HomeMy WebLinkAbout2004 02 23 Public Hearings 406
COMMISSION AGENDA.
ITEM 406
Consent
Informational
Public Hearing X
Regular
February 23. 2004
Meeting
Mgr. /
Att. /
Dept.
REQUEST:
The City Attorney requests that the City Commission adopt on Second and Final Reading Ordinance
2004-03 which shall regulate and prohibit slot machines and devices including casino-type slot
machines.
PURPOSE:
The Purpose of Ordinance 2004-03 is to amend the City Code to regulate and prohibit slot machines
or devices that are unlawful under Chapter 849, Florida Statutes or that are defined in this Ordinance.
The purpose ofthis Ordinance is also to adopt stricter regulations of casino-type slot machines than
as provided for in Chapter 849, Florida Statutes.
APPLICABLE LAW AND PUBLIC POLICY
1. Florida Municipal Home Rule Powers Act
2. Chapter 849, Florida Statutes
Page 1 of 3
CONSIDERATIONS:
1. The City of Sanford recently adopted an ordinance regulating adult arcades.
2. Several local cities adopted temporary moratoriums on adult arcades until regulatory
ordinances can be adopted.
3. The City Attorney has provided a white paper to the City Commission explaining the issues.
4. The studies referenced in this report indicate that gambling can create adverse socio and
economic effects on a community, including crime, lost productivity, negative effects on
education, and increased societal costs such as welfare.
ST AFF RECOMMENDATION:
The City Attorney recommends that Ordinance No. 2004-03 be adopted on Second and Final
Reading.
ATTACHMENT:
1. Index to Studies Relating to the Effects of Gambling. Copies of two (2) reports are attached
for a summary review and copies of all the studies indexed are on file with the City Clerk's
office.
2. Ordinance 2004-03.
3. City Attorney white paper, dated November 26, 2003.
COMMISSION ACTION:
On December 8, 2003, the City Commission directed that an ordinance be prepared to regulate
casino-like commercial activities.
On February 9,2004, the City Commission approved Ordinance 2004-03 on First Reading.
Page 2 of 3
INDEX TO STUDIES RELATING TO
THE EFFECTS OF GAMBLING
(On file in the City Clerk's Office)
1. Materials presented by Schwer, Thompson, Nakamuro at the 2003 Annual Meeting
of the Far West and American Popular Culture Association, Beyond the Limits of
Recreation: Social Costs of Gambling in Southern Nevada, Las Vegas, NV,
February 1,2003.
2. Florida Executive Office of the Governor, Casinos in Florida: An Analysis of the
Economic and Social Impacts, Florida Office of Planning and Budgeting (1994).
3. Grinols & Mustard, Business Profitability versus Social Profitability: Evaluating
Industries with Externalities, the Case of Casinos, 22 Managerial and Decision
Economics 143 - 162 (2001).
4. Kindt, Excerpt from The Business-Economic Impacts of Licensed Casino Gambling
in West Virginia: Short-Term Gain but Long- Term Pain, originally published in The
National Impact of Casino Gambling Proliferation: Hearing Before the House
Committee on Small Business, 103 rd. Congress, 2nd Session 787 (1994).
5. Volberg, The Lewin Group, Christiansen/Cummings Associates, Gambling Impact
and Behavior Study, Report to the National Gambling Impact Study Commission
(April 1, 1999).
6. Grinols, CUTTING THE CARDS AND CRAPS: Right Thinking About Gambling
Economics, December 21,2001.
G:\Docs\City of Winter Springs\Agenda\Adult_Arcades_ Ordinance2.wpd
Page 3 of 3
Casino Watch-Florida Casinos Report
Page 1 of3
Economic and Social Costs
of Gambling
rl{']i'i(;TIj6.If:~tIH'J. IZWfJJ
r.!..u:tj
rl[q!~l~
Back to previolls l1a~e
I.lome
Casinos in Florida
An analysis of the Economic and Social Impacts
Prepared by:
The Executive Office of the Governor
Office of Planning and Budgeting
The Capitol
Tallahassee, FL 32399-0001
EXECUTIVE SUMMARY
3. Recurring sales tax revenues would experience a net decrease of at least
$84.7 million as Floridians diverted some of their existing taxable spending to
casinos.
4. Crime and social costs attributable to casinos would total at least $2.16 billion
annually.
PROBLEM GAMBLING AND RELATED CRIMES
According to psychiatrist Richard J. Rosenthal, most gambling addicts "are seeking 'action', an
aroused, euphoric state comparable to a 'high' derived from cocaine and other
drugs" (Worsnop, page 250). Counselors Arnie and Sheila Wexler of New Jersey state "most
dual-addicted cocaine addicVcompulsive gamblers will tell you gamblling gives them a bigger
high. Some drug addicts, who are also gambling addicts, will sell their drugs for gambling
money." (Worsnop, page 250).
Several studies have been conducted in an attempt to quantify the costs of problem gambling
behaviors. In 1994, Rachel Vol berg reported that the average individual pathological gambler
cost the public $13,600 each year (in 1981 dollars). This includes income that would have been
earned for those who lost their jobs, costs of prosecuting and incarcerating individuals for
crimes caused by their gamblilng behavior, and bailout costs, such as family gifts. Other
problems include lost job productivity, impaired judgment at work, lost productivity of spouses,
divorces, unemployment compensation, depression, physical illness related to stress, and
suicide.
A Marylalnd Department of Health and Mental Hygiene task force determined that its 52,000
adult gambling addicts cost citizens $1,5 billion in lost work productivity, monies stolen and
embezzled, bad checks and unpaid taxes (Worsnop, 1990, page 644), The cost per individual
compulsive gambler exceeds $28,846 and significantly increases when related costs for social
services, health care, bankruptcies, legal and correctional fees are considered.
The American Insurance Institute estimates that 40 percent of all white collar crime has its roots
in both legal and illegal gambling. Problem gamblers are responsible for an estimated $1.3
billion worth of insurance-related fraud per year. Insurance companies paid fraud victims an
average of $65,000 (Lesieur, page 45).
The mean gambling debt of people in compulsive gambling therapy ranged from $53,000 to
~
Casino Watch-Florida Casinos Report
$92,000, not including debts paid.
As many as 10 to 17 people may be innocent victims of each compulsvie gambler (Lesieur,
1984), including spouses, children, parents, other relatives, employers, co-workers and friends.
Two out of three compulsive gamblers will commit illegal activities in order to pay gambling
related debts and to continue gambling.
One out of every four problem gamblers have been involved in an auto accident during the
worst of their gambling. Almost half of those surveyed were speeding on their way to gamble or
on the drive back (Florida Council on Compulsive Gambling).
New Jersey residents also report that 28 percent know of someone who gambles too much
(Gallup, page 4).
Poor and working people spend a disproportionate share of their incomes on gambling
(Goodman, page 8). As a result, state gambling revenues come disproportionately from lower
income residents, causing a regressive form of taxation, Problem gambling behaviors are
highest among the poor and minorities (Goodman, page 17).
YOUTH
A 1985 random sample of 332 students at an Atlantic City high school found that 64 percent of
the students had gambled illegally at local casinos. Over 40 percent had gambled in casinos
before the age of 14.
COST OF PROBLEM GAMBLING BEHAVIORS IN FLORIDA
1, As noted by Lesieur, two out of three problem gamblers will commit a crime in
order to support illegal gambling activities. This means that 266,667 individuals
in Florida will commit crimes such as burglary, larceny, theft, forgery or fraud,
While not all of the individuals will be convicted, they will likely continue to
commit crimes. Some will be convicted more than once,
2. The average sentence length of a person convicted of burglary in Florida is
5.7 years; theft, forgery and fraud 4,2 years; and robbery 8.8 years. First time
offenders and non-violent offenders will likely serve probation or community
control rather than prison time.
3. Of these categories, 41,3% will likely be convicted of burglary; 36,8% for theft,
forgery or fraud; and 21.9% for robbery based on current incarceration rates,
Not counting costs of prosecution, restitution or other related costs, incarceration and
supervision costs alone for problem gambler criminal incidents could cost Florida residents
$6.08 billion.
Florida's problem gamblers related to casinos would require $1.66 billion in prison construction
costs.
Deducting these positive social effects from the estimated social costs results in a high,
medium and low estimate of $3,25 billion, $2.65 billion, and $2.16 billion in net social costs,
respectively.
OTHER CRIME
While many believe that legalizing gambling activities will decrease illegal gambling, an
examination of dollars gambled does not support this belief. Dr. Vicki Abt, a sociologist at
Pennsylvania State University, says legalization "has not decreased the dollar amount of illegal
gambling; what is has done is decrease (illegal gambling) relative to legal gambling. Whereas
75 percent or so was wagered illegally before the 1960s, now about 75 percent is wagered
legally. But the total amount went up in the meantime," (Worsnop, page 643).
ATLANTIC CITY
Page 2 of3
Casino Watch-Florida Casinos Report
Page 3 of3
In just three years following the opening of its first casino, Atlantic City went from 50th in the
nation in per capita crime to first (Goodman, page 58). Crime spilled over into neighboring
communities that received no measurable economic benefits from casinos. Atlantic City's crime
rate rose 230 percent during the first three years of casino operations, over 25 times the growth
rate of 9 percent for the rest of the state (FDLE, 1994, page 3). It should be noted that Atlantic
City's population decreased by 20 percent during the same time period. Florida's crime rate
rose by 30 percent during these same years,
NEW ORLEANS
New Orleans estimates that crime-related costs will be just under $5 million for a single casino.
Widespread gambling could add an additional 10,000 new crimes at a cost to the city of an
additional $14.1 million.
COLORADO
Central City, Colorado has increased its police force from two full-time and one part-time officer
to 16 full-time officers in 1994 (Worsnop, page 256) to keep pace with its "skyrocketing
incidents" of disorderly conduct, assaults and DUls.
CONCLUSION
Since state tax revenues (adjusted for pari-mutuel and Lottery revenue losses) are estimated to
range between $155 million and $276 million while annual crime and social costs are estimated
at a minimum of $2.16 million annually (and up to $3.8 billion), it appears that casino costs
significantly outweigh the benefits of legalization.
According to University of Nevada professor of Public Administration William Thompson,
casinos generate significant new tax revenues "only if they can export their product. The local
economy doesn6t benefit if only local people are gambling."
Back to pre\'iolls page'
HOl11e
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 1 of 8
~
1::"
,"'"
gI
" ~~j1r;~~~p' "{::i~~~~~:~~')~~~:~~
E A S Y i\l () :\ J~ y! ,.' _.f->-, '-~?~~~,;;~ FRONTUNE
I,..,,,',. j"i" II". lli"'U."jUlI \,b,,( al''': th.;;;;.;,: ~I'" a;..I ";'rl-"i;lto:~',i,:.:J" ti"lt.jil"> ,;;,,:i. &. ~i;jb' , '
. ,t I I I '
,!:!:i
.J::;
"""
'~
*
.~~...'\, ..< ',f'.... ...((, .".~._~~. r .;;. '~l ~""~<""' ~ .I'."-j, ,'~~."""
'.' :1 ' ". '. i' '..' , ". I. .! f '
"'".. I i \, ',. ,I" ',,-
\.:CAMJ3.,ThING:)'FACTS, &,:STATS
An excerpt from:
,~
7=.:
e::;.
.J::;
,ioJ:
ra
~{;
The Business-Economic Impacts of Licensed Casino Gambling in
West Virginia: Short-Term Gain but Long-Term Pain
By John Warren Kindt
~~
fi
""
{~
I?:.
John Warren Kindt is a professor at the University of Illinois at
Urbana-Champaign. A faculty member since 1978, he teaches
courses in commerce and legal policy. The text and footnotes
have been updated from the original publication in The National
Impact of Casino Gambling Proliferation: Hearing Before the
House Committee on Small Business, I03rd, Congress, 2nd.
Session 77 (1994). Due to the fonnat of this publication,
substantial footnotes have been deleted.
~s
e-~k'
ch
;=:!:
~
.1'....
...1::;:
~l
c....
~
0;1.-
f::
;.~
E:.
-'.1
'<2;
:::i
The Economic Development Argument Exposed
,;;:..
f:'"'
.~
-'-
From a business-economic perspective, the main issue involved
in legalizing various fonns of gambling is whether gambling
activities constitute a valid strategy for economic development.
While the dollars invested in various legalized gambling projects
and the jobs initially created are evident, the industry has been
criticized for inflating the positive economic impacts and
trivializing or ignoring the negative impacts (Goodman 1994).
The industry's tendency to focus on specialized factors provides a
distorted view of the localized economic positives, while ignoring
the strategic business-economic costs to the state as a whole
(such as West Virginiatand to different regions of the United
States (California Governor's Office 1992, Kindt 1995). In 1994,
all of the various experts who testified before the U.S. House of
Representatives Committee on Small Business criticized the
impacts that casino-style gambling activities inflict upon the
criminal justice system, the social welfare, system, small
businesses, and the economy (Congressional Hearing 1994).
Utilizing legalized gambling activities as a strategy for economic
development was thoroughly discredited during the hearing.
Florida is the only state which has conducted a comprehensive
statewide analysis of the impacts of legalized ganlbling activities.
Its report concurred with the congressional hearing's conclusions
(Florida Budgeting Office 1994).
Since some issue areas have not received widespread public
attention in West Virginia, this analysis highlights some of the
neglected issue areas as they relate to tax revenues, social-welfare
costs, education, and job creation.
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 2 of 8
EconomicCycles and Gambling's Impact on Tax Revenues
From the perspective of U.S. economic history, the United States
has had previous economic cycles with widespread legalized
gambling activities. The most relevant cycle occurred after the
American C,ivil War and paralleled the post-bellum migration to
the "Wild West.'.' Although gambling proliferated during this
time~frame, within a few years the trend toward prohibiting
gambling activities had begun, and by ] 910 there was virtually
no legal gambling in the United States. Gambling activities were
not just prohibited via state statutes and local ordinances, but
more importantly, these prohibitions were incorporated into most
state constitutions. The fact that state constitution a] provisions
were utilized to make it as difficult as possible for future
generations to legalize gambling activities (and thereby
experiment once again with a classic "boom and bust" economic
cycle) lends substantial credence to arguments that both
historically and currently, the legalization of gambling activities
eventually causes: (I) increased taxes, (2) a loss of jobs from the
overall region, (3) economic disruption of other businesses, (4)
increased crime and (5) large social-welfare costs for society in
genera] and government agencies in particular. For example, two
studies of the riverboat casinos in Illinois concluded that for
every one job created by the riverboats, most of the surrounding
communities probably lost one or more jobs from pre-existing
businesses (Grinols ] 994; Grinols and Omorov ] 995).
In recent economic history, legalized gambling activities have
been directly and indirectly subsidized by the taxpayers. The
field research throughout the nation indicates that for every dollar
the legalized gambling interests indicate is being contributed in
taxes, it usually costs the taxpayers at least 3 dollars-- and higher
numbers have been calculated (Politzer, Morrow and Leavey
]98]; Better Government Association] 992; Florida Budget
Office 1994). These costs to taxpayers are reflected in: (I)
infrastructure costs, (2) relatively high regulatory costs, (3)
expenses to the criminal justice system, and (4) large social-
welfare costs (Illinois Governor's Office ]992). Accordingly,
several state legislators (e.g" in South Dakota) have called for at
least partially internalizing these external costs by taxing all
legalized gambling activities at a straight 50 percent tax rate.
Furthermore, as a matter of good public policy, state officials and
legislators in Illinois have proposed legislation to prohibit
contributions by legalized gambling interests to politicians and
political campaigns. In the case of casinos, New Jersey already
has such prohibitions, but other states have neglected to enact
similar prohibitions. Political scientists have raised concerns that
the newly developing constituencies in the licensed gambling
industry are becoming so widespread that the industry can dictate
economic, social, and tax policies. For example, the industry
drafted a state constitutional referendum in Florida which would
have mandated the introduction of casinos into communities even
if a particular community voted unanimously against a casino
(Dyckman ] 994). The industry spent approximately $3 million to
get the Florida referendum on the ballot and $6.5 million to
campaign for the casinos-- more than the combined gubernatorial
campaigns of Governor Lawton Chiles and his challenger Jeb
Bush (Lavelle 1994). In these contexts, an article in the Columbia
Journalism Review cautions the news media "flat out ask
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 3 of8
[experts, academics, and even other reporters] if they make
money off the industry" (Simurda 1994).
Social Welfare Costs
Legalized gambling activities act as a regressive tax on the poor
(Clotfelter and Cook 1989). Specifically, the legalization of
various fonns of gambling activities makes "poor people poorer"
and can dramatically intensify many pre-existing social-welfare
problems. Demographic analyses reveal that certain
disadvantaged socioeconomic groups tend to gamble
proportionately greater amounts of their overall income and
marketing efforts, particularly by state lotteries, have allegedly
been directed at these target groups.
In a specific example involving casinos, a 1995 Wisconsin report
concluded that "[w]ithout considering the social costs of
compulsive gambling, the 'rest-of-the-state' areas lose- or,
transfer in- $223.94 million to the local gaming areas.
Considering the lowest estimated social costs of problem
gambling, the rest of... [Wisconsin] loses $318.61 million to
gambling" (Thompson, Gazel, and Rickman 1995). This report
also concluded that without casino gambling, many local citizens
would have increased participation in other "outside" activities.
"More than 10% of the locals would spend more on groceries ifit
were not for the casino, while nearly one-fourth would spend
more on clothes. Thirty-seven percent said that their savings had
been reduced since the casino had opened ..." (Thompson, Gazel,
and Rickman 1995).
From the business perspective, businesses are not naive. For
example, "in a rare public stand on a controversial political issue,
the Greater Washington Board of Trade's 85-member board voted
unanimously against" Mayor Sharon Pratt Kelly's initiative to
bring casino-style gambling to Washington, D.C. (emphasis
added, Spayd and Woodlee 1993). With the exception of the
cluster services associated with gambling, new businesses tend
not to locate in areas allowing legalized gambling because of one
or more of the aforementioned costs. In areas saturated with
legalized gambling activities, pre-existing businesses face added
pressures that push them toward illiquidity and even bankruptcy.
