Daily Archives: January 1, 2013

The Hidden Costs of Gambling


Casino gamblingLet’s start 2013 with a sober consideration of the social and economic costs of gambling. Back n 2006, the Canadian Medical Association noted that,

“Provincial governments may be glossing over the societal and health costs of problem gambling, including depression and suicide, because of the significant income they gain from gambling, claim several public advocacy and mental health organizations.”

Glossing over is a polite way of saying “deceiving.” They’re hiding the facts from the public. The CMA called for a thorough and scientific study of the “relevance of depression and suicide among problem gamblers.”

“The normal system that provides checks and balances around this area is compromised because government in every province is responsible for alcohol and gaming regulation of the industry — and the welfare of those with gambling problems,” says Neasa Martin, a researcher at the Mood Disorders Society of Canada. “Their revenues are closely tied to the gambling industry, putting a pall on normal advocacy around the issue.”

This article has a parallel piece in the CMAJ that lists the three “elephants in the room” that have to be part of any discussion about gambling. These, the author states, are:

  1. …the inequitable distribution of the risks and benefits of gambling in our society;
  2. …treatment of problem gambling cannot undo the damage caused by lost wealth;
  3. …health promotion. It is time for governments and public health advocates to stop being seduced by the promise of anti-gambling campaigns and education that place the onus of self-control on the shoulders of the very individuals who have a serious disorder of impulse control.

New Zealand Mental Health
Gambling is, of course, voluntary taxation.* I have no moral issue over gambling; however, if people wish to gift the government their wages, they could just as easily mail the Minister of Finance a cheque every few weeks, and save us the contentious debate over gambling in our municipality. I do, however, believe that government dependence on gambling revenue is a fool’s economy. Any government – municipal and higher.

Back to the numbers

Let’s take a moment and consider some numbers before we continue. The Ontario Lottery Gaming Corporation (OLG)** has accepted the studies that show roughly 3.4% of the population as “moderate to serious” gambling problems. The Problem Gambling Association (PGIC) suggests a range between 1.2% and 3.4%, depending on which study you read and how you define “problem.” There are other studies that show the percentage in Ontario is as high as 4.8%.

The population of Ontario in 2011 was 12,851,821. At the low end of that range, there are 154,222 Ontarians with severe gambling problems. At the upper end, there are 436,962. A London Free Press article on gambling ups this to “almost 500,000″ without a source reference.

In other studies, the national average of problem gamblers (moderate is not mentioned) was estimated to be 2%. At 2%, there are still more than 257,000 problem gamblers in Ontario; but that media figure is lower than most experts estimate. In young gamblers aged 15 to 24 years, the percentage is much higher as this study found:

 “…the national prevalence of moderate-risk or problem gambling was 2.22% (3.30% in male respondents and 1.10% in female respondents). …Regional prevalence estimates of youth moderate-risk or problem gambling were… 2.75% in Ontario….”

This article repeats another warning from CAMH that, “Young people are twice as likely as their adult counterparts to develop serious gambling problems.”

 Younger Ontarians are also more likely to gamble online, as the Centre for Addiction and Mental Health (CAMH) reported in 2005: “…aged 18-34 were most likely to gamble money over the Internet compared to the older age groups.” And to capture more of that market, the OLG is launching a whole new internet gambling initiative in 2013.
Until recently, it was illegal for Ontarians to gamble online, but because the OLG now wants a slice of that pie, the laws got changed to accommodate them. Seems the MOH’s recommendations in 2005 were overlooked: “Prior to the adoption of any new technological innovations proposed by the gaming industry, consideration should be given to their potential impact on problem gambling…”.

This 2007 study suggests online gambling will create even more problem gamblers:

“…the inherent nature of Internet gambling would seem to make it conducive to increasing the rates of problem gambling. …The increasing patronage of online gambling sites will also increase the actual numbers of problem gamblers in the general population.”

The OLG contributes about $40 million a year to problem gambling treatment and services throughout the province, but it’s a small figure compared to its revenue:

“The OLG’s current contribution to education, prevention and treatment is less than 1% of its total revenues of $6.5 billion. There’s no indication this tokenism will be replaced with a serious investment in curbing the social and economic impacts of gambling.”

Based on the accepted number of problem gamblers in Ontario, that divides out to between a high of $259 each person per year, to a low of $91.54 each. There are 52 agencies treating gambling in the province right now. Dividing $40 million by 52 and you get less than $750,000 a year per centre.

