Critics: Problem gambling analysis flawed

By Liz Benston

Tuesday, March 25, 2003 | 10:59 a.m.

Nevada's casino industry has issued a rebuttal to a controversial study by two University of Nevada, Las Vegas professors who generated figures on what addicted gamblers in Clark County cost society.

Last month's study by UNLV professors Keith Schwer and Bill Thompson estimated that addicted gamblers in Southern Nevada cost the community from about $300 million to $470 million each year and about $8,000 per problem gambler each year. The methodology of the study sparked instant criticism from casino insiders who suggested that the data could be used as a tool for anti-gambling crusaders or as an excuse to hike casino taxes.

The study also emerged amid a renewed Legislative effort to earmark state funds for problem gambling treatment programs.

In a report prepared last week for the Nevada Resort Association -- the chief lobbying group for Nevada casinos -- Georgia College & State University Assistant Professor of Economics Douglas Walker said the results of the UNLV social costs study are "unreliable because their analysis is seriously flawed."

The study generates a cost estimate that uses an inappropriate characterization of gambling's social cost and relies on "numerous arbitrary assumptions," the report said.

Under Walker's revised estimates, the UNLV study's estimated social costs would be dramatically reduced to about $12 million for problem gamblers as a whole per year and to $881 per problem gambler per year.

In a 1999 article in the Journal of Gambling Studies, an academic publication, Walker published a similar critique of a social costs study authored by Thompson in 1996 of problem gamblers in Wisconsin. Walker also expects to publish an article in the journal's summer issue criticizing another Thompson study on the social costs of problem gamblers in South Carolina.

Walker will be paid up to $2,000 by the Nevada Resort Association for the report.

The study includes the effects of welfare, food stamps, theft and bad debts -- all elements that indicate transfers of wealth from one person to another rather than economic costs to society, Walker said. Other included costs, such as work missed and lost output, are costs that are internalized by third parties such as private companies and also shouldn't be part of a true social cost estimate, he said.

The findings, based on surveys filled out by about 100 participants in Gamblers Anonymous treatment programs, included costs such as bankruptcies, thefts, lost productivity at work, unemployment benefits and welfare support that occur as a result of gambling addictions.

"Adding transfers and internalized costs to social costs produces a gross overestimate of the social costs of problem and pathological gambling," he wrote. Eliminating such costs yields a revised figure of $32.4 million to $50.5 million per year -- much lower than the $300 million estimate, he said.

"You need some way of comparing (social costs) to other states," Walker said in an interview with the Sun. "Standardized economics provides some definition of that."

The study also extrapolates cost estimates to the general population by using a screen that has been dismissed by the research community, Walker said.

The screen -- which other researchers have used in surveys to detect problem gambling activity -- has since been replaced by another screen that some consider more reliable, he said. Under the newer screen, total estimated social costs would be further reduced to $12.2 million per year.

The study also doesn't address the fact that many pathological gamblers also suffer from other addictions, he said.

Problem gambling "may be independent of other afflictions or it may only be a reflection of other problems," he said. "Then if a person has addictions other than gambling, it is inappropriate to attribute all the social costs of his antisocial behavior to the gambling problem."

Thompson said he cast a wider net by including costs paid by people who aren't gambling as the "price tag gamblers impose upon non-gamblers."

"We do not call it economic costs. We call it social costs because these are all costs to society."

Thompson said he low-balled his cost estimates and also left out others that would have been too difficult to evaluate, such as cost of attempted suicides, the cost of divorce beyond court costs and emotional distress costs.

Up to 30 percent of the sample said they had attempted suicide -- about 10 times the population average, he said. Real suicides that researchers don't know about also have costs, he said.

Also, the rate of pathological gambling activity in the general population is based on a range of 1.9 percent to 3.5 percent -- with the low end double the rate found by the congressionally-appointed National Gambling Impact Study Commission in 1999 and the upper end determined by researcher Rachel Volberg in a Nevada study released a year ago, Thompson said.

Volberg was commissioned by the state to conduct the first comprehensive study of gambling addiction in Nevada.

In the study, a median of 6.4 percent of Nevadans were identified as either pathological gamblers or "problem gamblers," a less severe form of gambling addiction. Using the newer screen Walker advocates, the extent of problem and pathological gambling comes in at 2.1 percent.

Thompson said he expects to present a future report in the coming weeks that will further explore the issue of co-existing addictions by separating estimated social costs for problem gamblers without other addictions from costs attributed to those with other problems.

Social cost studies are controversial because they attempt to assign values to human suffering and also suggest the industry is to blame, observers say.