Although South Dakota does not constitute a saturated gambling
state, this trend has already been reported there. South Dakota
basically had no gambling in 1988 and then instituted casino
gambling and video lottery tenninals by the end of 1989. Within
two years legalized gambling a,ctivities constituted one of the
leading causes of business and personal bankruptcies among
South Dakota residents (whereas this cause was virtually
nonexistent in 1989) (Nelson 1993). More subtly, traditional
businesses in communities which initiate legalized gambling
activities can anticipate increased personnel costs due to
increased job absenteeism and declining productivity (Kindt
1994a). The best blue-collar and white-collar workers, the Type-
A personalities, are the most likely to become pathological
gamblers (Kindt 1994b). A business with 1,000 workers can
anticipate increased personnel costs of $500,000 or more per
year-simply by having various fonns of legalized gambling
activities accessible to its workers.
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 4 of 8
To some extent businesses must already internalize the societal
costs associated with assisting personnel with drug or alcohol-
related problems. Lega]izing various gambling activities
increases the number ofprob]ems related to pathologica]
gambling in the context of the workforce, and these costs are
reflected in increased personnel costs-- such as "rehabilitation
costs..." which can easily range from $3,000 to $20,000 (or more)
per pathological gambler (Kindt 1994b). In the context of the
hea]thcare debate, the spectre of these unanticipated costs can
raise further concerns to businesses already being asked to bear
certain health care costs.
Education Costs
Gambling activities and the gambling philosophy are directly
opposed to sound business principles and economic development.
Lega]ized gambling activities also negatively affect education--
both philosophically and fiscally (Better Government Association
1992; C]otfelter and Cook 1989). Adherence to a philosophy of
making a living via gambling activities not only abrogates the
perceived need for an education, but also reinforces economically
unproductive activities (and is statistically impossible since the
"house" always wins eventually). In states with legalized
gambling activities which were initiated allegedly to bolster tax
revenues to "education," the funding in "rea] dollars j, has almost
uniformly decreased.
The Patho]ogical Gamb]er Problem
States which embrace ]egalized gambling activities can expect
enormous socioeconomic costs and a decline in the quality of
life. Unlike traditional business activities, legalized gambling
activities cater to a market consisting of addicted and potentially
addicted consumers, and most pre-existing traditional businesses
will find it quite difficult to compete for "consumer dollars"
which are being transformed into "gamb]ing dollars." For
example, the field research strongly suggests that the introduction
of widespread legalized gambling in South Dakota, including
casinos and video lottery terminals (VL Ts), over a two-year time
span caused a one percent increase in the num ber of problem and
probable pathological gamblers-- a recognized addictive behavior
pursuant to the American Psychiatric Association (C]otfe]ter and
Cook 1989; Better Government Association 1992). Each new]y-
created pathological gambler has been calculated to cost society
from $13,200 to $52,000 per year (Mary]and Department of
Health 1994; Better Government Association 1994). These costs
are not just reflected in society as a whole, but impact on all
businesses. In particular, small businesses could easily
experience disproportionate negative impacts, and unlike large
corporations, small businesses would be less likely to have the
asset base necessary to cushion against those negative impacts.
Sociologists almost uniformly report that increased gambling
activities which are promoted as sociologically "acceptable" (the
acceptability factor) and which are made "accessible" (the
accessibility factor) to larger numbers of people will increase the
number of pathologic a] gamblers (Goodman ]994: Po]itzer,
Morrow and Leavey] 98]; Better Govemment Association 1992;
Mary]and Department ofHea]th 1994). The baseline of
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 5 of8
patbological gamblers as part of the population begins at .77
percent as reported by the 1976 U.S. Commission on Gambling
(U.S. Commission 1976). Since gambling has been legalized and
made accessible in several states, the range has increased to ].5
to 5 percent in those states (Alberta Lotteries and Gamb]ing
1994). This phenomenon was specifically confirmed by a 1995
study which concluded that the lifetime probable pathological
and problem gamblers in Iowa increased from 1.7 percent of tbe
public in ]989 to 5.4 percent in 1995 (Iowa Department of
Human Services 1995, Petroski 1995). Similarly, a limited study
of Native Americans revealed a rate for lifetime probable
pathological and problem gamblers of 14.5 percent in casino
areas (A]berta Lotteries and Gamb]ing 1 994). These
developments translate into increases in socioeconomic costs
which must be addressed and absorbed primarily by taxpayers,
but also by businesses, charities, social-welfare organizations and
governmental units.
Negative Impact on Job Creation
On a regional leve], the combined ranges of these various
socioeconomic costs are so large that they tend to dwarf the
localized economic positives (California Governor's Office
1992). These drains on society could easily translate into a net
loss of jobs on a statewide or regional level. Furthermore, it can
be argued that the combined economic positives and negatives
result in a net negative economic multiplier (Goodman] 994;
Teske and Sur 199 I). From the perspective of business-
economics and strategic development, major businesses are and
should be concerned with the trend toward expanding various
forms of legalized gambling activities, Among other reasons,
nongambling related businesses will not be competing for
consumer dollars or recreational dollars on a "level playing
field," because legalized gambling activities can cater to an
addicted and potentially addicted market segment. Since the U.S.
economy and most state economies are extensive in scope, the
socioeconomic negatives associated with legalized gambling
activities can remain hidden for long periods of time. However,
just because a particular activity is "legalized" by a state
government does not mean that thf: negative business or societal
impacts have been eliminate-- or even reduced.
Conclusion
Increasingly, taxpayers and businesses are beginning to realize
that, as Professor Jack Van Der Slik has summarized for much of
the academic community, state-sponsored gambling "produces no
product, no new wealth, and so it makes no genuine contribution
to economic development" (Van Der Slik 1990). Business-
economic hIstory supports this proposition. The recriminalization
of gambling activities occurred] 00 years ago after a brief
gambling boom following the Civil War. Most state legislatures
utilized constitutional provisions to recriminalize gambling,
because lawmakers wanted to make it as difficult as possible for
future generations to experiment with the classic "boom and
bust" cycles and the concomitant socioeconomic negatives
occasioned by legalized gambling activities. To paraphrase
Georg Hegel's common quote, "those who forget the lessons of
economic history are condemned to relive them" (Bartlett 1968).
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 6 of 8
References
Alberta Lotteries and Gaming. 1994. Gambling and Problem
Gambling in Alberta. Edmonton, Alberta. January.
American Psychiatric Association. 1994. Diagnostic and
Statistical Manual of Mental Disorders. ("pathological
gambling") Vol. 312.31, pp. 615-618.
Bartlett, Jolm. 1968. Familiar Quotations. 14th. Edition. Boston;
Little Brown.
Better Government Association. 1992. Staff White Paper. Casino
Gambling in Chicago. Chicago, 11.
California Governor's Office of Planning and Research, 1992.
California and Nevada: Subsidy, Monopoly, and Competitive
Effects of Legalized Gambling. ES-l. Sacramento, CA:
California's Governor's Office. December.
Clotfelter, Charles and Phillip Cook. 1989. Selling Hope,
Cambridge, MA: National Bureau for Economic Research,
Harvard University Press.
Congressional Hearing, 1994. The National Impact of Casino
Gambling Proliferation: Hearing Before the House Committee on
Small Business. I03rd Congress, 2nd. Session. Washington, D.C.
Dyckman, Martin. 1994. "Misleading the Public," St. Petersburg
Times. November I, p. AI3.
Florida Governor's Office of Planning and Budgeting. 1994.
Casinos in Florida: An Analysis of the Economic and Social
Impacts. Tallahassee, FL: Florida Governor's Office.
Goodman, Robert. 1994. Legalized Gambling as a Strategy for
Economic Development. Amherst, MA: University of
Massachusetts-Amherst, Center for Economic Development.
Grinols, Earl. 1994. "Bluff or Winning Hand? Riverboat
Gambling and Regional Employment and Unemployment."
Illinois Business Review (Spring ): 6-11.
Grinols, Earl and J. Omorov. 1995. "Development or Dreamfield
Delusions?: Assessing Casino Gambling's Costs and Benefits."
University of Illinois, September,
Illinois Governor's Office. 1992. Governor James Edgar, Press
Release. "Governor Warns Land-Based Casinos Could Bring
Crime Surge as well as Overall Loss of Jobs and State
Revenues." September 29 (summarizing several Illinois State
reports).
Iowa Department of Human Services. 1995. Gambling and
Problem Gambling in Iowa: A Replication Survey. July 28,
Kindt, John. 1995. "U.S. National Security And The Strategic
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 7 of 8
Economic Base: The BusinesslEconomic Impacts Of The
Legalization Of Gambling Activities." Saint Louis Law School
Journal 39, p. 567,
Kindt, John. 1994a. "The Negative Impacts Of Legalized
Gambling On Businesses." University of Miami Business Law
Journal 4, p. 93.
Kindt, John, 1994b. "The Economic Impacts Of Legalized
Gambling Activities." Drake Law Review 43, pp. 51,66 n. 119.
Lavelle, Louis. 1994. "Voters Deal Loss to Casinos; Gambling
Backers Lose Despite $16-5 Million Campaign." Tampa Tribune.
November 9, pp. 1,5.
MaIYland Department of Health and Mental Hygiene, Alcohol
and Drug Abuse Administration. 1994. Task Force on Gambling
Addiction in Maryland, Annapolis, MaIYland.
Nelson, Todd. 1993. "S.D. bankruptcies down 5 percent: Judge:
Gambling caused most cases." Argus Leader (Sioux Falls, S.D.)
JanuaIY 15, p. 1.
News Leader (Springfield, Missouri). 1995. "Jeff City rejects
riverboat gambling approved in '92." November 8, p. 1.
Petroski, William. 1995. "Study: More gamblers in jeopardy."
Des Moines Register. August 25, pp. AI, A2.
Politzer, Robert, James Morrow and Sandra Leavey. 1981.
Report on the Societal Cost of Pathological Gambling and the
Cost-Benefit Effectiveness of Treatment. 5th National
Conference on Gambling and Risk Taking.
Teske, Paul and Bela Sur. 199 I. "Winners and Losers: Politics,
Casino Gambling, and Development in Atlantic City." Policy
Studies Review IO (Spring! Summer): 130-137.
Simurda, Stephen F. 1994. "When Gambling Comes To Town."
Columbia Journalism Review (January-February): 36-38.
Spayd, Liz and Yolanda Woodlee. 1993. "Trade Board Rejects
D.C. Casino Plan." Washington Post. September 25, pp. AI, A8.
Thompson, William, Ricardo Gazel, and Dan Rickman 1995. The
Economic Impact of Native American Gambling in Wisconsin.
Wisconsin Policy Research Institute Report.
U.S. Commission on the Review of the National Policy Toward
Gambling. 1976. Gambling in America. Washington, D.C.:
United States Government Printing Office.
Van Der Slik, Jack. 1990. "Legalized gambling: predatory
policy." Illinois Issues (March): 10.
Wartzman, Rick. 1995. "Bayou Backlash: Gambling Is Proving
To Be a Poor Wager For State of Louisiana: Business Is
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 8 of 8
Disappointing And an FBI Graft Probe Roils a Jaded Electorate."
Wall Street Journal. September 11, p. 1.
Worthington, Rogers. 1995. "Poor get poorer at tribal casinos,
says study of Wisconsin 'garners'." Chicigo Tribune. April 11, p.
1.
ioin the discussion / what are the odds" / gambling' pro and con / ~ / ~ / f.1cts & stats / tapes & transcripts / llIlli
exolore FRONT] INE
New Content Copyright <<:> 1998 PBS and WGBH/FRONTLINE
, m~>.
!onhne
ORDINANCE NO. 2004-03
AN ORDINANCE OF THE CITY OF WINTER SPRINGS,
FLORIDA, AMENDING CHAPTER 10 OF THE CITY CODE,
ENTITLED "LICENSES AND BUSINESS REGULATIONS";
ESTABLISHING A NEW SECTION 10-115, ENTITLED "SLOT
MACHINES OR DEVICES"; PROVIDING FOR A DEFINITION
OF AND PROHIBITING THE COMMERCIAL USE OF SLOT
MACHINES OR DEVICES IN THE CITY OF WINTER SPRINGS;
PROVIDING FOR THE REPEAL OF PRIOR INCONSISTENT
ORDINANCES AND RESOLUTIONS, INCORPORATION INTO
THE CODE, SEVERABILITY, AND AN EFFECTIVE DATE.
WHEREAS, the City is granted the authority, under S 2(b), Art. VIII of the State Constitution, to
exercise any power for municipal purposes, except when expressly prohibited by law; and
WHEREAS Section 849.15, Florida Statutes, prohibits businesses from permitting the operation
of or using or keeping slot machines or devices; and
WHEREAS, Section 849 .16( 1), Florida Statutes, defines slot machines as any machine or device
used in such a way that as a result of the insertion of money or other object, the machine is caused to be
operated, and if the user by reason of any element of chance or other outcome unpredictable by the player
may be entitled to a prize or replay; and
WHEREAS, Section 849.161, Florida Statutes, provides a limited exemption for amusement
games or devices, which are operated within a bona-fide arcade or amusement center, with fifty (50) or
more machines or devices and which strictly through the application of skill a player may become entitled
to receive points or coupons which may be exchanged for merchandise, excluding cash and alcoholic
beverages; and
WHEREAS, businesses known as "Adult Arcades" or "Adult Arcade Amusement Centers" have
begun to operate within cities and counties across Florida providing casino styled gaming, which
significantly and adversely affect the public health, safety, and welfare of the citizens, residents and
tourists of those areas; and
WHEREAS, despite the fact that these businesses use slot machines, these businesses are
attempting to use the state law exemption, commonly referred to as the "Chuck E' Cheese" exemption,
found in Section 849.161, Florida Statutes, in order to operate their businesses; and
WHEREAS, it has been determined by the Florida courts that machines or devices that are
determined to have an inherent element of chance are not exempt under Chapter 849, Florida Statutes.
Deeb v. Stoutamire, 531 So. 2d 873 (Fla. 1951);In ReForfeiture of Forty-Seven Video Games v. Hunter,
799 So. 2d 221 (Fla. 2nd DCA, 2001) (Based upon the testimony ofthe owner/operator of an adult arcade
amusement center the court found that there was an inherent element of chance in the forfeited slot
City of Winter Springs
Ordinance No. 2004-03
10f5
machines, as the player could win, without any application of skill, purely in proportion of the natural
odds, i.e., there was a 1 in 26 chance that the spinning reels would stop on the proper icon, should the
player randomly press the stop button.); and
WHEREAS, the State of Florida has prohibited all forms of gambling, including the operation of
any kind of slot machines, and gambling is punishable as a misdemeanor under the Florida Statutes; and
WHEREAS, the City Commission finds that slot machines or devices can lead to gambling
addictions, along with the secondary consequences that such additions impose upon the community,
including increases in crime, debt, suicide and other negative effects on the health, safety and welfare of
the residents and visitors of the areas where these devices exist; and
WHEREAS, the City of Winter Springs has the duty and authority, through its police powers, to
ensure appropriate regulation, licensing and permitting of businesses; and
WHEREAS, both the State of Florida and Seminole County have prohibited all forms of
gambling, including the operation of any kind of slot machines, and gambling is a misdemeanor under the
statutes; and
WHEREAS, stricter local regulation of these gaming or slot machines is permissible provided
such regulations are not in conflict with the State laws; and
WHEREAS, this ordinance is intended to provide stricter regulations of gaming or slot machines
than those provided by Chapter 849, Florida Statutes and to prohibit all commercial casino type slot
machines as defined herein; and
WHEREAS, the City Commission of the City of Winter Springs, Florida, hereby finds this
ordinance to be necessary in the best interests of the public health, safety, and welfare of the citizens of
Winter Springs.
NOW THEREFORE, The City Commission ofthe City of Winter Springs, Florida hereby
ordains as follows:
Section 1. Recitals. The foregoing recitals are hereby fully incorporated herein by reference
as legislative findings and the intent and purpose of the City Commission of the City of Winter Springs.
Section 2. Code Amendment. That Chapter 10, entitled "Licenses and Business
Regulations", ofthe Code of Ordinances, City of Winter Springs, Florida, are hereby amended as follows:
(underlined type indicates additions and strilceout type indicates deletions, while asterisks (* * *) indicate
a deletion from the Ordinance of text existing in Chapter 10. It is intended that the text in Chapter 10
denoted by the asterisks and set forth in this Ordinance shall remain unchanged from the language
existing prior to adoption of this Ordinance).
CHAPTER 10 - LICENSES AND BUSINESS REGULATIONS
City of Winter Springs
Ordinance No. 2004-03
2of5
* * *
ARTICLE IV. AMUSEMENTS
* * *
Sec. 10-115 Slot Machines and Devices
f!l Purpose and Intent. The purpose and intent ofthis section is to regulate and prohibit, within the
City of Winter Springs, Florida. slot machines or devices and casino-type slot machines as
defined in this section or as prohibited by Chapter 849, Florida Statutes.
.au Definitions. The following words, terms, and phrases, when used in this section, shall have the
meanings ascribed to them in this section, except where the context clearly indicates a different
meamng:
ill Slot machine or device shall mean an amusement machine, device, game or instrument
operated bv means of the insertion of a coin, bilL currency, credit card, debit card, ticket.
token or slug, for use as a game, contest of skill or amusement of any description, which
device, machine, game or instrument is similar to, or in the nature of a casino-type slot
machine
ill Casino-tvpe slot machine shall mean a machine, device, game or instrument which is
prohibited by Section 849.16, Florida Statutes, and which, by means of skill or chance, or
both, has as its object the lining UP of two or more colors, lights, or spinning obiects, or
any combination thereof.
!tl Slot machine or device prohibited.
ill It shall be unlawful to have. own, operate, use or allow another to have, own operate or
use a slot machine or device in the City of Winter SPlings, and slot machines or devices
are hereby prohibited in the City of Winter Springs.
ill No occupational licenses shall be issued by the city for any business, enterprise, or
organization that proposes to operate a slot machine or device in the City and no
occupational license shall be issued by the City for a slot machine or device.
@ Exemption. This section is not intended to and specifically does not regulate:
ill merchandise vending machines;
ill coin-operated mechanical or electrical musical instruments or devices;
ill coin-operated amusement devices that are not in the nature of, or similar to, casino- type
City of Winter Springs
Ordinance No. 2004-03
30f5
slot machines: or
ill coin-operated amusement devices that mav be in the nature of or similar to casino-type
slot machines or devices, which are not used for commercial or betting purposes but kept
only for purposes of disp1av.
Section 3 Repeal of Prior Inconsistent Ordinances and Resolutions. All prior
inconsistent ordinances and resolutions adopted by the City Commission, or parts of prior ordinances and
resolutions in conflict herewith, are hereby repealed to the extent of the conflict.