If people wish to gift the government their wages, they could just as easily mail the Minister of Finance a cheque every few weeks.

So how much does problem gambling cost, and is $40 million sufficient to counter the social and economic costs?  Does the profit from gambling offset its cost? This is hugely complex and difficult to adequately measure, because are so many factors to take into account – medical, economic, social, legal, many of which are difficult to quantify, or may be intrinsically linked with other issues like substance abuse.

This PDF is a 24-page chapter of a larger US report on gambling, and it reviews numerous state studies. One such study that estimates that, in Wisconsin, there is a net benefit from gambling: “social costs represent about 42 percent of the economic gain…

The OLG net benefit to the Province of Ontario is approximately $2 billion. In its strategic plan (pgs 3-4), the OLG has stated it wants to increase that contribution by $1.3 billion, to more than $3 billion a year by “modernizing” gambling (making it more “accessible” by increasing the opportunities and locations so more Ontarians gamble). For all their high and mighty words about social responsibility, to me, it looks like simply an outline of a plan to grab more money from our pockets.

If our social costs are roughly the same as Wisconsin’s, then 42% of $2 billion is $840 million. Subtracting the $40 million currently paid to problem gambling issues, problem gambling costs the province $800 million every year. At $3.1 billion, it will cost us $1.26 billion, assuming the OLG does not increase its contribution to problem gambling programs. Gambling is, if the situations are similar, profitable for the province in this comparison, but not nearly as much as we are led to believe: there are high social and economic costs not being disclosed.

In response to this plan to expand gambling, in its Winter 202 newsletter, the PGIC wrote,

Without question, this will increase the incidence of gambling problems and associated negative impacts, including harm to families and the community.
The OLG plan means that, more than ever, Ontario needs an accessible, skilled, comprehensive and well-resourced problem gambling (PG) treatment system.

However, nothing in the OLG strategic plan indicates to me a significant increase either in funding or treatment system. At the very least that a 50% increase in revenue should be matched by a 50% increase in funding. Or is the net income the only measure of success? The Toronto Star  noted,

“…there is no indication that funding for education, addiction counselling and Trillium Foundation grants will be increased proportionately to the gambling universe that’s being created. It’s already inadequate.”

Back in 2005, a report to the Ministry of Health on gambling in Ontario noted (emphasis added):

“…if a culture of responsibility is fostered in the decision making process, then decisions will be made taking into account the impact of any proposal on the treatment of problem gamblers, the provision of information to gamblers and the public on problem gambling and responsible gaming,the research that should precede the decision and whether an element of consumer protection is involved. Those involved in each component of the Strategy should have a role to play in contributing to initiatives that are designed to advance the minimization of potential harm from gambling activities. A culture of responsibility reaches out to other Ministries as well whose programs impact on problem-gambling issues.”

The OLG and the rest of the Ontario government seems to have overlooked that in the new strategic plan. They just want the money. Responsibility? No our job.

A study in Florida, described in that US report, used costs-per-problem-gambler “…calculated by Volberg (1994) of $13,600 on average per pathological or problem gambler.” That study looked into collateral costs such as prisons and increased pressure on the justice system to arrive at that figure. If Ontario is similar per-capita, the cost of problem gambling ranges between $2.1 and $5.9 billion a year. In this model, there is no net benefit from gambling revenues, just losses.

This 2007 study showed that the number of people with moderate to severe gambling problems is much higher than previously estimated. This is based on a sample size of 6,654 Ontarians. It found,

“…3.74% were moderate problem gamblers (CPGI = 3-7); and .99% were severe problem gamblers (CPGI = 8+), with an overall of prevalence of 4.73% for moderate and severe problem gamblers combined… this adjusted Ontario prevalence rate of 4.76% is significantly higher than two

previous Ontario prevalence studies: 3.8% obtained by Wiebe, Single, & Falkowski-Ham in 2001 and 2.0% obtained by the Canadian Community Household Survey (CCHS 1.2) for Ontario in 2002 (Statistics Canada, 2002).”

Sources of Revenue

The study also concluded that the OLG gets more than a third of its revenue from problem gamblers, but also that gambling machines (slots) take a higher proportion of that:

“Both the winsorized and losses-only data suggest that this proportion is approximately 36%… This evidence indicates that gambling machines and horse racing derive a much larger portion of their revenue from problem gamblers, in comparison to other forms of gambling. In rank order, the rough proportions are: 61% gambling machines; 45% horse racing; 32% casino table games; 22% bingo and raffles; and 18% lotteries, instant win, and Sports Select.”

