5): 564-74.
Walker D. 2007. The economics of casino gambling. New York, NY: Springer.