Section 4. Incorporation Into Code. This ordinance shall be incorporated into the Winter
Springs City Code and any section or paragraph number or letter and any heading may be changed or
modified as necessary to effectuate the foregoing.
Section 5. Severability. If any section, subsection, sentence, clause, phrase, word or
provision of this ordinance is for any reason held invalid or unconstitutional by any court of competent
jurisdiction, whether for substantive, procedural, or any other reason, such portion shall be deemed a
separate, distinct and independent provision, and such holding shall not affect the validity of the
remaining portions of this ordinance.
Section 6. Effective Date. This Ordinance shall become effective immediately upon
adoption by the City Commission ofthe City of Winter Springs, Florida, and as provided by City Charter.
ADOPTED by the City Commission of the City of Winter Springs, Florida, this _ day of
,2004.
JOHN F. BUSH
Mayor
ATTEST:
ANDREA LORENZO-LUACES
City Clerk
City of Winter Springs
Ordinance No. 2004-03
40f5
Approved as to legal form and sufficiency for
the City of Winter Springs only
Anthony A. Garganese, City Attorney
First Reading:
Second Reading:
Effective Date:
City of Winter Springs
Ordinance No. 2004-03
50f5
BROWN, SALZMAN, WEISS & GARGANESE, P.A.
Attorneys at Law
Usher L. Brownt
Suzanne D'AgrestaO
Anthony A. GarganeseO
Gary S, Salzman'
John H. Wardt
Jeffrey S. Weiss
Offices in Orlando, Kissimmee,
Cocoa & Viera
Debra S. Babb-Nutcher
Jeffrey P. Buako
John U. Biedenharn, Jr.
Joseph E. Blitch
Douglas Lambert
Katherine Latorre
Michelle A. Reddin
tBoard Certified Civil Trial Lawyer
'Board Certified Business Litigation Lawyer
oBoard Certified City, County & Local Government Law
Erin J. O'Leary
Of Counsel
November 26, 2003
The Honorable Mayor John F. Bush and
Members of the City Commission
1126 East State Road 434
Winter Springs, Florida 32708-69l2
Re: Adult Arcade Amusement Centers - Slot Machines
City of Winter Springs / General File No. 1193
Dear Mayor Bush and Members of the City Commission:
This correspondence follows a recent flurry of news reports regarding casino style gaming
rooms which are opening around the state. It is the intent of this letter to bring the City up to date
on the issues surrounding these establishments and to provide several possible alternatives for
dealing with such an application.
Typical Scenario
These gaming rooms or arcades often look like traditional casinos inside, except they are in
neighborhood strip centers. The owner/operators typically lower the lights, the staff is dressed in
traditional casino attire, the owners/operators often serve complimentary wine and food, and usually
require all patrons to be over the age of twenty-one (21) years. The arcades typically have slot
machine style games. Some are the traditional reel styled machines, while others are video images
that appear to spin. The insertion of money into the machine provides credits for play. Typically
each game costs One Dollar ($1.00), but players can place as much as One-Hundred Dollars
($100.00) into the machine and receive one-hundred (100) credits. The machine is then activated
by placing a bet or using a credit. Each of the three reels is then stopped individually by pushing a
stop button, sometimes labeled by the word "skill". Should a player end up with the necessary
images or icons in the proper location, the player wins and is awarded a coupon. The coupons may
then be added together and traded for gift cards, restaurant coupons, discounts for services, as weB
as other merchandise.
225 East Robinson Street, Suite 660 . P.O. Box 2873. Orlando. Florida 32802-2873
Orlando (407) 425-9566 Fax (407) 425-9596. Kissimmee (321) 402-0144. Cocoa & Viera (866) 425-9566
Website: www.orlandolaw.net . Email: firm@orlandolaw.net
The Honorable Mayor John F. Bush and
Members of the City Commission
November 26, 2003
Page 2
Issue
The question posed for state and local officials is whether these establishments violate
Chapter 849,Florida Statutes, by the use or operation of a slot machine or if these establishments are
exempt from the regulation as the slot machines are games of skill, as opposed to ones of chance.
Short Answer
We believe the exemption provided in Section 849.161, Florida Statutes, for amusement
games or machines located within an arcade amusement center is limited. Notwithstanding the
foregoing, it appears possible that a casino styled arcade amusement center may exist lawfully under
this exemption.
Legal Analysis
Pursuant to Section 849.15(1), Florida Statutes, it is unlawful to possess or pem1it the
operation of any slot machine or device. Section 849.16, Florida Statutes defines those machines
which qualify as a slot machine or device as:
[O]ne that is adapted for use in such a way that, as a result of the
insertion of any piece of money, coin, or other obj ect, such machine
or device is caused to operate or may be operated and if the user, by
reason of any element of chance or of any other outcome of such
operation unpredictable by him or her, may:
(a) Receive or become entitled to receive any piece of money,
credit, allowance, or thing of value, or any check, slug, token,
or memorandum, whether of value or otherwise, which may
be exchanged for any money, credit, allowance, or thing of
value or which may be given in trade; or
(b) Secure additional chances or rights to use such machine,
apparatus, or device, even though it may, in addition to any
element of chance or unpredictable outcome of such
operation, also sell, deliver, or present some merchandise,
indication of weight, entertainment, or other thing of value.
Thus, if the receipt ofa prize is dependent on any element of chance, the machine would fall within
the above definition of a slot machine proscribed by Section 849. I 5, Florida Statutes.
The Honorable Mayor John F. Bush and
Members of the City Commission
November 26, 2003
Page 3
There is an exemption, however, for amusement games or machines, operated within a
bonafide arcade or amusement center, which in accordance with Section 849.161, Florida Statutes
are:
[C]oin operated and which by application of skill may entitle the
person playing or operating the game or machine to receive points or
coupons which may be exchanged for merchandise only, excluding
cash and alcoholic beverages, provided the cost value <!f the
merchandise or prize awarded in exchange for such points or coupons
does not exceed 75 cent~ on any game played.
"Arcade amusement center" is defined in Section 849.161, Florida Statutes as "...a place of business
having"at least 50 coin-operated amusement ganles or machines on premises which are operated for
the entertainment of the general public and tourists as a bona fide amusement facility."
Therefore, Section 849.161, Florida Statutes, creates an exemption for a machine or device
that would otherwise be prohibited by Chapter 849, Florida Statutes, ifthe machine is located in an
arcade amusement center and if, by application of skill, the player may receive a prize worth no more
than seventy-five cents ($0.75). This exemption has been referred to as the "Chuck E Cheese
Exemption".
The owner's and operators of the casino style arcades claim this exemption applies to their
establishments. They claim the fact that a player must push a button, sometimes labeled with the
word "skill", to stop the spinning reels or images is a function of skill and not chance. It is claimed
that if a player plays enough, concentrates enough, and has fast enough reflexes, the player will be
able to stop the reels, through the application of skill, and win a coupon redeemable for prizes.
The Florida Attorney Generat was asked in 1995 whether the pushing of a button marked
"skill" equals the application of skill required by Section 849.161, Florida Statutes. The Attorney
-General found that while the outcome ofthe game does not have to be wholly dependent upon skill,
skill must be a factor in the outcome. But, the mere pushing of a button marked "skill" would not,
in and of itself, appear to constitute the application of skill for purposes of qualifying for the
exemption under Section 849.16l, Florida Statutes. The Attorney General went on to note that the
information provided indicated that the objects painted on the reels were impossible to identify due
to the speed at which the reels turned, and that when the button was depressed the reels did not
immediately stop spinning, therefore under those facts, it appeared that the knowledge, attention, or
practice of the player would not affect the element of chance that dominates the result of the play,
and therefore the machine would not qualify for the exemption in Section 849.161, Florida Statutes.
The Honorable Mayor John F. Bush and
Members of the City Commission
November 26, 2003
Page 4
The Florida courts have also looked at this issue. In Deeb v. Stoutamire, 53 So.2d 873, (Fla.
1951) the Florida Supreme Court considered the meaning of the phrase "any element of chance or
of other outcome of such operation unpredictable by him", and determined the element of chance
or unpredictability is not supplied by the player, but must be inherent in the machine. In other
words, when considering whether a machine or device is exempt from the requirements of Sections
849.15 and 849.16, Florida Statutes, it must be determined that there is no chance orunpredictability
in the machine.
Later, in State et al. v. Broward Vending, Inc., 696 So.2d 85l (Fla. 4th DCA 1997) the Fourth
District Court of appeal applied this standard when considering whether slot machines were properly
seized under the Florida Contraband Act. The Court found that if a player does not manipulate the
levers to improve the score, the machine was preset for the player to win fifty-five (55) percent of
the time, therefore while ski II will improve the winning percentage, the game has an inherent
element of chimce. As such, it is not exempt from Chapter 849, Florida Statutes.
Most recently, In re forfeiture of Forty-Seven Video Redemption Games v. Don Hunter, 799
So.2d 22l (Fla. 2nd DCA 200l), the Second District Court of Appeals found machines seized to be
slot machines as defined by Section 849.l6, Florida Statutes, and not subject to the exemption of
Section 849.161, Florida Statutes. At trial the owner of the machines testified that there was a
chance a player could win, without skill,just in proportion to the natural odds. He testified that each
reel contained twenty-six (26) icons and that without any use of skill, there was a 1 in 26 chance that
the reel would stop on the proper icon.
It is important to note that Chapter 849, Florida Statutes, contains no express preemption
from further regulation by local governments.
Conclusion
While the exemption provided in Section 849.16l, Florida Statutes, has been found to be
limited in scope, there remain fact based determinations regarding whether a machine or device is
exempt as an amusement game of skill.
I) If the machine or device has an element of chance inherent to it, the machine or
device is not exempt from the prohibitions of Section 849.l5, Florida Statutes, and
cannot be operated or possessed.
2) If the machine or device does not have an inherent element of chance, but is instead
a skill based machine, the same is exempt and may be permitted, provided the arcade
amusement center has at least fifty (50) machines, is open to the general public and
tourists, and any award has a value of less than $0.75 per game.
The Honorable Mayor John F. Bush and
Members of the City Commission
November 26, 2003
Page 5
Recommendations
1. The City could do nothing and handle any application on a case by case basis. Any applicant
which demonstrates to the satisfaction of staffthat the machines are skill oriented, that there
will be more than fifty (50) machines within the amusement center, that the arcade
amusement center will be open to the general pubic (which could or should include children),
and that will only provide coupons with a value of less than $0.75, may be approved.
2. The City could amend the City Code to provide that any such use would be permitted by
special exception only in an appropriate zoning district. This would provide a quasi-judicial
forum whereby the applicant would have to demonstrate compliance with the Florida
Statutes and City Code. This would allow the presentation of witnesses regarding fact
determination of skill versus chance. Where compliance with the Florida Statutes is
demonstrated, the City could impose additional reasonable conditions to protect the
surrounding property owners and ensure the lawful existence of the business.
3. The City could amend the City Code to provide additional regulations relating to gaming and
arcades to effectively close the "loophole" claimed to exist in Chapter 849, Florida Statutes.
I look forward to discussing this issue at the December 8, 2003 City Commission meeting.
Anthony A. Garganese
City Attorney
AAG:j f
INDEX TO STUDIES
RELATING TO
THE EFFECTS OF
GAMBLING
TABLE
OF
CONTENTS
.
.
.
INDEX TO STUDIES RELATING TO
THE EFFECTS OF GAMBLING
(On file in the City Clerk's Office)
1. Materials presented by Schwer, Thompson, Nakamuro at the 2003 Annual Meeting
of the Far West and American Popular Culture Association, Beyond the Limits of
Recreation: Social Costs of Gambling in Southern Nevada, Las Vegas, NV,
February 1, 2003.
2. Florida Executive Office of the Governor, Casinos in Florida: An Analysis of the
Economic and Social Impacts, Florida Office of Planning and Budgeting (1994).
3. Grinols & Mustard, Business Profitability versus Social Profitability: Evaluating
Industries with Externalities, the Case of Casinos, 22 Managerial and Decision
Economics 143 - 162 (2001).
4.
Kindt, Excerpt from The Business-Economic Impacts of Licensed Casino Gambling
in West Virginia: Shorl.- Term Gain but Long- Term Pain, originally published in The
National Impact of Casino Gambling Proliferation: Hearing Before the House
Committee on Small Business, 103 rd. Congress, 2nd Session 787 (1994).
5. Volberg, The Lewin Group, Christiansen/Cummings Associates, Gambling Impact
and Behavior Study, Report to the National Gambling Impact Study Commission
(April 1, 1999).
6. Grinols, CUTTING THE CARDS AND CRAPS: Right Thinking About Gambling
Economics. December 21,2001.
TAB
" 1 "
[sO] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada by... Page 1 of 20
. NCALG National CoalitionAgainst Legalized Gambli"!l.
Together we
can protect
our children
from a terrible
m,enace.
Beyond the Limits of Recreation: Social Costs of
Gambling in Southern Nevada
Information Center
R. Keith Schwer
William N, Thompson
Daryl Nakamuro
Document Librarv
NCALG Newsletters
Board of Directors
2003 Annual Meeting of the
Far West and American Popular Culture Association
.
Contact NCALG
SEARCH
Las Vegas, NV
February 1, 2003
I. THE SETTING: HIGH-STAKES MECCA
Legalized gambling generates more revenues than any other popular leisure-time
activity in America. In 2001, gambling activities generated over $63 billion for their
operators, who included private companies, state governments, Native American tribes,
and non-profit charity groups. Casinos produced nearly $42 billion of this amount.
(American Gaming Association) The revenues from gambling equate to over 1 % of the
national domestic product. (United States Statistical Abstract). Clearly, gambling is a
major element in American popular culture.
The Las Vegas Strip is Mecca for gambling, and particularly for legalized casinos.
Nevada casinos reported gross gambling profits (player losses minus prizes given to
players) of about $9.3 billion last year, with about $7 billion produced by the casinos of
Clark County, which includes the city of Las Vegas as well as the Las Vegas Strip,
(Nevada Gaming Commission),
Though the benefits are widely touted, they do not reveal the full picture. In particular,
the cost side to the gambling equation is often debated, but seldom measured. Some
argue that there is no economic gain from gambling activity as it represents only a
neutral exercise in exchanging money from one set of hands to another-in short, an act
of income redistribution. As such, they argue that no product is created to add wealth to
society, whereas the costs of the exchange (time and energy of players, dealers, and
other casino employees) represent a net economic loss for society. In short, they argue
that gambling offers no ~ecreation value.
:.
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada by... Page 2 of20
The leading textbook author in economics has stated:
.
(There is) a substantial economic case to be made against gambling...lt involves simply
sterile transfers of money or goods between individuals, creating no new money or
goods, Although it creates no output, gambling does nevertheless absorb time and
resources. When pursued beyond the limits of recreation, where the main purpose after
all is to "kill" time, gambling subtracts from the national income. (Samuelson, 1970, p.
402) Thus, this viewpoint identifies beyond the limits of recreation as a cost to society,
what has come to be known as "the social costs of gambling," This study focuses on
social costs of gambling in a single economic region-Southern Nevada, the location of
the Mecca of gambling.
In the highly charged environment of public policy on gambling, some have used the
religious and moral framework to advance the argument that gambling for recreation is
a small fraction of the total. On the other hand, gaming proponents have argued the
opposite, Only slowly have people come to the measurement and evaluation of
gambling costs hat go beyond the limits of recreation. The research to date suggests a
range of possible outcomes. Nevertheless, left unanswered is the magnitude of the
social costs of gambling in a mature gaming market such as Las Vegas.
.
This study reports findings gathered from 99 members of Gamblers' Anonymous (GA)
groups residing in the Las Vegas Metropolitan Area. The estimated social costs from
gambling reflect information data gathered from a sUNey of problem gamblers who are
most likely to avoid denial. The problem of denial limits the interpretation of findings
based on general sUNeys. Participants in the study of GA members have come to
recognize the presence of addictive behavior in their life, have begun treatment, and are
more likely to reveal information in a forthright manner than persons who might respond
to a general sUNey, Working from the descriptive statistics of our sUNey, we estimate
the cumulative social costs resulting from the presence of residents with serious
problem gamblers in Southern Nevada, a mature gaming venue.
The identification of actual social costs of problem gambling activity, if possible, has
practical value for th~ gambling industry and for governmental policy makers
establishing rules pertaining to gambling. For example, findings could direct industry
leaders toward strategies to prevent and mitigate problems and to aid in treatment of
troubled gamblers. Similarly, a knowledge of the extent of social costs in a mature
environment should be evidence for policy makers in other jurisdictions as they decide
whether or not to legalize casino gambling, permit the expansion of gambling, or
evaluate policy alternatives.
Additionally, the cost estimates could be used to develop mitigation and treatment
programs, including their funding. As such, the industry and regulators might better
fashion policies and programs to address social costs, thereby avoiding more costly
approaches such as class-action suits.
II. PAST STUDIES OF THE SOCIAL COSTS OF GAMBLING
The notion that gambling activity carries with it negative problems and costs for
societies is not a new notion. The social problem or concern for them prompted the
development of rules regarding gambling--even rules prohibiting gambling--since early
recorded history. Most often the rules were incorporated in religious doctrines and in
stories that become part of religious lore.
.
Our British legal heritage is replete with concerns about gambling. Richard II prohibited
games in 1388 as he felt they interfered with the performance of military duties. In 1710
the Statute of Anne 1712 banned enforcement of gambling debts through government
courts, as such, enforcement was viewed as detrimental to the property holdings of the
"
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada by... Page 3 of 20
.
noble classes. The religious and legal heritage from days of antiquity may give us
perspectives about the vexing questions posed above. However, it was not until very
recent years that scholars attempted a frontal assault on the questions. Especially, this
has been the case of making attempts to put actual cost figures on adverse social
phenomena resulting from activity of gamblers. (Cabot, Thompson, Tottenham, and
Braunlich, 1999)
Contemporary studies chronicle the impacts that the pathological gambler imposes-- not
only onto himself or herself, but also onto family members, friends, co-workers, those
with whom he or she has business relationships, and onto the general public as well. It
has been estimated that between 10 and 15 persons are directly and adversely affected
by the pathological gambler. The gambler will borrow from close associates, the
gambler very likely will also steal. And when the associates can't pick up the pieces, the
entire society may have to pay for welfare, for treatment costs, for police service, for
jails and prisons. (Leisure, 1998)
It is not easy to come up with definitive money figures which can discern the exact
social costs caused by each compulsive gambler.