Similar studies in the USA have reported this as well. Salon noted in a recent article:

“A significant portion of gambling revenues — one-third to one-half — is derived from problem gamblers, says Grinols, who, in a 2006 Review of Economics and Statistics article concluded that 8 percent of crime in casino counties can be attributed to the presence of legal gambling.”

The Canadian study recognizes that estimates of income are problematic because in their survey, only a small number of people who identified themselves as “severe” gamblers responded to the questions about gambling expenditures:

“Expenditures from the prospective diaries of 364 individuals tentatively indicates that about 36% of Ontario gambling revenue is derived from moderate and severe problem gamblers.”***

What I found most compelling about this figure is in the policy implications section near the end of the report (emphasis added):

“Thirty-six percent would be a problematic figure for private industry, but is especially problematic for a government-run operation, when the purpose of government is to serve the people, not to exploit the people.

Ontario spends 13 times more money advertising and promoting gambling as they do on prevention and treatment (Williams, 2006). Furthermore, the $36 million put into gambling prevention, treatment and research in 2003/2004 only represents 2.6% of the $1.41 billion dollars estimated to have derived from problem gamblers in that time period. It is also far from clear whether gambling revenues represent true economic gain. Gambling revenues largely come from a transfer of wealth,rather than creation of wealth (e.g., Grinols, 2004). Furthermore, this is not an innocuous transfer, as it harms a significant minority of people (problem gamblers) in the process, and it tends to generate its revenue through the cannibalization or crowding-out of other (privately owned) 15 entertainment industries (e.g., Grinols, 2004).

There is another factor to consider when trying to estimate the economic impact of gambling: productivity loss. An Australian study found gambling was, overall, economically positive (for the government) but it recognized that it cost the Australian economy. Based on the loss of one hour per week per problem gambler:

“The study was able to “cost out” a number of factors associated with problem gambling. The effects of gambling on employment, consisting of job change costs, unemployment, and productivity loss, were estimated at A$27.8 million annually. The largest component of this estimate was productivity loss, accounting for almost A$20 million, followed by A$5.2 million for job change and A$2.7 million for unemployment.”

I’m not sure how they measure the economic value of lost productivity in this study. One simplistic way is to simply multiply the average hourly wage times the number of problem gamblers for the losses per week, then times 50 (allowing a two-week holiday per person) for the annual impact. If one hour per week is lost in Ontario, then based on 257,000 problem gamblers working at minimum wage ($10.25/hour), the resulting loss in productivity is about $132 million per year. If we use the upper end estimate (437,000 problem gamblers), and an average $15 an hour wage, the resulting loss is almost $328 million.

And what about suicide? The CMAJ article notes that, “problem gambling as a contributor to suicide is difficult to measure,” however, it does report an Alberta study that indicated gambling is a key factor in roughly 10% of suicides:

“In Alberta, gambling was listed “in the files” of suicides about 10% of the time; 46 out of a total of 482 suicides in 2001, and 54 out of 430 suicides in 2000.”

Can we quantify the impact of suicide? Not really. The emotional impact, however, is devastating to family and friends. Ontario has been been tracking gambling-related suicides since 1998 – there are on average five a year. The number may be higher because the method of identifying them is solely through a note left behind. The Globe & Mail reported 13 in 2007 and noted the number was rising, in part encouraged by casino tactics:

“A Globe and Mail investigation last fall revealed government-owned casinos are spending hundreds of millions of dollars on freebies – trips, dinners, theatre tickets – that keep gamblers coming back. Gambling debts have led to bankruptcies and even suicides.

Although there are no countrywide statistics, Canada Safety Council estimates 200 problem gamblers kill themselves every year. The Globe and Mail contacted each province to track the numbers of gambling-related suicides, finding only 50 such deaths were recorded in 2007 among the seven provinces reporting.”

Problem Gambling Institute of Ontario has a short FAQ on the potential impact of expanded gambling with many links to studies on gambling. In one called the Social and Economic Impact of Gambling, which tallies an objective view of both positive and negative effects, the authors write,

“One of the main negative impacts of gambling introduction is an increase in problem gambling and its related indices (e.g., bankruptcy, divorce, suicide, treatment numbers). The bulk of the impacts tend to be social/nonmonetary in nature because only the minority of problem gamblers seek or receive treatment, and only a minority typically have police/child welfare/employment involvement. Most of the increase in problem gambling occurs after the initial introduction of gambling, with progressively less impacts on problem gambling occurring with extended exposure… Research confirms that lower income people consistently contribute proportionally more of their income to gambling than do middle and high income groups (‘socioeconomic inequality’)…”

The report concludes that, gambling’s impact is a mix of good and bad, but that overall the good is limited, and not enough to fully offset the bad:

“…in most jurisdictions, in most time periods, the impacts of gambling tend to be mixed, with a range of mild positive economic impacts offset by a range of mild to moderate negative social impacts.”