There are definitional issues in deciding exactly what a "social" cost is, and there are
methodological problems in calculating costs, even where one knows the specific cost
item. Several have offered opinions about the societal costs. Lesieur and Puig(??)
examined several illegal behaviors in general and insurance frauds in specific. They
indicate a monumental cost for society from this fraudulent activity--one third of
insurance fraud is assigned to gamblers. John Kindt testified that the social costs of an
individual compulsive gamblers was between $13,000 and $52,000 a year. (Kindt,
1994 ).
.
In 1981 Robert Politzer, James Morrow, and Sandra Leavey made an analysis of the
annual costs to society of untreated pathological gamblers. The costs included lost
productivity, criminal system costs, and "abused dollars," an illusive term that included
not only bad debts but also all money lost at gambling. Their information was gathered
from 92 persons receiving treatment at the Johns Hopkins Compulsive Gambling
Counseling Center. They found that the average "bottomed-out" gambler imposed a
cost of $61,000 upon society over the last year of gambling, A "more average"
pathological gambler imposed an annual cost of $26,000 upon society, or approximately
43% of the costs from a "bottomed- out" pathological gambler.
There is a wide range of factors that could be included in a cost analysis. One writer of
this paper participated with others in studies of the costs in Wisconsin, Connecticut, and
South Carolina. Our methodology was direct. In those studies we asked members of
Gamblers Anonymous groups as well as others in treatment about their gambling
behavior and the consequences of the behavior. (Thompson, Gael and Rickman, 1996,
and 1999; WEFA Group, 1997; Thompson and Quinn, 1999)
Where we could reasonably determine cost parameters for behaviors, we did so. Where
we could not, for example, suicides and suicide attempts, we simply did not include a
cost, even though there can be no doubt that there are social-cost consequences, We
utilized a survey instrument developed by Henry Lesieur.
.
Each survey revealed that between 20% and 30% of the respondents made actual
suicide attempts (we could not assess how many were successful). No other addictive
population has had as high a prevalence for attempts; and, therefore the misery is
surely large, and perhaps not fully understood by others, Still, the cost of the misery--
ergo personal suicide watches, close family counseling, unreported lost job productivity-
-could not be estimated. These costs were not considered. Similarly, though we could
have considered the court costs involved with divorces, we did not assign a dollar cost
"
[5d] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada by... Page 4 of20
.
for the effects of the divorce on family members, particularly children. Further research
will be need to include these social costs. Thus, we excluded some social costs for lack
of adequate measurement.
In 1999, the National Gambling Impact Study Commission (National Study) released
results of a survey that identified several costs of problem gambling including
bankruptcies, imprisonment, and legal fees for divorce, but they stopped short of
examining theft and bad debts as costs, and they did not seek to find a total cost impact
from the activity of a pathological gambler. They did present information suggesting that
costs emanating from problem gamblers were approximately one-half those from
pathological gamblers. (NGISC) Clearly, studies have identified a wide range of social
costs, but left their measurement for further study.
In addition to the range of social costs from problem gambling, measuring the number of
persons with gambling problems is also a problem. The National Study reported data
collected by the National Research Council on the prevalence of pathological and
problem gambling in society, They reported that .9% of the population was currently in a
condition of severe or pathological gambling activity, and 2% problem gamblers. The
national study indicated that the numbers of abusive (pathologicar and problem)
gamblers doubled when a casino was within 50 miles of their homes. (NGISC)
The presence of readily available gambling opportunities resulting in higher incidences
of problem gambling is further supported by a recent study in Nevada. In 2001 Rachel
Volberg studied prevalence rates in Nevada. Her research found that 3.5% of the adults
in the state were current probable pathological gamblers and another 2.9% were current
problem gamblers. (Volbe
.
III. SURVEY OF TROUBLED GAMERS IN LAS VEGAS
The survey questionnaire for this study was developed from the original Leisure model,
with several modifications, The instrument appears in Appendix A. A gambler in
recovery who was a member of several Gamblers Anonymous (GA) groups in the Las
Vegas area personally distributed the surveys to leaders of fifteen G.A. groups, The
groups met on different days and nights, The leaders then distributed the surveys to
members of the groups. The members were cautioned not to divulge their identities in
any manner whatsoever as they filled out the questionnaires. They returned the written
questionnaires to an envelope outside of the view of the leader. The envelope was
unmarked. It was then picked up by the leader and returned to the coordinating gambler
in recovery. The gambler then forwarded the surveys to an assistant who entered the
data and recorded any written comments, and then returned the questionnaires. No
handwriting was preserved.
The data were collected over a two-month period from late August 2002 through
October 2002. Returns were received from 99 GA members. To be sure, the survey
responses doe not represent a random survey of G.A. members; nevertheless, the
purposive sample is a representation of the behaviors of those who filled out the
questionnaires, We also consider that these representations are, if not a random
sample, generally typical of behaviors of serious problem gamblers in Las Vegas.
.
Two important factors impact the surveying of problem gamblers. On the one hand, it
can be argued that persons who seek treatment for compulsive gambling behaviors will
be more serious problem gamblers, giving reason to be concerned with overestimates
of the actual level of problems in the community at large, On the other hand, the most
serious compulsive gamblers are most likely to exhibit a self-denial that will cause an
avoidance of treatment. In such a case, we may be missing the most serious problem
gambler behaviors in the community with this survey. In short, these factors are at cross
purposes, working toward self-correction. Also, when we consider total community
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada by... Page 5 of 20
.
costs, we purposely left out some social cost estimates, again suggesting that the
numbers presented underestimate actual costs because we do not factor in costs that
we cannot quantify (e,g, suicide attempts, divorce). As a result, we have reason to
believe that our sample is fairly typical and avoids major biases.
In our estimates of the total social costs we include both serious problem gamblers--
pathological gamblers--and less serious problem gamblers--problem gamblers. We do
not include costs imposed upon society from gamblers whose behaviors result in unpaid
debts, thefts, welfare costs, etc., even though the gamblers remain in control of their
gambling behaviors to a degree that they may not be clinically defined as problem
gamblers. All in all, measurement difficulties suggest the conservative approach used
here.
IV. DEMOGRAPHICS OF THE G.A. RESPONDENTS
a. Gender
In the recent past, GA meetings were almost the exclusive domains of male problem
gamblers. However, as the gambling industry has spread beyond the realm of illegal
bookies, clandestine craps games, poker games, and racetracks, and beyond the legal
casino enclave of Nevada, gambling has engulfed many female customers, Lotteries,
free standing slot and video machines, bingo games, and in Las Vegas "friendly"
neighborhood casinos can be very welcoming for female participants.
.
The prevalence study for the National Gambling Impact Study Commission found that
men still were more likely to become pathological gamblers, but the rates for women
were at least half as high, Whereas, the Wisconsin study of 1996 found that only 28% of
the G.A. respondents were women, the Connecticut study of 1998 found only 22%
women, and the South Carolina survey 33% women; we found that in Las Vegas, 45 of
93 (48.4%) of the G.A. respondents were women. In Las Vegas, the problem gambling
threshold knows no glass ceiling.
b. Ethnicity.
One might conclude that the close-knit nature of GA meetings creates barriers to
participation across ethnic groups. Though several minority groups do have higher
prevalence rates for pathological gambling, and the Las Vegas Metropolitan Area has
minority populations comprising about one third of the area's population, only 18 of 93
respondents were either Hispanic or non-white. Nonetheless this 19.4% is considerably
greater than the 9% in found in South Carolina, the 3% in Wisconsin or the 4% found in
Connecticut.
White 75 (80.6%)
Black 6 (6,5%)
Hispanic 5 (5.4%)
Asian-Pacific 4 (4.3%)
Other 2 (2.2%)
.
Native American 1 (1,1%)
c. Age
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada by... Page 6 of 20
.
Of the respondents, 93 reported their current age. The median age of the respondents
was 46, while the mean age was 47.25.
Under 3010 (10.8%)
31-4017(18.3%)
41-5032 (34.4%)
51-6020 (21,5%)
61-7011 (11.8%)
Over 703 (3.2%)
The median age of respondents in Wisconsin was 43 years, in Connecticut 47 years,
and in the South Carolina the average age was 44,63 years.
d. Marital status and children
Only a minority (28 of 92, or 30.1 %) of the G.A. respondents was currently married. Six
were "living with someone" and 29 were single or widowed. Twenty-nine (31.5%) were
either divorced or separated, Of these, 19 (or 65.5%) indicated that gambling was the
cause of their break-up. (In Wisconsin 70%, Connecticut 55%, and South Carolina 48%
of separated respondents gave gambling as the reason for their break-up.)
.
The average number of children was 1,8. Only 21 still had children living with them. Of
the respondents who had children, 31.8% had children living with them.
single 25 (27,2%)
married 28 (30.4%)
cohabitating 6 (6,5%)
divorced 26 (28.3%)
separated 3 (3.3%)
widowed 4 (4.3%)
f. Education
.
The Nevada population displays a low level of educational achievement--among the
lowest level of all the states. The national survey found that the prevalence rates for
pathological gambling were higher among those with lower education attainment.
Nonetheless, the Las Vegas G.A. respondents reflect relatively high levels of education,
Thirty had college degrees, almost twice the portion in the population as a whole. Only 5 '
had not graduated from high school. In comparison the G.A. members surveyed in other
states had lower educational levels, In Connecticut 22% were college graduates, in
South Carolina 20%, and in Wisconsin only 13%.
College degrees 30 (31,6%)
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada by... Page 7 of 20
Attended College 47 (49,5%)
.
High School grad 13 (13,7%)
Not H.S. grad 5 (5.3%)
g. Employment
Occupations were not ascertained in the survey; however, 29 of 91 (31.9%) indicated
that they did work in the casino gaming industry. This reflects the fact that about one
third of the employment base in southern Nevada consists of jobs in the casino hotel
resort economy. It also suggests a major occupational hazard for employees who are
constantly exposed to gambling activity,
Surveys from the other venues showed fairly even distributions between white collar
(including professionals) and blue collar occupations.
h. Income
The median income was $52,000. This was in the same range as the median household
income in Connecticut and South Carolina, but less than above that in Wisconsin,
Assigning a mid-point value for the top income category, we can estimate that the
average income for the group was $54,495.
V. THE GAMBLING CAREERS OF THE RESPONDENTS
.
The survey identified the age that the respondent's gaming began, when gaming began
on a weekly basis, when the individual started borrowing money to gamble, when
gambling problems began. They were also asked how long they were participating in
G.A. From the information and the age of the individual we can gain a notion of the
duration of their troubled gambling career,
Range Median Mean
Age Now 23-86 46 47.25
Age Started Gaming 4-90 21 26.82
Age Weekly Gaming 6-74 30 31.84
Age Borrowing 13-71 3033.43
Age Problem Start 8-67 30 34.12
Years in GA 0-31 0.62.31
.
Whereas an assessment of the averages and medians above show problem gambling
careers to be 10.8 years or 15.4 years, they are found to be lower when the information
is calculated for each individual first, and from that number averages and medians are
determined. When the data are analyzed for individuals, we find that the average
problem gambling career is 10,2 years and the median career is 7.6 years.
The median and mean careers respectively were 2.96 and 6.66 years in Wisconsin, in
Connecticut the median was 9 years, and in South Carolina 8,2 years. We did expect
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada by... Page 8 of20
.
longer durations of careers in Las Vegas as the gaming industry has been well-
established and entrenched in the local society for many more years than elsewhere.
We can also suggest that gambling careers would be 'a longer in Connecticut than in
Wisconsin or South Carolina as forms of gambling have been openly present there for
decades, but came to the latter states only in the late 1980s. The fact that gambling
activity is well ingrained into the fabric 0'[ Las Vegas society means that there are
considerably more incentives for problem gamblers to be in denial ("it's normal,
everyone does it all the time"). Commercial practices are also more tolerant of problem
gambling ("Have problem with credit--'no problem' we can help you").
We will return to the concept of the problem gambling career later when we seek to
estimate annualized social costs of patholo~lical gambling from data on lifetime costs.
VI. WHERE THEY PLAY AND WHAT THEY PLAY
We addressed the GA members regarding the locations of their play. We focused upon
opportunities for gambling that are present in the Las Vegas area. We also looked at the
types of games played and the extent to which they precipitated problems among the
respondents,. .
Comparisons with other studies are limited as Nevada has a unique array of easily
accessible gaming opportunities.
.
The locals-oriented (neighborhood) casino was the favorite venue of the respondents. A
big majority, about 83% gambled at least: two times a week at local casinos, 51 %
frequented major resort casinos, 44% at bars and taverns, 41 %convenience stores,
40% at supermarket casinos.
There was no doubt but that the video poke-r machines were the game of choice for the
G.A. members. Eighty-eight responded to the specific inquiry. Over two-thirds found the
machines to constitute "serious" problems for them. Only 14 (15.9%) found no problem
at all with the machines.
In descending order, 49.4% found other machines (i.e., traditional type slot machines),
38.7% table games, 23.0% sports betting to cause serious problems for them; and,
15.6% found serious problems in other gaming activities.
VII. VOLUME OF GAMBLING LOSSES, DEBTS AND SOURCES OF FUNDS
a, Losses
.
Ninety of the respondents estimated the amount of money that they had lost in their
lifetimes at gambling activity. Four suggested they had lost less than $1000 each, 18
estimated losses to be in excess of $250,000, with one of these saying the losses
exceeded one million dollars, The median loss was between $50,000 and $100,000,
and the mean loss was $112,400. The mean was determined by assigning $250,000 as
the mid value for the category of $250,000 and above. The $112,400 loss is a very high
loss amount compared to the other surveys, The median losses were $82,500 in
Connecticut, $45,000 in Wisconsin, and the trimmed average losses were $79,434 in
South Carolina.
b. Sources
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada by... Page 9 of 20
.
The troubled gambler will typically seek funds from others only when his or her personal
funds have been exhausted. After other legitimate sources are tapped, the problem
gambler may consider seeking money from illegal sources. Over two-thirds of other
respondents indicated that they had to go to other people for gambling money. Sources
included:
Spouse 52 (57.7%)
Children 27 (30.0%)
Sold Personal Property 54 (60.0%)
Cashed in Securities 50 (55.5%)
Passed bad checks 57 (63.3%)
Used Casino Credit 31 (34.4%)
Used Bookies 15 (16.7%)
It can be added that 17 of 92 (18,5%) gambled with social security funds. However, as
only 14 were over 60 in age, and a small number of younger respondents could have
been on social security disability, it is likely that all receiving social security monies did
use these for gambling activities.
.
In other surveys almost all the gamblers utilized credit cards, The source was not
included on this survey as it was assumed that Nevada gamers heavily used them as
well.
c, Debts
The gambling activity did cause major financial problems for the respondents. All but
five indicated the scope of their debts when they joined GA. Specifically, because of
gambling activity five owed less than $1000, and six owed more than $250,000 each,
with one owing $1.6 million. The median debt was $24,500, while the mean debt was
$78,305, A trimmed mean took 5% off the extreme ends of the continuum of
respondents. That mean was $57,160. The debt figures were higher than those found
elsewhere: the average debts were $38,664 in Wisconsin and $29,586 in South
Carolina.
d. Bankruptcies
Gambling led 45.4% of the respondents to bankruptcy court for protection from their
creditors. These 44 had median debts of $38,750, while their average debt was
$121,646. A trimmed mean found debts of this group to be $85,551.
The other studies did not find bankruptcy rates nearly as high as in Las Vegas. In both
Wisconsin and Connecticut the rates were 23%; 26% of the respondents in the South
Carolina study had gone through bankruptcy proceedings.
e, Creditors in Court
.
In addition to bankruptcy court, the gambIE~rs' debts also resulted in other legal actions.
Fourteen were sued in courts by others seeking repayment of debts. Eleven were sued
a single time, one twice, and two three-or-more times for a total of at least 19 suits,
[50'] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page lO of 20
.
(Respondents were asked if they were sued one, two, or three-or-more times.)
VIII. THEFTS AND OTHER ILLEGAL ACTIVITIES
a. Thefts
When pathological gamblers run out of legitimate sources of money they consider illegal
sources. Starting close at hand, they pass bad checks, We found that 63.3% wrote such
checks. They also look for money in the workplace. Also, 30.1% admitted to stealing
from the workplace in order to gamble or pay gambling debts. This is about the same
portion who stole from the workplace in other surveys: 31,7% in Wisconsin, 37,1 % in
South Carolina, and 40,7% in Connecticut.
A majority, 50,6%, of the respondents indicated that they had stolen money or things
and used it to gamble or to pay gambling-related deaths, (In Wisconsin, 49% claimed to
have stolen for gambling, 53% in South Carolina, and 55% in Connecticut). Whereas
many in Las Vegas admitted to stealing only small amounts of money (29 said they
stole less than $1000), a number had very large thefts as well. Fourteen indicated
stealing over $10,000, The largest amount stolen by a single gambler was well in
excess of a million dollars. That theft caused the survey average to be $91,696,
however, using the trimming methodology, we find a trimmed average of $7277. In
Wisconsin the average was $5738, $8487 in South Carolina, and $22,533 in
Connecticut.
b. Criminal Justice System Activity
.
The thefts reported by a majority of respondents (43 of 85) certainly led to many police
investigations, However, the respondents were quite adept at avoiding the criminal law
consequences of much of their illicit activity.
Only 9 of 93 were ever arrested for any gambling-related activity. The nine had 13 (or
more) arrests, resulting in nine (or more) trials. Seven were convicted nine times (six
one time each, and one twice). Five were then incarcerated, with average sentences of
just over three months. Spread over the 93 respondents, the average sentences
were .17 months. Ten (of 96) were also placed upon probation for gambling-related
offenses. Of all, we can say each endured an incidence of probation of .10.
IX. EMPLOYMENT IMPACTS
Problem gamblers are not as productive as others in the workplace. The Politizer study
suggested that a pathological gambler was only 20% effective in the workplace, While
we will not attempt to assess financial costs for reduced productivity for those who were
on the job, such costs do exist and must not be considered nonexistent because they
are not calculated here. What information we can use for cost calculations comes from
answers to our survey questions.
We found that 50 of 89 (56.2%) had lost time by absenting themselves from the
workplace in order to gamble or to participate in activities related to their gambling (i.e.,
seeking funds with which to gamble or to pay gambling related debts). The 50 reported
missing an average of 17.22 hours of work each month due to gambling. Averaged over
all 89 this represents a loss of 9.67 hours a month, or 116,1 hours a year.
.
Comparable respondents from Wisconsin led to assessments of 7,5 hours lost per
month, in Connecticut 9.8 hours, and in South Carolina 23,9 hours.
Twenty-two of 96 (22.9%) respondents quit work because of gambling activity and/or
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page 11 of20
.
gambling problems. These 22 averaged times of unemployment averaging 18,77
months. Spreading the idle months over the 96, we find lost employment averaging 4.3
months due to gambling. An additional 21 of 89 (24.0%) responded that they had been
fired due to their gambling activity. These 21 averaged 11.57 months of unemployment
as a result; spread across the 89, this represents an average loss of 2.73 months of
work because of discharges from the workplace. The portion of respondents who
separated (by choice or otherwise) from work due to their gambling was larger in
Nevada than in the other jurisdictions, In Wisconsin 21 of 98 lost jobs, in South
Carolina, 19 of 70, and in Connecticut, 25 of 112.
X. WELFARE SERVICE IMPACTS
Whereas the employment impacts seem to be greater among problem gamblers in
Nevada, the social consequences of welfare services are less, Only 3 turned to general
welfare because of gambling problems, and 5 received food stamps as a result of
gambling. (Of 112 respondents in Connecticut, 7 took welfare payments and 10 food
stamps because of gambling; in Wisconsin one of 98 went on welfare and 3 took food
stamps, while in South Carolina, 2 of 70 went on welfare and 4 took food stamps.)
XI. TREATMENT COSTS
A smaller portion of Las Vegas respondents sought professional help as compared with
gamblers in the Connecticut, Wisconsin, and South Carolina studies. Thirteen of 87
were hospitalized, while 20 had out patient care.
.
Though we specifically asked for information regarding treatment related to gambling
problems, the respondents here, as elsewhere, indicated that they had a wide range of
physical and emotional problems. In this survey 58 of 88 (65.9%) indicated that they
suffered from other addictions.
Twenty-nine (of 70) indicated treatment costs averaged $7,022. Spread over all 87 (who
responded to the treatment question), the average cost is $2,340.
Of 33, 7 indicated they paid the costs "out of pocket." Ten were fully covered by
insurance, while 8 used a combination of personal payments and insurance. Eight said
they had not paid their treatment bills.
XII. SUICIDE
The survey found that 60 of 91 (65,9%) respondents had planned suicides as a result of
gambling. This compares with 71 % in South Carolina, 55% in Wisconsin, and 44% in
Connecticut.
Twenty-six (of 94, or 27.7%) indicated that they had made actual attempts to take their
own lives. In South Carolina the portion was 30%, in Wisconsin 24%, and Connecticut
17%.
XIII. DEVELOPING A COST PROFILE
.
We utilize the information above to develop a social cost profile for a single-problem
gambler. In doing so we are mindful that some costs of the gambler's activity are
absorbed by the gambler and his or her family, others are imposed (against the will) of
others, while some are imposed upon governments, and yet these and other costs may
result in general losses for the full economy of a society, As we indicate cost values we
will seek to explicitly label the "victim" of the financial loss due to the gambler's activity.
In general, we use the term "social cost" to indicate costs that are imposed upon people
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page 12 of 20
.
other than the gambler and his or her family, that is, people who do not participate in the
gambling process. As a result of the gambling these people receive one of two results
from the wagers made--they "break-even" or they lose. Given the odds structure of the
games involved, inevitable, they lose.
There are many very real costs that are not included in the cost profile presented. There
is a very real cost to employers and society when work productivity of problem gamblers
decreases. It happens, and only because it would be very difficult (and costly) to
develop methodologies to capture the cost, we exclude the cost from analysis. Students
of pathological gambling find that major life costs are generated because of the
problems of spouses of troubled gamblers. These include workplace costs such as
missed work and lost productivity, as well as health and other treatment costs. Spouses
who are "enablers" may also participate in illicit fund raising to deal with situations
brought about by the gamblers. There are also major costs to children, and these costs
can be relayed onto society as a whole through their dysfunctional behaviors. When a
pathological gambler "steals" from a child's college fund, and education is stifled,
society is also a big loser. Again, these matters defy easy interpretations in terms of
cost, and therefore will not be found in the numbers presented.
For these reasons, when we do come up with real numbers, we can suggest that the
numbers are conservative and very much below what the real social costs are. We will
also purposely use conservative methodologies to assess the costs that are identified.
a, Annualizing Costs
.
We will use a factor of four in determining the annualized social costs. The initial
Wisconsin study found that a problem gambling career was almost three years in
duration. The time represents that span between the beginning of participating in GA
and the time that the respondent indicated that "problems" with gambling began. In
Connecticut the span identified in the survey was approximately 9 years, though lifetime
costs were divided by three. Also, without assessing the time span independently, the
study conducted on behalf of the National Gambling Impact Study Commission also
used the three-year figure and cited the Wisconsin study as the basis for doing so. In
South Carolina, the "career" was found to be approximately 8.2 years and that number
was used as the divisor. However, we arE~ persuaded that a shorter career length for
"pathological" is in order. Troubled gambling is identified as a progressive disease. It
grows worse over time. Gambling increases in volume, so does the gambler's resort to
seizing funds from others, and in turn, undergoing treatment for gambling and related
problems. Hence, in later years of a problem career, costs to society must increas~.
Indeed, while a problem gambling career may last 3, 8, or 9 years, the time span of
pathological gambling within this period is undoubtedly much shorter. Though we may
be inclined to simply utilize the three-year factor found in most other studies, here we
will look at the median and mean problem gambling careers both in the aggregate and
for the individual. We presented two numbers, a mean of 10.2 years, and a median of
7.6 years. The midvalue of the two figures is 8,9 years. During a majority of these years
the problem gambler will cope for the most part on the basis of his or her own
resources. Considering that the pathological-problem phase will not be as much as half
this time, we will very CDn'servatively divide the full social costs determined here by a
favor of four so that they can be annualized.
b, Cost Parameters
.
For purposes of clarity and comparisons, and in recognition of past efforts to calculate
specific costs for matters such as arrests and cost appearances, we will simply use the
costs calculations identified in the 1996 Wisconsin study. Again, this will yield
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page 13 of20
.
conservative results as both inflation and other factors have without questions caused
these costs to rise today. Here they will simply be identified and the Wisconsin study will
be cited as a source for finding the original source of the information and the
methodology for developing the cost factors.
c. Employment Costs
The missed work of the gamblers constitutes an average over eighty respondents of
$116,1 hours a year. This factor is already presented in an annualized format, so we
assume that the cost factor endured throughout the pathological gambler's career. At
the same $15 an hour used in the earlier studies we find an annual social cost (in this
case employers and their customers) to be $1742,
Productivity losses attend those fired and quitting their jobs. Those quitting (21 of 86
respondents) were unemployed an average of 18,52 months; spread over the 86 this
represents 4.5 months of lost work for each pathological gambler. Valued at $15 an
hour or $2500 a month, this represents a social loss of work valued at $11,250. The
annualized loss of productivity for society is $2813. These are costs to the~ull society.
Those fired (16 of 81) were out of work for an average of 11.53 months, Spread over 81
gamblers, this represents a loss of work averaging 2,28 months, valued at $5693 each
for their careers. The annualized cost of lost productivity due to firings because of
gambling is $1423.
.
Eight of the fired workers also secured unemployment compensation. Over the average
of 2.28 months for the eight at $732 a month this is a career social cost (to all society)
per fired gambler of $1669, Spread over the 81, the amount averages $164. Annualized
this social cost per gambler is $41,
d. Debts, Bankruptcies, and Civil Suits.
Bad debts represent costs imposed upon other people. If the other people are
businesses (e.g., credit card companies), the costs are spread out across society,
directly or indirectly. Hence when you do business with a merchant that has filed to
collect an obligation owed by a gambler, you pay part of the cost-even though you did
not enjoy the excitement of the poker machine, nor did you receive a "chance" to win
the casino's "big prize."
Though almost all these gamblers probably failed to pay debts to some degree, we
have calculated debts by looking only at the gamblers who went through bankruptcy
proceedings. These averaged debts of $121,646; however, the trimmed average debt
was $85,551. Spread over 94 respondents who reported debts, we find an average
(trimmed) debt of $38.225. The annualized social cost of these bad debts is $9,556.
.
There are also social costs that are imposed upon governments as a result of court
actions. Additionally, the gambler incurs legal costs that represent lost resources for
society, that is, they are resources not available for positive needs. There were 44
bankruptcy actions and were also 19 civil suits relating to debts. (Reported by 93
respondents), The earlier study found that each federal court action costs $7500,
Considering that these actions may not be as complicated or long enduring as some
others, we assign a 50% cost factor of $3750 for each of the 63 cases. This represents
an average cost spread over the 93 respondents of $2,540, which is annualized to
$635.
The 63 court actions at legal fees of $2,500 each, take another average of $1676 from
the gamblers (annualized to $418), money that could be better spent on positive things
in the economy.
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page 14 of20
.
The 19 divorce actions (among 92 respondents) that are put at the feet of gambling
impose court costs on society of another $774 per gambler, or an annualized cost of
$194. They impose legal costs on the gambler of $516, costs annualized to be $129.
e. Thefts and Criminal System Actions
Thefts are social costs. (The attached note discussed the rationale for considering these
to be costs imposed upon the full society.) The average (trimmed) costs of thefts as
reported by the respondents (an average spread over all those responding to the
question of thefts) were $7,277. This cost can be annualized to $1,819.
The thirteen arrests for 95 respondents cost society $2,900 each (according to the study
performed for the National Gambling Impact Study Commission), This represents a
career cost of $397, or an annualized cost of $99.
Nine criminal trials at a cost of $3,750 each represent another average gambler cost of
$355, a cost to society annualized to $89.
Sixteen-months incarceration costs $32,000, or an average of $337 per gambler career,
or an annualized cost of $84,
Ten probation cases (among 96 GA respondents) cost society $9,600 each or an
average of $1 ,000 per gambler for the career, or annualized to be a cost of $250.
Legal fees of $2,500 per trial result in average costs of $237 per gambler, annualized to
$59.
.
f. Social Costs of Treatment
The twenty-seven gamblers who undertook to have professional treatment for their
problems, either in a hospital or as an outpatient, spent an average of $7,022 each on
the treatment. Spread over 87 respondents answering questions about treatment, we
find an average career treatment cost to be of $2,179, and an annual cost $545. One-
fifth of this cost was paid directly by the gambler. One-fourth was not paid at all, making
it a "social cost," ($136), while 55% was paid by insurance providers. Of the latter
amount, we will assign one-half to social costs, or $150. Hence we find an annual social
cost of treatment (to others) to be $286.
g. Welfare Services.
We assume that those taking welfare and food stamp provisions do so for two years
each. Three of 89 took welfare. The average payment identified in the Wisconsin study
was $460 a month, making the total costs for three people for two years to be $33,120.
Spread over the 89 respondents, this represents an average career social cost of $372.
Annualized, this is a cost of $93.
.
Food stamp costs were set at $2,000 a year. Five (of 87) gamblers impose a two-year
cost of $20,000 as a result. This represents a career average cost of $230, or an
annualized cost of $57.
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southem Nevada... Page 15 of20
.
h. Estimated Average Social Cost per Problem Gambler by Item
TOTALS (Costs to Society)
Employment Costs ($6,017)
"
Missed Work........................ $1,740
Productivity Losses (quit jobs).... 2,813
Fired from Work (productivity lost). 1,423
Unemployment Compensation.......... 41
Bad Debts and Civil Court ($10,291)
Bankruptcy Debt Loss............... $9,556
.
Civil Court Costs (bankruptcy/debUdivorce)....... 735
Criminal Justice System ($2,341)
Theft...,."., .....,......, ".,.... $1,819
Arrests...,.,......,......."., ,... 99
Trials............................. 89
Incarceration...,....., ....... ..... 84
Probation...........".........,... $250
Treatment and Social Services ($436)
Treatment Costs.................... $286
Welfare.....................:...... 93
Food Stamps........................ 57
Total Social Cost.........,..................... $19,085
.
XIII. USING SURVEY TOTALS TO ESTIMATE COSTS FOR ALL SOUTHERN
NEVADA
Earlier in this report we suggested that the gamblers we surveyed at GA meetings were
likely to be typical of pathological gamblers in southern Nevada. On the one hand, the
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page 16 of 20
.
more desperate pathological gambler may be drawn to seek treatment such as can be
offered in the group counseling atmosphere of GA, it can also be surmised that the most
severe pathological gamblers would avoid such treatment and other treatment. The
pathological gambling disease does incorporate a strong notion of denial that precludes
treatment. So in a sense we believe that some severe gamblers come, and others do
not, hence balancing the survey results toward the typical.
The Politizer group looked at pathological gamblers who were having treatment at
inpatient treatment centers. These were found to have imposed social costs of $61 ,000
a year onto other people.
However, the group considered these gamblers to be at an extreme level. They
assessed others not in treatment and found they produced social costs of $26,000
annually. They concluded that the typical gambler out of treatment cost society amounts
equaling 43% of those in treatment. Though we might agree that ones in actual hospital
settings for treatment for pathological gambling would have more severe problems, we
would think those at GA might not be so severe. Still, using social-cost figures which are
43% of t~ose we found in our survey of GA members, the pathological gambler
produces social costs of $8,207 per year.
The cost of pathological gambling for the Southern Nevada community is then
determined by how many pathological gamblers there are in the community. The
population of Southern Nevada (Clark County) for 2002 is estimated to be 1,578,322. Of
this number, 69.8% are over 21, and therefore allowed to gamble in the casinos of Las
Vegas. This adult population is 1,102,033.
.
The number of pathological gamblers would range from a low of
19,836, based upon a doubling of the National study's prevalence rate of 0,9% of
pathological gamblers in society, to a high of 38,571 based upon Volberg's study of
Nevada which found that 3,5% of the adults were probable pathological gamblers at the
current time.
The range of social costs for Clark County derived from the activities of these
pathological gamblers would, therefore, range from a low of $162,794,052 per year to a
high of $316,552,197 per year.
These cost ranges ignore costs imposed upon society by troubled gamblers who do not
meet the full definition of "pathological" gambler, as well as costs imposed upon others
by non-problem gamblers as a result of their gambling activity,
The Volberg study found that 2.9% of the Nevada adults were current problem
gamblers. This figure yields a number equaling 31,959 adults for Southern Nevada,
which will be our low estimate, This corresponds closely to the National Study's 1,6%
which would be doubled in a region close to casinos, yielding 3.2% or 35,265, which
becomes our high estimate for the number of problem gamblers in Clark County,
Nevada.
The National Study presented partial information on social costs, and in doing so they
suggested that social costs emanating from problem gamblers amounted to 53% of the
value of those from pathological gamblers. Thus, we can set the social cost of a
problem gambler in Clark County to be $4,350 per year,
.
Projected to the full population we, thereforE~, find an additional social cost ranging from
a low estimate of$139,021,463 (based upon 2.9%) to a high estimate (based upon
3,2%)of $153,402,750 coming from abusive gambling in Southern Nevada.
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page 17 of20
.
The overall range of social costs, therefore, can be estimated to be from a low of
$301,815,515 to a high of $469,954,947.
XIV, DISCUSSION
Faced with estimates of the social cost of problem gambling in Las Vegas ranging from
$273 to $413 per adult per year suggests an important problem, The magnitude of
these costs lead one to think about ways to address such concerns, Of course, one
option available is to continue to leave individuals and families to address the problem
as best as they can. As such, this is the baseline case to compare with other
alternatives. Yet, the magnitude of the estimates, even if preliminary and subject to
greater refinement, would suggest that proactive alternatives are likely to yield a better
outcome and that inaction could create a backlash against the legalized gambling
industry.
Leaving individuals to address the social problem of gambling invites possible litigation
similar to the experience of the tobacco industry, The possibility of an outcome similar to
tobacco, however, would surely place the gaming industry at risk, particular, given the
magnitude of the recent tobacco settlements. As a result, one might reasonably expect
the industry to support efforts to bring action on the issue, two proactive positions come
quickly to mind.
.
One possible proactive approach would be for public support of facilities and staff. In the
current environment of many state and local governments struggling with budget
shortfalls, however, one might expect that tllis approach to be difficult to accomplish,
Furthermore, the gambling industry, similar to alcohol and tobacco, pay excise taxes,
also referred to as externalities. These taxes have not, however, historically been tied to
the social costs of these activities, rather they have been used as general operating
monies, Thus, it might not be easy for the problem gambling problem to receive
consideration without further increases in rE~venue. Earmarking of additional taxes paid
by the industry for treatment might be one possible outcome.
Another possible proactive approach to addressing the concern of the problem gambler
would be self-assessment of the industry. This approach would be similar to the use of
self-assessment in the travel and tourism industry for local convention and tourism
authorities. That is, there is an assessment earmarked to support treatment that is
related to a firms level of activity. Again, the self-assessment approach works best when
an industry finds that there is consensual support for the collective effort to establish a
fund-sharing arrangement for eduacation, warnings, intervention, and treatment.
Whether private, public, or public-private approaches are forthcoming is unknown;
nevertheless, the magnitude of the problem gambler in Las Vegas, a mature gaming
venue, and the spread of gambling through out the U,S. suggests that the social cost of
gambling will become larger in the future and that the concern over these costs will also
increase.
References
American Gaming Association, 2002. The Economic Benefits of Gaming.
.
Cabot, Anthony, and William N. Thompson, Andrew Tottenham, and Carl Braunlich,
1999, International Casinos Law. Reno: Institute of Gaming Studies, University of
Nevada,
Grinols, Earl. 2003 (forthcoming). The House Rules: Cost Benefit Analyses of
Gambling,
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page 18 of20
.
Kindt, John W. 1994. Testimony to U. S. House of Representatives
Committee on Small Business. September 21.
Lesieur, Henry. 1984. The Chase: Career of the Compulsive Gambler, Cambridge, MA:
Schenkman.
Lesieur, Henry and S. B. Blume. 1987. "The South Oaks Gambling Screen. American
Journal of Psychiatry, V. 144: 1184-1188.
Lesieur, Henry. 1998. "Costs and Treatment of Pathological Gambling," Annals of the
American Academy of Political and Social Science, March. Pp. 153-171.
National Gambling Impact Study Commission. 1999. Final Report.
Nevada Gaming Commission. 2002. Nevada Gaming Avstract. Carson City: State of
Nevada.