My point here in this rambling (and incomplete) survey of the literature, is twofold:

  1. The true costs of gambling are not being disclosed by the government or the OLG, while the benefits are being exaggerated;
  2. Problem gamblers are contributing a significantly large portion to the OLG revenue, but also to the economic and social expenses caused by their gambling.

I’ll have more to post in the future, as we approach the debate at the council table (again). In a future post, I want to tackle the question of pork-barrel politics: what turns politicians into shills advocates for gambling? (It’s a story well, but partially, documented in the book, Betting the House, by Brian Hutchinson; it needs to be brought up to date). Does all the support for gambling and OLG’s planned expansion pass the smell test?


* The Fraser Institute, a right-wing think tank, framed the debate in 2002, ideologically: as one of “Individual freedom versus government paternalism.” It concludes as one might expect of conservatives, that social and economic impacts on individuals, families, or even communities, are not the concern of government:

“…the small number of people who are unable to control their gambling does not merit heavy-handed government intervention. Although there are socially harmful activities that require government intervention, gambling should not be considered one of those activities. Furthermore, those who become addicted to gambling rarely threaten the overall harmony of the community …”

The report concludes with a bit of Old Testament thunder:

“…intruding on gamblers’ liberties, prohibition makes a mockery of individual responsibility, which is hardly the best way to sustain the nation’s moral health.”

Ayn Rand must be smiling in her grave at that line.

** Gambling advocates like the OLG often use carefully chosen words to present gambling as entertainment. They call slot warehouses “gaming facilities” as if they were on a par with sports facilities. They don’t call them “gambling facilities.” Terms like “slot barns” annoy them. The Toronto Star calls them “…those windowless structures on the outskirts of cities, surrounded by huge parking lots.

*** The full disclaimer and explanation of the tentative results reads (emphasis added):
Limitations of these Findings
Regular gamblers occasionally have very large wins and losses. These statistical outliers have a major influence on the averages, making it very difficult with small sample sizes to establish what the “true” average expenditures are, so as to compare them with actual revenues.
Gambling is glamourousRealistically, there would have to be thousands of people completing prospective diaries from each of the four categories of gamblers to offset the impact of these outliers. The present study compensated for this by using winsorized data and data sets that eliminated winners. This is a reasonable but not perfect solution to this problem.
The proportion of revenue from severe problem gamblers is very tentative because of the small number of severe problem gamblers completing prospective diaries (n = 32). There is more certainty in the proportion derived from moderate and severe problem gamblers combined (n = 92). Similarly, the proportion of revenue derived from problem gamblers for particular forms of gambling is also tentative; not all problem gamblers participate in all forms of gambling and so some of these estimates are based on small sample sizes. It seems certain that gambling machines derive more revenue from problem gamblers then other forms of gambling. However, the actual portion for each form of gambling is less certain.
There is not a perfect match between reported expenditure and actual revenue for the prospective diaries. The total winsorized expenditures are 36% below actual revenues, and the losses-only total is 37% higher than actual revenues. This makes some sense considering that the largest expenditures have been winsorized in the former and all wins have been eliminated in the latter. On the other hand, it is also important to realize that the present study found gambling expenditure exaggeration and minimization to be equally common for all four types of gamblers, as evidenced by the uniformly low correlations between retrospective estimates and subsequent prospective diary amounts.
The implication here is that if there is an over or underestimate of expenditures relative to revenues, it probably does not affect the proportion derived from problem gamblers because of equivalent exaggeration/minimization in each group.

A separate, Canada-wide, 2004 study on the percentage of revenue from problem gamblers is here. It notes:

“… the most meaningful figure is the proportion of revenue derived from problem gamblers averaged across all jurisdictions: 23.1%. …problem gamblers report a proportion of expenditure that is more than five times their proportion among the Canadian population.”

Another study on Ontario demographics for gambling found:

“…converging lines of evidence indicating that a substantial portion of gaming revenue derives from people who are negatively impacted by their involvement in this activity.”