Politzer, Robert M., James S. Morrow, and Sandra B. Leavey. 1981. "Report on the
Societal Cost of Pathological Gambling and the Cost-Benefit/Effectiveness of
Treatment." Paper presented to the Fifth National Conference on Gambling and Risk
Taking. October 22,
Samuelson, Paul. 1970. Economics. New York: McGraw-HilI.
.
Thompson, William N., Ricardo Gazel, and Dan Rickman. 1996. The Social Costs of
Gambling in Wisconsin, Wisconsin Policy Research Institute.
Thompson, William N., Ricardo Gazel, and Dan Rickman. 1999. "Social Costs of
Gambling: A Comparative Study of Nutmeg and Cheese State Gamblers." Gaming
Research and Review, V. 5, no. 1. Pp. 1-16.
Thompson, William N" and Frank Quinn, 1999. "An Economic Analysis of Machine
Gambling in South Carolina." Report to the Educational Foundation of the South
Carolina Policy Council. Columbia, May 19.
United States Statistical Abstract.
Volberg, Rachel. 2002. Gambling and Problem Gambling in Nevada. Report to the
Nevada Department of Human Resources,
WEFA Group. 1997. A Study Concerning the Effects of Legalized Gambling on the
Citizens of the State of Connecticut. Eddyston, PA: WEFA (June). (Co-authored by
Henry Lesieur and William Thompson),
AJl-p-endix 1: Theft is a Social Cost--Bigger Than We May Have Thought
.
Several studies of cost impacts of gambling have been criticized for including the value of stolen
property by pathological gamblers as a negative social cost. The studies have considered the
worth of the property as a social cost while the critics have indicated that the cost is merely a
transfer of wealth from one person (legitimate owner) to another (new illegitimate owner). The
critics agree with the inclusion of the costs of police investigations and court deliberations in
property crimes related to gambling, but they assert that the value of the property stolen is still in
the society, and therefore can't be considered a factor decreasing the wealth of the society, As a
,.
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page 19 of20
.
person involved in three gambling social cost studies, I wish to offer a dissent to the criticism,
albeit an admission as well.
In a Wisconsin study my colleagues and I found that each pathological gambler imposed annual
costs (while they were active pathological gamblers) of $1733 because of property they stole. A
Connecticut study found $2503, while a social impact study in South Carolina discovered costs of
$1035 due to the value of stolen property. A study by Henry Lesieur and Chris Anderson found
career costs of thefts by each pathological gambler to be $76,309 while Westphal, Johnson, and
Stevens found Louisiana theft costs to be annualized at $1929. The findings were based upon
surveys of gamblers in treatment. While the methodologies utilized certainly can be improved, the
crux of the criticism of the studies has not focused upon methods but upon the theoretical
decision to include the costs in and of themselves,
I would offer the following caveat regarding the use of these figures in overall cost impacts per
pathological gambler. The costs are not exact, and the costs do involve some transfer values--
that is some of the value of the stolen property may indeed remain in the community, and hence,
should not be considered in assessing the overall loss of value (wealth) for a community (society),
That being said, I can offer that the treatment of the survey results is such that we are leaving out
many social costs (loss of wealth to the entire society) that are attached to the stealing of the
property, so that the figures presented must only be seen as a surrogate for the full costs, and as
such they probably grossly undervalue the social cost of thefts that are imposed by the
pathological gamblers. It is time to elaborate why values of stolen property should be counted as
social costs of gambling. Hopefully at some future time the dollar figures can be made more
exact, and the uncounted costs referred to below can be included with precise dollar values.
.
Let us look directly at the value of the items stolen by pathological gamblers. The items may be
cash, instruments representing cash value, or tangible items of property, Let's look at a tangible
item of property, For instance, a new television in a person's home. Why a new television?
Simply, because the thief has been drawn to the item because he has "cased" out a
neighborhood, and he has seen a television box in the discarded trash by a home--ergo, he
reasons the home must include the valuable item. The store cost of the television was $500, But
any understanding of micro economics would say to Buyer Bob it is worth more, otherwise he
would not have made the purchase. We will say the television has a value of $600 to Bob, the
legal owner.
What is it worth after it has been stolen? Consider how much it is worth to the thief. Not $600. Not
$500. If it had that worth the thief would have gone to the store and purchased a similar type
television set. No. It is worth--well, the time necessary to case the neighborhood (two hours), the
time to wait for the owner to leave the house (two hours), the time to purchase burglar tools (one
hour), the value of the use of the tools ($20--they can be used again). The cost of the use of a
vehicle and gasoline ($1 O)--perhaps a total of $75. Now these are also social costs, because in
the absence of the theft by the pathological gambler, they would not have been expended, The
thief also incurs a risk cost of detection and police action, but to the thief it is very low because he
would not attempt the theft if he thought it was a large cost. So in the first instance, we can see
that the television has decreased in value by perhaps $500. Now consider the many transactions
needed to get the television back into active use, The thief did not steal it in order to watch it. He
sells it to a Fence Frank for $100, taking his profit and running back to the casino (or paying off
his debt to Vinnie). The fence rents his space at the flea market and sells the set to Dealseeking
Dan for $150. Is it worth $600 again? Hardly, when it breaks down Dan has no guarantee,
Moreover, he has to worry that the repairman will see Buyer Bob's 10 mark on the set and report it
to the police (the police have a report describing the set, and they distribute the description to
repair people). He may also lose sleep over the thought. He may come to think it is worth much
less than he paid for it. The social cost of the transaction, the time and energy necessary for both
Dan and Frank to go to the flea market is a loss to society.
.
By now the social costs are close to the cost of the set as it was originally sold. But there are
more social costs. Buyer Bob is now afraid to go home sometimes, and afraid to leave at other
times. He has lost security and freedom. Social costs. His doctor has also decided that he needs
medicine and therapy--more social costs, He buys a new set of locks and window alarms for his
house. Social costs. He buys an expensive Doberman Pincher--more social costs, He convinces
the city to assign extra police to his neighborhood. More costs.
Of course, not all these costs would attend a robbery of negotiable cash or checks, but many of
the costs would, as some instruments cannot be easily sold, and if there is a home invasion or a
[50] 2003 - Beyond the Limits of Recreation: the social costs of gambling in Southern Nevada... Page 20 of20
.
personal robbery, the fear factor--which carries costs-has been imposed on the entire society,
Now of course, the pathological gambler may have multiple addictions. We recognized that, and
in all the studies. subjects were specifically asked the cost of the goods the stole because of their
gamblinq problem,
In the future we will develop better methodologies, which may accurately capture many of the now
unmeasured social costs that attend thefts by pathological gamblers. For now we take comfort
that we have not overestimated the extent of the costs by using the value of the goods stolen as a
surrogate value for all the social costs. When we assess them all in an acceptable scientific
manner we will find that the costs we have used in our studies will be multiplied many times over.
We have also been criticized for including debts as a social costs--well, that's a matter for another
essay.
See: D. M, Walker and A. H. Barnet, The Social Costs of Gambling: An Economic Perspective,"
15 Journal of Gambling Studies (1999); Michael Walker, The Psychology of Gambling (1992);
Thompson, Gazel and Rickman (1996) (see above); WEFA (Thompson and Lesieur)(see above);
H, Lesieur and C. Anderson, Results of a Survey of Gamblers Anonymous Members (1995); W.
N. Thompson and F. Quinn, The Social Costs of Machine Gambling In South Carolina (2000);
J.Westphal, L.J. Johnson, and L. Stevens, Estimating Social Costs of Gambling In Louisiana for
1998 (1999).
This article, By William N. Thompson, appeared in the Report on Problem Gamblinq
(December/January 2001), pp. 44-46
NCALG Information Librarv
~t:;;'~~~;~'~:J:=~~.:~~:-~'.r.-,A::;7?l~1~1
.
Why]. - Who Are We? - You Can Help - Donate - More Info - Contact NCALG
For information about NCALG, email ncalq@vcn.com or call us at
the National Information Center 1.,800-664-2680/ FAX (307) 587-8082
100 Maryland Avenue NE, Room 311, Washington, DC, 20002
Web Design by Anderson Research -- Copyright 2003 All Rights Reserved.
.
TAB
"2 "
.
.
.
Casino Watch-Florida Casinos Report
Economic and Social Costs
of Gambling
r i{!}i'itflJ ~ If:~'l H [1.11 iQIiLI
ur']/~ 1 ~
Back to prt:\'iolls D3QC
Home
Casinos in Florida
An analysis of the Economic and Social Impacts
Prepared by:
The Executive Office of the Governor
Office of Planning and Budgeting
The Capitol
Tallahassee, FL 32399-0001
EXECUTIVE SUMMARY
3. Recurring sales tax revenues would experience a net decrease of at least
$84.7 million as Floridians diverted some of their existing taxable spending to
casinos.
4. Crime and social costs attributable to casinos would total at least $2.16 billion
annually.
PROBLEM GAMBLING AND RELATED CRIMES
According to psychiatrist Richard J. Rosenthal, most gambling addicts "are seeking 'action', an
aroused, euphoric state comparable to a 'high' derived from cocaine and other
drugs" (Worsnop, page 250), Counselors Arnie and Sheila Wexler of New Jersey state "most
dual-addicted cocaine addict/compulsive gamblers will tell you gamblling gives them a bigger
high. Some drug addicts, who are also gambling addicts, will sell their drugs for gambling
money." (Worsnop, page 250).
Several studies have been conducted in an attempt to quantify the costs of problem gambling
behaviors. In 1994, Rachel Volberg reported that the average individual pathological gambler
cost the public $13,600 each year (in 1981 dollars), This includes income that would have been
earned for those who lost their jobs, costs of prosecuting and incarcerating individuals for
crimes caused by their gamblilng behavior, and bailout costs, such as family gifts. Other
problems include lost job productivity, impaired judgment at work, lost productivity of spouses,
divorces, unemployment compensation, depression, physical illness related to stress, and
suicide.
A Marylalnd Department of Health and Mental Hygiene task force determined that its 52,000
adult gambling addicts cost citizens $1.5 billion in lost work productivity, monies stolen and
embezzled, bad checks and unpaid taxes (Worsnop, 1990, page 644). The cost per individual
compulsive gambler exceeds $28,846 and significantly increases when related costs for social
services, health care, bankruptcies, legal and correctional fees are considered.
The American Insurance Institute estimates that 40 percent of all white collar crime has its roots
in both legal and illegal gambling. Problem gamblers are responsible for an estimated $1.3
billion worth of insurance-related fraud per year. Insurance companies paid fraud victims an
average of $65,000 (Lesieur, page 45).
The mean gambling debt of people in compulsive gambling therapy ranged from $53,000 to
Page 1 of 3
r.!./lLt:,4
.,
Casino Watch-Florida Casinos Report
.
$92,000, not including debts paid.
As many as 10 to 17 people may be innocent victims of each compulsvie gambler (Lesieur,
1984), including.spouses, children, parents, other relatives, employers, co-workers and friends.
Two out of three compulsive gamblers will commit illegal activities in order to pay gambling
related debts and to continue gambling.
One out of every four problem gamblers have been involved in an auto accident during the
worst of their gambling. Almost half of those surveyed were speeding on their way to gamble or
on the drive back (Florida Council on Compulsive Gambling).
New Jersey residents also report that 28 percent know of someone who gambles too much
(Gallup, page 4).
Poor and working people spend a disproportionate share of their incomes on gambling
(Goodman, page 8). As a result, state gambling revenues come disproportionately from lower
income residents, causing a regressive form of taxation. Problem gambling behaviors are
highest among the poor and minorities (Goodman, page 17).
YOUTH
A 1985 random sample of 332 students at an Atlantic City high school found that 64 percent of
the students had gambled illegally at local casinos. Over 40 percent had gambled in casinos
before the age of 14.
COST OF PROBLEM GAMBLING BEHAVIORS IN FLORIDA
.
1. As noted by Lesieur, two out of three problem gamblers will commit a crime in
order to support illegal gambling activities. This means thai 266,667 individuals
in Florida will commit crimes such as burglary, larceny, theft, forgery or fraud.
While not all of the individuals will be convicted, they will likely continue to
commit crimes. Some will be convicted more than once.
2. The average sentence length of a person convicted of burglary in Florida is
5.7 years; theft, forgery and fraud 4.2 years; and robbery 8.8 years. First time
offenders and non-violent offenders will likely serve probation or community
control rather than prison time.
3. Of these categories, 41,3% will likely be convicted of burglary; 36.8% for theft,
forgery or fraud; and 21.9% for robbery based on current incarceration rates.
Not counting costs of prosecution, restitution or other related costs, incarceration and
supervision costs alone for problem gambler criminal incidents could cost Florida residents
$6.08 billion.
Florida's problem gamblers related to casinos would require $1.66 billion in prison construction
costs.
Deducting these positive social effects from the estimated social costs results in a high,
medium and low estimate of $3.25 billion, $2.65 billion, and $2.16 billion in net social costs,
respectively.
OTHER CRIME
.
While many believe that legalizing gambling activities will decrease illegal gambling, an
examination of dollars gambled does not support this belief. Dr. Vicki Abt, a sociologist at
Pennsylvania State University, says legalization "has not decreased the dollar amount of illegal
gambling; what is has done is decrease (illegal gambling) relative to legal gambling. Whereas
75 percent or so was wagered illegally before the 1960s, now about 75 percent is wagered
legally. But the total amount went up in the meantime." (Worsnop, page 643).
ATLANTIC CITY
Page 2 of3
\
Casino Watch-Florida Casinos Report
Page 3 of3
.
In just three years following the opening of its first casino, Atlantic City went from 50th in the
nation in per capita crime to first (Goodman, page 58). Crime spilled over into neighboring
communities that received no measurable economic benefits from casinos. Atlantic City's crime
rate rose 230 percent during the first three years of casino operations, over 25 times the growth
rate of 9 percent for the rest of the state (FDLE, 1994, page 3). It should be noted that Atlantic
City's population decreased by 20 percent during the same time period. Florida's crime rate
rose by 30 percent during these same years.
NEW ORLEANS
New Orleans estimates that crime-related costs will be just under $5 million for a single casino.
Widespread gambling could add an additional 1 0,000 new crimes at a cost to the city of an
additional $14.1 million.
COLORADO
Central City, Colorado has increased its police force from two full-time and one part-time officer
to 16 full-time officers in 1994 (Worsnop, page 256) to keep pace with its "skyrocketing
incidents" of disorderly conduct, assaults and DUls.
CONCLUSION
.
Since state tax revenues (adjusted for pari-mutuel and Lottery revenue losses) are estimated to
range between $155 million and $276 million while annual crime and social costs are estimated
at a minimum of $2.16 million annually (and up to $3.8 billion), it appears that casino costs
significantly outweigh the benefits of legalization.
According to University of Nevada professor of Public Administration William Thompson,
casinos generate significant new tax revenues "only if they can export their product. The local
economy doesnOt benefit if only local people are gambling."
Back to previolls page
'Home
.
TAB
"3 "
.
MANAGERIAL AND DECISION ECONOMICS
Manage, Decis. Econ, 22: 143-162 (2001)
DOl: 10,1002/mde.1004
Business Profitability versus Social
Profitability: Evaluating Industries with
Externalities, The Case of <:asinos
Earl L. Grinolsa.* and David B. Mustardb.*
. Department of Economics, University of lllinois, USA
b Department of Economics, Terry College of Business, University of Georgia, USA
Casino gambling is a social issue, because in addition to the direct benefits to those who own
and use casinos, positil'e and negative externalities are reaped and borne by those who do not
gamble. To correctly assess the total economic impact of casinos, one must distinguish
between business profitability and social profitability. This paper provides the most compre-
hensive framework for addressing the theoretical cost - benefit issues of casinos by grounding
cost-benefit analysis on household utility. It also discusses the current state of knowledge
about the estimates of both the positive and negative externalities generated by casinos.
Lastly, it corrects many prevalent errors in the debate over the economics of casino
gambling. Copyright @ 2001 John Wiley & Sons, Ltd.
INTRODUCTION
.
Between 1990 and 1998, commercial casino rev-
enues increased from $8.7 billion to over $22.2
billion, or 156%.1 The number of counties with
casinos rose from 26 to almost 200 in the same
time. Including Class III American Indian casi-
nos,2 casino revenues totaled $29.5 billion in 1998,
representing expenditures of $153 per adult aged
20 or over.
The rapid expansion of casinos to new parts of
the country generated extensive debates about the
impact of casinos on a range of social, economic,
and political issues.3 These concerns were suffi-
ciently pronounced to cause Congress to establish
the National Gambling Impact Study Commis-
sion (NGISC) in 1996 to conduct an exhaustive
study of the impact of casinos4 At the conclusion
of its investigation, the commission recommended
a national moratorium on the expansion of gam-
* Correspondence to: 'Department of Economics. University
of Illinois. 1206 S. 6th Street, Champaign. Illinois 61820.
E-mail: grinols@uiuc,edu; bDepartment of Economics, Terry
College of Business, University of Georgia, 528 Brooks Hal!,
Athens, Georgia 30602. E-mail: mustard@terry,uga,edu
Copyright <9 2001 John Wiley & Sons, Ltd.
.
bling and more study of gambling's effects, costs
and benefits, before making further decisions
about it.
The literature on the costs and benefits of
casino gambling is fraught with inadequacy and
confusion. Even studies that purport to evaluate
the economic impact of casinos commonly exhibit
a great deal of misunderstanding about what
should be included among benefits and costs, and
provide little or no guidance about how the costs
and benefits relate to one another or should be
computed. Many studies pay a great deal of atten-
tion, for example, to estimating the number of
direct and indirect jobs that casinos create and to
tallying the taxes casinos pay, but do not explain
the social value of an additional job or calculate
the lost taxes of competing non-casino busi-
nesses.5 In general, the costs and benefits dis-
cussed are casually listed, vary by study, and are
commonly presented with little or no justification
of how they were selected or why other potential
costs and benefits were excluded.
A recent paper, Eadington (1999), is instructive.
It identified three principal benefits of casinos: (1)
144
E.L. GRINOLS AND D,B, MUSTARD
.
gain in utility (for those gambling in moderation
for entertainment), (2) ancillary economic benefits
such as 'job creation, investment stimulation,
tourism development, economic development or
redevelopment, urban or waterfront revitalization,
or the improvement of the economic status of
deserving or underprivileged groups', and (3) ad-
ditional revenues to the public sector. He lists two
principal costs: (1) 'moral disapproval' and (2)
'fears of adverse social impacts', such as patholog-
ical gambling, crime, or political corruption. The
net increase in profits to business, unless this is
meant to be part of ancillary economic benefits, is
absent from the list of benefits.6 Although Ead-
ington lists gain in utility (clearly internal to the
individual or household) as a benefit, he writes
that 'many of the costs identified are internal to
the individual or the household, as opposed to
external-borne by society-and are therefore
difficult to place into a cost/benefit framework'.
This view of costs (including the references to
moral disapproval and fears of consequences in-
stead of the actual consequences) suggests that the
author believes costs are more subtle and possibly
less tangible than benefits, However, because the
process to determine how items are included is
not explained, there is little theoretical guidance
about how the identified co'st-benefit components
relate to one another in an overall assessment of
the impact of casinos or how competing costs and
benefits are reconciled. We will show how cost-
benefit components based on utility can be placed
into the evaluation framework.
To bring uniformity and more theory to bear
on the cost- benefit treatment of casinos, this
paper demonstrates the construction of an ex-
haustive and utility-grounded framework to iden-
tify costs and benefits. It outlines an explicit
taxonomy for costs and benefits based on the
principle of real resource use, and reviews the
available studies that contain original research
estimating one or more cost-benefit components.
Although the primary purpose of this paper is to
rectify theoretical cost- benefit reasoning as it ap-
plies to casinos, the methodology applies more
generally to the evaluation of projects in other
industries. We also review existing empirical esti-
mates of the costs and benefits of casinos ar-
ranged according to the theoretically grounded
principles. Unfortunately, there has been rela-
tively little research on many of the most impor-
tant social cost-benefit components, while much
. Copyright lD 2001 John Wiley & Sons, Ltd.
of research has examined less significant issues or
issues that are not even part of a properly defined
analysis of social costs and benefits. Some re-
search that purports to evaluate costs or benefits
actually examines local and not total social costs
or benefits. Another concern is that much of the
research has been conducted by organizations
with a vested interest in the outcome of the re-
search, institutes with industry ties, or state agen-
cies. Relatively little research is in peer-reviewed
journals. A review of the empirical literature that
estimates correctly defined components of social
costs and benefits indicates that the costs of casi-
nos are at least 1.9 times greater than benefits.
The remainder of the paper is arranged as
follows. The next seotion constructs a theoretical
cost-benefit measure based on economic funda-
mentals. The third and fourth sections examine
the social benefits and costs of casino gambling,
respectively. The fifth section concludes by sum-
marizing our contributions and outlining the im-
plications of this work for future research.
THEORY
.
Linking Cost-Benefit to Utility
In this section, we lay. out the foundations of
cost- benefit analysis for casino gambling. To
avoid the mistakes that have plagued cost-benefit
analyses, especially confusion about what can be
included on each side of the cost-benefit ledger
and how each item should be computed, we start
from the most fundamental cost-benefit concept
possible-individual utility. The framework we
employ can be as comprehensive and general as
desired, although our objective is to provide just
enough detail to include all of the major elements
commonly considered relevant to the economic
effects of gambling and enough explanation to
indicate what would change in a more detailed
application of the framework.
Our starting point is the change in the individu-
al's utility, u I - uO, where superscripts distinguish
utility in two situations. In one, casinos are wide-
spread geographically (alternative I) and in the
other, casinos are less widely spread (alternative
0). We assume that u(x) is a continuous utility
function representing locally non-satiable prefer-
ences defined on consumption xERn. A positive
element of x denotes consumption of a good or
service, while a negative component stands for the
Manage. Decis. Econ. 22: 143-162 (2001)
.
BUSINESS PROFITABILITY VERSUS SOCIAL PROFITABILITY
145
provision of a good or service.7 For example, the
provision of 10 hours of labor by the individual
would appear as - 10 in the labor component of
x. We define the expenditure function e(d, p, u) as
the minimum expenditure needed to achieve util-
ity u when the consumer buys and sells at prices p,
and d is the distance to the nearest casino. It is
strictly monotonic in u for any choice of fixed d
and p. The sign of e(d', pi, u1) - e(dl, pI, UO) is,
therefore, identical to the sign of u I - uO. In other
words, for fixed distance and prices d and p, e(d,
p, u(x)) is a utility function whose natural money
metric records utility in dollars.s
We compare the social welfare between the two
situations. We presume for simplicity that gam-
bling is a standardized good; casinos offer gam-
bling on essentially the same terms as casinos in
other locations.9 The primary advantage to the
consumer of more casinos, therefore, is closer
proximity to the nearest one. Let d} be the dis-
tance to the nearest casino for consumer i in the
post casino alternative I. Our measure of social
profitability is the change in welfare for all
consumers
.
~W= L wi[ei(d},p}, u}) - ei(d},p}, u?)),
where Li Wi = m, and m is the number of con-
sumer households. Equation (I) allows for differ-
ent weights for dollar gains to different
households, a topic to which we will return below.
However, in applying (I) to produce a working
measure of social profitability, we explicitly ad-
dress many issues left unspoken in some studies
and that are a source of confusion in others. The
initial model provides the simplest framework for
analyzing the impact of casinos. We list our as-
sumptions at the outset for clarity.
- We assume that a dollar of utility to one
household is equal to a dollar of utility to
another.lO With respect to Equation (I) this
implies that Wi = I for all households. It also
means that firm profits do not need to be
assigned artificial premia or discounts based
on which individuals or households happen to
own them.
- Firm profits are equally important to social
welfare regardless of which firm generates
them. For example, casino profits are valued
the same as the profits of a non-casino firm.
. Copyright ~ 2001 John Wiley & Sons, Ltd.
- To allow for regional tax differences, con-
sumers and firms may face different prices. In
the limit, each firm and household could have
a different, personalized set of prices. House-
hold i faces price vector Pi' finn j faces price
vector Pj' and endowments are traded at prices
Po'
- We allow for the possibility that consumers
may be constrained in their labor supply deci-
sions, resulting in unemployment. People have
a reservation wage but cannot always find a
job at that wage, and lowering their wage will
not increase the chances of their getting a job,
- Firms and economy endowments are owned by
households. Household i owns share Bij of firm
j, Li Oij = 1 and endowment niER~, where
Li ni = n, the economy endowment vector.
- The government spends tax revenues to pur-
chase goods and services, and private house-
holds receive utility from these expenditures.
To implement this assumption, we employ the
artificial device of having the government re-
turn tax dollars to households in a lump-sum
fashion. Households then spend the transfers
as part of their income and experience utility
gains based on their purchases.
- In addition to direct benefits and costs, casinos
may generate positive or negative externalities.
Positive externalities add value to the economy
not reaped by the agent creating them, while
negative externalities remove value not paid by
the causing agent, following the usual defini-
tion. For example, if a casino's presence re-
duces crime in an area, leading to less need for
police presence, this frees real resources to the
rest of the community and represents a posi-
tive externality. If the reverse is true, and the
casino increases the need for police, real re-
sources are removed from final consumption
x, and this is a negatIve externality. The third
and fourth sections discuss the nature of bene-
fits and costs in more detail. The net resources
gained or lost to the system are denoted by g.
If g > 0 negative externalities outweigh positive
externalities, which decrease the resources
available for consumption x, and thereby
lower social welfare. Social cost accounting in
real terms requires
x+g = y+.0.+z,
where x == Li Xi is aggregate consumption, and
Y == Lj Yj is aggregate production. For each
Manage. Decis. Econ. 22: 143-162 (2001)
.
146
E.L. GRINOLS AND D.B, MUSTARD
firm j, y) is the associated production vectorll;
z is the economy trade vector.12
The above remarks provide the simplest frame-
work that is sufficiently inclusive to discuss an
economy's social costs and benefits of gambling.
Application
Consider now the following carefully chosen iden-
tity, a telescoping sum where each term cancels
part of the preceding term.
I [e;(d}, p}, uf) - e;(d}, p}, u?)] =
I [e;(d},p},uf)-p}'xiJ
(Consumption Constraints in Situation I)
+ I fp}'x} -p?'x?J
(Income Effects)
+ I fp?' x? - ei(d?, p?, u?)J
(Consumption Constraints in Situation 0)
.
+ I [e;(d?, p?, u?) - e,(d}, p?, u?)J
i
(Distance Benefits)
+ I [e;(d}, p?, u?) - ei(d}, p}, u?)l
i
(2.5)
(Consumer Surplus)
Expression (2.1) measures the welfare impact of
constraints on the consumer's choice that prevent
him from being at his optimal bundle given the
prices he faces. The primary example of this kind
of constraint is unemployment. ei(d}, p}, ul) by
definition is the least costly way of achieving the
utility achieved in situation 1. Consumption bun-
dle x! satisfies u I = u(x f) and also achieves utility
u I. Because choice of x} was constrained (in the
case of unemployment, by the consumer's ability
to supply labor), it will lead to a greater expendi-
ture than ei(d}, pi, uf). Therefore, the difference
in expression (2. I) is the amount of money the
individual would be willing to pay to remove the
constraint. The same argument applies to expres-
sion (2.3) in situation O.
Expression (2.4) measures the value to the con-
sumer of having the nearest casino distance d}
away compared to distance d? For example, in
. Copyright @ 2001 John Wiley & Sons, Ltd,
(2.1)
the initial situation the consumer needed e;(d?, p?,
u?) to reach initial utility. When the nearest casino
is closer, distance d} <: d?, the income needed to
maintain original utility, ei(dl, p?, u?), is smaller
(presuming the individual gambles). The differ-
ence in expression (2.4), therefore, is the amount
the consumer would be willing to pay to have the
nearest casino closer.
Expression (2.5) is the conventional measure of
consumer surplus. The only difference between
the two terms in the expression is the price vector.
If prices p 1 are better for the household than
prices p? (lower for goods purchased and/or
higher for goods sold, such as labor), then expres-
sion (2.5) is positive and measures the amount of
money the consumer would be willing to give up
to have the better set of prices.
Now examine expression (2.2). Use the house-
hold budget identity
(2.2)
(2.3)
P.'x= '\' O.rr,+p^.Q.+ T-E.
I I ~ I} j H I I ,
j
(3)
(2.4)
to transform the income effects in (2.2) where IT)
is the profit of firm j, Po' Oi is earning from the
household's endowment, Ti is the household's
share of taxes, and Ei is the household's share of
the cost 'of gambling-induced externality expendi-
tures. Summing (3) over households and differ-
encing between the initial and final situationsl3
yields
I fp}' xl - p?' x?l = I .1IT) (.1Profits)
i j
+ .1Po.Q (Endowment Capital Gains)
+.1T (.1Taxes) -.1E (.1Externality Costs) (4)
Substituting (4) into (2); writing the distance ef-
fects in differential form and rearranging gives the
taxonomy of cost-benefit elements that we seek:
I [ei(d},p}, u}) - e(d},p}, u?)] == .1w=
i
f.d} ae,
I .1il) + I dO adddi + .1T - .1E
j I I t
+ Consumption Constraints + .1Po.Q
+ I [ei(d}, P?, u?) - ei(d}, p}, u?)],
(5)
where 'Consumption Constraints' IS the sum
(2.1) + (2.3).
The seven components in Equation (5) are an
exhaustive, exact tabulation of the cost- benefit
Manage. Decis. Econ. 22: 143-162 (2001)
.
BUSINESS PROFITABILITY VERSUS SOCIAL PROFITABILITY
147
.
elements for evaluating the economic effects of
casinos. Moreover (5) shows precisely how each
term should be computed theoretically. For exam-
ple, the effect of casino gambling on firm profits
should be summed over all firms, not just casinos.
The increased profits of the casinos should be
netted against lost profits of other firms that
compete for consumer spending. Comparable
statements apply to the computation of employ-
ment benefits and costs, taxes, and social costs.
There is one obvious simplification we can
make to (5). Because gambling industry revenue
(casinos, lotteries, racetracks and other forms of
gambling) is relatively small,14 it will have a negli-
gible effect on creating capital gains or losses on
endowments, It is unlikely that the cost of capital,
for example, will differ because of the presence or
absence of casinos in the economy. A similar
statement applies to consumer surplus effects that
depend on gambling to influence overall prices. 15
Therefore, for the remainder of the paper we
assume that firm and household prices are invari-
ant to the amount of gambling (P? = pi, pJ =
p), p~ = pb), which means that the last two terms
in Equation (5) related to capital gains on endow-
ments and consumer surplus gains and losses drop
out.
Conceptual Corrections
Equation (5) allows us to address some common
errors and misconceptions of cost- benefit analy-
sis applied to gambling.
The first error is the tendency to identify busi-
ness profitability, Lj TIj' and its improvement,
Lj ~TIj, with social profitability. The two are dif-
ferent. Business profitability is clearly important
to social profitability and contributes to it, but the
two are not synonymous. Failure to account for
all of the components of social profitability is
perhaps the most common mistake. Casino profits
are visible and prominent. Other costs and bene-
fits may be less so.
The second error is to evaluate the economic
impact of gambling with respect to the taxes and
profits of a subset of firms-typically the profits
of firms in one state or region and sometimes the
profits of local gambling firms only. Equation (5)
sums profits over all firms, not just casinos or
firms in one location. Ignoring firms that lose
profits due to the expansion of gambling is equiv-
alent to selecting weights for them in Equation (1)
. Copyright <<J 2001 John Wiley & Sons, Ltd.
that are zero. Because households own these other
firms, this violates the assumption that house-
holds are treated equally,
, The third is to consider only the taxes of a
subset of households or regions. It is not uncom-
mon, for example, for studies to focus only on
costs within the state, even though casinos that
border another state have ramifications for citi-
zens of the neighboring jurisdiction. Equation (5)
sums taxes over all households and regions.
Evaluations that consider only the costs or
benefits of a subset of households or regions are
inaccurate and incomplete. For example, the
cost- benefit measure in (5) does not treat a job in
a given location as more valued than a job in
another location. Many economic impact studies
perform regional net export multiplier analyses of
the effects of casinos. They erroneously report the
number of jobs in a given location as a benefit.
According to (5) the value of employment in one
location (part of the determination of firm profits)
must be netted against the value of employment in
another location. There is no net gain to the
economy from shifting a job from one location to
another unless it increases profits to the
economy. 16
The last common error is that much empirical
work purports to show casinos decrease unem-
ployment, but fails to prove what employment
would have been in the absence of casinos. Most
casinos were introduced after 1991, when the
country was recovering from the recession of
1990-1991. The period from 1991 to 2000 also
coincided with the longest economic expansion in
American history. As the country emerged from
the recession, unemployment declined in areas
with and without casinos. If casinos temporarily
reduced unemployment faster than it would have
fallen otherwise, this transitory effect could cor-
rectly be counted as a benefit of casinos. How-
ever, we know of no study that has made this
case. On the contrary, the failure to account for
the decline in unemployment that would have
occurred anyway leads to a classic post hoc, ergo
propter hoc fallacy of logic. For a more detailed
example, see Appendix A, which discusses The
Evans Group (1996). Although it argued that
casinos reduced unemployment, it did not report
that areas without casinos with comparable start-
ing unemployment rates experienced comparable,
and in many cases, larger reductions in the unem-
ployment rate.
Manage. Decis. Econ. 22: 143-162 (2001)
.
148
BOUNDING BENEFITS
E,L. GRINOLS AND D,B, MUSTARD
This section reviews the studies that estimate the
benefits from casinos based on the theoretically
correct cost- benefit computation in Equation (5).
We discuss in order the net increase in firms'
profits plus taxes paid due to the presence of
casinos, the consumer distance benefits of nearer
casinos, employment benefits and total benefits
from the expanded gambling opportunities.
.
Profits and Taxes
This benefit is calculated by determining the
casino profits and taxes minus the reduction in
profits and taxes of other businesses caused by.
casinos. Although casino profits and taxes are
highly visible, they are invalid measures of social
benefits because they do not adjust for the entire
economy for the lost profits and taxes of compet-
ing businesses. This point is not special to casinos.
Any business-be it Wal-Mart or a drugstore
chain, that attracts consumer sales, employs labor
and other inputs, and displaces competing busi-
nesses-should be evaluated on the same basis,
Because many casinos do not have to report
their profits or pay taxes (for example, casinos
owned by American Indian tribes), there are no
data on industry profits. However, we can esti-
mate revenues from annually published informa-
tion. We provide a brief overview of casino
gambling in the US before estimating the benefits.
Table 1 reports total and per capita gambling
revenue.17 For comparison, we provide data on
the tobacco industryY Many studies estimate po-
tential casino revenues using the amount of gam-
bling per person in areas where casino gambling is
a prominent activity. For example, the City of
Chicago Gaming Commission funded a study
(Deleitte and Touche, 1992) that reported that
adults within 35 miles of Atlantic City lost $198
per adult annually to casinos. Adjusted for eight
years of price changes, this figure is approxi-
mately $230. In its study, the Mirage Hotel (1993)
estimated that annual per capita gambling rev-
enues for persons residing within a 50-mile radius
of its proposed Chicago suburb gambling facility
would be $200.19 In Iowa, in 1995, a Christiansen
and Cummings Associates study for the state
Racing and Gaming Commission found that the
average adult lost $172 to the casinos (this figure
is lower than $230 because casinos are still not in
close proximity to all parts of Iowa). These data
are comparable to revenue for other areas.
In addition to avera.ges we are interested in the
concentration of gam bling among users. Many
studies examined gambling markets in different
locations and at different times. Taken together,
they provide a general estimate of how frequently
residents gamble. In a market with readily avail-
able gambling opportunities including casinos, ap-
proximately 30% of the population does not
gamble, meaning that they will not have gambled
in the past year20 Another 50-60% could be
termed light bettors, who gamble less than once
per week. This group includes those who enjoy a
night out at the casino once in a while, but do not
frequent casinos. About 5-15% could be termed
heavy bettors who gamble twice per week or
more. The last 2-5% of the population consists of
problem and pathological (P&P) gamblers, who
suffer from compulsive gambling disorders, which
are expressed when the opportunity to gamble is
present and sufficient time has elapsed for the
problem to become evident. This group might be
in the casino daily, for long periods of time, and
at unusual hours. Two..thirds to 80% of gambling
revenues come from the 10% of the population
that gambles most heavily.21 Expressed in reverse,
90% of the population may provide as little as
20% of casino revenues. Consequently, the great
majority of adults are indifferent, or nearly indif-
ferent, to the availability of casino gambling.
Although the average American adult loses
Table 1. The Casino Market
Casino gambling revenues
per adult ($)
All gambling revenues
per adult ($)
All gambling
revenues ($)
Total revenues ($)
US 1998
'Saturated market'
Tobacco industry
282
359
54.4 billion
69.3 billion
153
::::: 230
29,5 billion
44.4 billion
39 billion
. Copyright <92001 John Wiley & Sons, Ltd.
Manage. Decis. Econ. 22: 143-162 (2001)
.
BUSINESS PROFITABILITY VERSUS SOCIAL PROFITABILITY
149
.
approximately $153 per year and might lose closer
to $230 per year were gambling more widespread,
these revenues come from a few who gamble a lot,
instead of many who gamble a little.
We now return to our original question-what
is the social value of this amount of casino gam-
bling? According to Equation (5) we need the
profits and taxes attributable to casinos, minus
the reduction in profits and taxes of other busi-
ness due to casinos. To these we must add the
consumer distance benefits of casinos (which we
address in the next section). Because profits are a
function of market structure and the presence of
free entry and exit, if casinos were deregulated,
market contestabi]ity and free entry of casinos
would drive economic profits to zero. In that
event, from the perspective of profits, a larger
casino sector and smaller remainder of the econ-
omy would represent a net wash because eco-
nomic profits in the economy would be no greater
with casinos than without. The sole contribution
of casinos to social welfare in that case would be
the direct consumer benefits.
However, in the current legal environment,
casinos in many locations are effectively region a]
monopolies sustained by government licensing re-
strictionsY We, therefore, make the following
adjustment to allow for the higher monopoly
profits of some casinos, In ] 998, profits before
taxes23 of all non-financial corporate business in
the United States were 13.8% of sa]es.24 Assuming
that casinos average 30% profit rates before taxes
(more than double the norma] business rate of
profit) implies that socia] benefits in the form of
profits and taxes from shifting $153 of revenue
from other businesses to casinos is (0.30-0.138)
]53 = $25 rounded up to the nearest dollar. In the
next section, we add to this consumer distance
benefits of casinos to produce an upper bound on
total casino socia] benefits.
Consumer Distance Benefits
Equation (5) also idenifies s:;~ (ae /ad)ddj as a
direct social benefit of casinos, where di is con-
sumer i's distance to the nearest casino. Distance
benefits have been ]ittle studied, even though they
constitute a primary direct benefit of casinos. To
our knowledge, only Grino]s (1999) estimated
these benefits and compared them with the other
components of (5), Assuming that utility depends
.
Copyright ~ 2001 John Wiley & Sons, Ltd.
on goods x, the number of casino VISIts V, the
amount gambled (spent) per visit g, and the dis-
tance traveled to the casino, u = u(g, V, leg, d))
where [(g, d) is an enjoyment factor or visit 'in-
tensity' factor that rises with g and falls with d
and g is consumption of other non-casino, goods.
The envelope theorem and consumer optimization
. . Sd1 Sd1
condItIOns show that d'O (ae /ad)ddi $, '0 Vdg.
, dj
This inequality allows inferences about welfare to
be made from data that relate to the number of
visits and amount gambled per visit to the dis-
tance from the casino. Grinols (1999) estimated
that the upper bound for direct conumer benefits
of casinos was $50 per adult (again, rounding up
to the nearest round fIgure to produce an upper
bound on casino benefits) when no allowance is
made for the significant portion of revenues from
problem and pathological gamblers. If the rev-
enues of non-P&P gamblers only are used to
calculate consumer distance benefits, then the
benefit figure falls to under $3425 This number
can be interpreted as the answer to. the question,
'How much would you be willing to pay each year
to have the opportunity to gamble in a casino
nearby compared with the alternative where casi-
nos are 1000 miles away?'
Employment Benefits
Although the topic of employment benefits is one
of the most studied issues about casino gam-
b]ing,26 it also contain~; a widespread and centra]
misunderstanding-that the benefits of new busi-
nesses are measured by the jobs they create in a
given location. While it may be legitimate to ask
what effect a new business will have on employ-
ment, what taxes it will pay, and from where its
revenues will come, these answers do not assess
tbe social benefits and costs of the business. In-
creasing jobs in one location at the expense of lost
jobs in another is not a social benefit. Business
profitability is not social profitability. Socia] cost
benefit is grounded on consumer utility and re-
sults in a list of relevant factors different from
tracking income and employment effectsY
Total Social Benefits
Based on the previous sections, if casinos were
fully deregulated and allowed to spread freely
nationwide, economic profits would be driven to
zero. The net increase in profits and taxes from
Manage, Decis. Ecan. 22: 143-162 (2001)
.
150
E.L. GRfNOLS AND D.H, MUSTARD
.
expanding the casino sector at the expense of the
rest of the economy, therefore, would be zero.
The consumer distance benefits of casinos would
be less than $50 per adult, or if the revenues of
P&P gamblers are subtracted, $34 per adult.
If casinos are regulated and granted regional
monopoly status in some jurisdictions, the eco-
nomic profits of casinos will remain positive, but
the distance benefits will drop. Assuming average
pre-tax profits equal to 30% of sales (more than
double the rate for non-financial corporate busi-
ness in the US) implies that the net profit and tax
benefits of casinos are less than $25 per adult.
However, if there is not free entry, distance bene-
fits will average less than $50 per adult (less than
$34 adjusting for P&P gamblers) because some
areas will not have casinos close to consumers.
We are, therefore, left with three upper bounds.
The preferred number, $34 per adult, is the most
correct upper bound because it represents the full
social value of casinos under circumstances in
which all of the benefits would be captured by
consumers if the industry were deregulated to
allow free entry. Fifty-nine dollars combines the
full estimate of consumer distance benefits ad-
justed for P&P gamblers with a generous profit
figure. It is too high because the consumer benefit
is overstated, and in addition, because it fails to
recognize that distance benefits would decline
with regional monopolies present that do not put
casinos close to all consumers. Finally, $75 per
adult adds consumer benefits to profits without
making any adjustments. We emphasize that these
numbers are upper bounds on the estimated
benefits.
COUNTING COSTS
Researchers estimate the social costs of casinos
using two methods. The first is through the study
of problem and pathological gamblers. The sec-
ond is through statistical analyses of cost-creating
activities such as crime, suicide, and bankruptcy.
The former approach ties the cost activities to
gamblers, but overlooks social costs that do not
derive from problem and pathological gamblers,
The latter approach, determining the effect of
casinos on social costs such as crime by examining
aggregate statistics, is direct and more inclusive
because it looks at more than just the crimes
committed by P&P gamblers,
. Copyright @ 2001 John Wiley & Sons, Ltd.
The remainder of this section consists of two
parts. The first derives a detailed taxonomy of
cost classifications tied to the -theoretical analysis
in the second section. When discussing these clas-
sifications, we cite cost studies of both types listed
above. The second part of this section is a more
detailed review of all the studies that focused
specifically on problem and pathological gam-
blers, We calculate costs per pathological and
problem gamblers, and estimate the costs for the
entire nation. These sections constitute the most
comprehensive compilation of the social costs of
gambling available to date,
Cost Taxonomy
The underlying principle, based on Equation (5),
is that each social cost uses physical resources g in
ways that do not directly enter utility or that
reduce economic efficiency. We arrange social
costs into nine disjoint groups and discuss each
one briefly. '
1. Crime: Of all the social costs, the link between
casinos and crime has received the most atten-
tion.28 Crime costs are real resources used for
the apprehension, a.djudication, incarceration,
and rehabilitation of criminals, or the police
costs that result from the need for increased
police presence in areas of greater gambling
activity. One significant problem that has
plagued the majority of the casino-crime liter-
ature is analogous to the problem present in
calculating the profit and tax benefits of casi-
nos: To estimate social costs, one should not
count new crime around the casino only, but
also consider whether casinos reduced crime in
other locales (for example, this could happen
if casinos move crime from other locations).
Counting only local crime as a cost is similar
to counting only local profits as a benefit. -
The most comprehensive analysis of the
casino-crime link is Grinols et at. (2000),
which evaluated county-level data for seven
offenses in every US county over 20 years, and
controlled for about 50 variables. It concluded
that on average, 8 -10% of crime in casino
counties in 1996 could be attributed to the
presence of casino gambling in the county,
resulting in costs of $63 per adult annually in
these counties. Furthermore, counties that
border casinos also experience increased
crime rates, which suggests that casinos truly
Manage. Decis. Ecan. 22: 143-162 (2001)
.
BUSINESS PROFITABILITY VERSUS SOCIAL PROFITABfLlTY
151
increase crime, not merely shift it from one
location to another. Estimates of the cost of
non-Index crimes would add to total crime
costs. For example, insurance fraud is not an
FBI Index I crime. Estimates of the fraud
committed by gamblers is $1.3 billion per
year,29 or $6.61 per adult annually.30
Studies of problem and pathological gam-
blers have found similar effects. Maryland De-
partment of Health and Mental Hygiene (1990)
reported that 62% of gamblers in treatment
committed illegal acts as a result of their gam-
bling, 80% committed civil offenses and 23%
were charged with criminal offenses. Lesieur
(l998b) surveyed nearly 400 members of Gam-
blers Anonymous, 57% of whom admitted
stealing to finance their gambling. On average
these 400 people stole $135000, and their total
theft was over $30 million. Lesieur (1992) re-
ported on illegal activities and civil fraud en-
gaged in by pathological gamblers to gamble
or to pay gambling debts in five samples from
hospital inpatients, Veterans Administration
and Gamblers Anonymous groups, male pris-
oners, female prisoners, and a female Gam-
blers Anonymous sample that includes the
white-collar crime and other crimes listed in
item 1.31
2. Business and Employment Costs: These costs
include lost productivity on the job, lost time
and unemployment: sick days off for gambling,
extended lunch hours, leaving early to gamble,
and returning late after gambling. Problem and
pathological gamblers often impose costs on
their employers (in addition to theft or em-
bezzlement discussed in the section on abused
dollars below) in the form of an unreliable
presence on the job and reduced productivity
when present. Between 21 and 36% of problem
gamblers in treatment reported losing a job
because of their gambling (Lesieur, 1998b).
Firing an employee imposes costs on the
worker in terms of lost output during the
period of unemployment and on the employer
in terms of additional costs of hiring and
training new employees. These costs are higher
the greater the firm-specific human capital.
3. Bankruptcy: Lawsuits and legal costs, and bill
collection costs, bill collector harassment are
among the consequences of bankruptcy.
Pathological gamblers often follow a pre-
dictable path of exhausting personal resources,
. Copyright (Q 2001 John Wiley & Sons, Ltd.
.
selling insurance policies, selling possessions,
and 'borrowing' from family and friends. Their
search for funds may lead them to acquire
multiple credit cards that they use to the limit.
Debts will be paid off, of course, when the
individual wins big in his next gambling
episode. Bankruptcy entails costs to creditors
attempting to collect and costs to the legal
system in court time and resources. SMR Re-
search Corporation (1997, p. 118) indicated
that these costs may be quite large, 'We set out
this year to interview many of the leading US
experts on gambling, gambling addiction, and
the financial impacts of gambling. Their stud-
ies have suggested, fairly consistently, that
more than 20% of compulsive gamblers has
filed for bankruptcy as a result of their gam-
bling losses',
4. Suicide: Lesieur (1992) concluded that problem
and pathological gamblers have higher suicide
rates than the general public.32 Dozens of sto-
ries have been reported of gamblers killing
themselves after losing at the casino, some-
times on the premises.J3 Consistent with this,
Phillips et at. (1997) found that deaths in Las
Vegas were 2.5 times more likely to be a result
of suicide than deaths in other comparably,
sized metropolitan areas. Visitors to Atlantic
City and Reno were 1.75 and 1.5 times more
likely to die in suicides than tourists to other
non-gambling areas, and in Atlantic City the
suicide rates did not become elevated until
after casinos were introduced in 1978.
McCleary et at, (1998), funded by the Ameri-
can Gaming Association, contested Phillips'
findings. While we recognize the impact of
casino gambling on suicide, the literature has
not provided sufficiently reliable social cost
estimates, and, therefore, we do not account
for such costs in th(: table below.
5. TIlness: Among the forms of sickness associated
with gambling or affected by it are depression,
stress-related illness, chronic or severe
headaches, anxiety, moodiness, irritability, in-
testinal disorders, asthma, cognitive distor-
tions, and cardio-vascular disorders. Many
sickness costs are borne by the gambler, but
they also appear as resource costs when the
gambler seeks treatment. Gambler-borne costs,
even when not absorbing resources, however,
are tangible costs to the extent that the gam-
bler would be willing to pay to eliminate the
problem.
~
Manage. Decis, Econ. 22: 143-162 (2001)
.
152
E.L. GRINOLS AND D,B, MUSTARD
6. Social Service Costs: This category of costs
includes therapy/treatment costs, unemploy~
ment and other social service costs (includes
welfare and food stamps).
7. Government Direct Regulatory Costs: Social
service and government direct regulatory costs
are paid primarily through the government.
The gambling industry has been regulated be-
cause it has historically been subject to fraud
and abuse. Social service costs transfer re-
sources from one segment of society to an-
other, consuming resources in the process. If
social costs include the financial burden placed
on the non-gambling society that would not be
present in the absence of gambling, then these
costs should be included for a complete assess-
ment of the effects of gambling. Regulatory
costs differ by state and depend on the type of
casinos (i.e. riverboat, Indian reservation,
etc.), and extent of the responsibilities of the
regulatory agencies.
8. Family Costs: Families of problem and patho-
logical gamblers bear gambling-related costs
of divorce, separation, spousal abuse, and
child neglect. Although these costs are non-
pecuniary, they are, nevertheless, tangible and
real. They can be quantified in terms of the
amount of money an individual would be will-
ing to pay to remove the problem. In practice,
such costs are rarely measured. When social
services become necessary, as when gambling
leads to divorce proceedings, they represent
resources lost to other uses in society and can
be measured by the cost of the services
provided.
9. Abused Dollars: The final category represents
lost gambling money acquired from family,
friends, or employers under false pretenses.
Two examples are stealing that is never re-
ported because the thief is a relative, and
money 'loaned' under duress that is never
repaid. Abused dollars represent costs to the
non-gambling population. To the extent that
abused dollars represent purchases of gam-
bling services that are inefficiently sub-optimal
from the gambler's perspective or create mar-
ket inefficiencies, a significant portion repre-
sents social costs to society as a whole even
allowing for gains by the gambler or gambling
sector.34
. Copyright @ 2001 John Wiley & Sons, Ltd.
.
Social Cost Estimates Tied Directly to P&P
Gamblers
Table 2 reports the results of all eight studies that
contain original research that ties social costs
directly to pathological gamblers.35 The first two
rows show the location studied and the author(s),
respectively. The first column shows the category
of costs, as outlined in the previous section. The
studies are listed in order of date of publication.
With the exception of the path breaking paper by
Politzer et al. (1981), the studies were published
between 1994 and 1999. The column totals range
from a low of $1,195 (Gerstein et at., 1999) to a
high of $30,235 (politzer et al., 1981). The Execu-
tive Office of the Governor (1994) is the highest
post-1994 estimate, Because all studies omit some
of the costs, these totals will understate the actual
totals.
A large share of the differences in the totals is
explained by differences in the number of cost
components the studies estimated. The Executive
Office of the Governor (1994) estimated only
crime costs in Florida, while Thompson and
Quinn (1999) estimated ten components. The
study with the lowest total cost (Gerstein et al.,
1999) estimated only four categories. No study
estimated all the components.36 By far, crime and
abused dollars are the largest cost estimates. Ger-
stein et al. (1999) is the only study that completely
omits crime costs, and only the Executive Office
of the Governor (1994) and Gerstein et at. (1999)
omit estimates of abused dollars. One important
common characteristic of all but one of these
studies is that they are not published in peer-
reviewed journals. The Executive Office of the
Governor (1994), Ryan et a/. (1999), Thompson
and Quinn (1999) and South Dakota Legislative
Research Council (1998) were either published by
or prepared for state agencies. Thompson et al.
(1996) was published by the Wisconsin Policy
Research Institute. Politzer et al. (1981) was pre-
sented at the Fifth National Conference on Gam-
bling and Risk Taking, Gerstein et at. (1999) was
presented to the NGISC. The paper by Thompson
et al. (1998) was presented at the Twelfth Na-
tional Conference on Problem Gambling, and
later published in Gaming Research and Review
Journal.
We used many strategies to ensure that the final
estimates of costs per pathological gambler were
Manage. Decis. Econ. 22: 143-162 (2001)
. . .
()
0
"0
~
.....
~'
S':
@
tv
0
0
...... Table 2. Annual Social Costs per Pathological Gambler
0
:;J'" MD FL WI CT SD LA US SC Row
::J Politzer Exec. Thompson Thompson SD Leg, Ryan Gerstei n Thompson ttl
averages C
~ el at. Office el ai, el at. Research el ai, el ai, and Quinn for studies en
Z
0' (1981) of Gov (1996) (1998) Council ( 1999) (1999) (1999) 1994-1999 t'I1
~ ($) (1994) ($) ($) 1998-1999 ($) ($) ($) ($) '"
'"
R<> ($) ($) "'d
C/.l ::<l
0 Crime 0
::l .."
-'" Apprehension and increased police costs 44 71 1000 53 116 257 ::j
t""" Adjudication (criminal and civil 1788 1234 994 27 649 476 676 :>
.... justice costs) ~
P. t'"'
Incarceration and supervision costs 2828 15221 758 889 382 690 451 3065 ::j
Business and employment costs 11265 ><
<
Lost productivity on job 1082 1082 t'I1
Lost time and unemployment 2717 3436 5936 320 2156 2913 i<'
en
Bankruptcy 515 1I8 316 C
en
Suicide '"
0
()
~ Illness 700 700 ;;
Social service costs t'"'
;::s "'d
~ Therapy/treatment costs 437 114 75 396 30 83 189
OQ i<'
~ Unemployment and other soc. svc. 606 971 549 60 145 318 442 0
.."
b (inc!. welfare and food stamps) ::j
'" :>
'" Government direct regulatory costs
~. ~
Family COSts r
~ Divorce, separation 111 III ~
C)
:=s Abused dollars 14354 3802 9519 240 3175 2436 3834
N 13586
!';J
.j:>.
w
I
0\
Iv
.--
tv
0
0 Vl
'-" W
.
.
.
154
E.L. GRINOLS AND D,B. MUSTARD
lower bounds.3? First, in calculating the average
annual cost per pathological gambler by category
(shown in the last column of Table 2 on the right)
we omitted Politzer et al. (1981).38 This study had
the highest cost estimates, but was conducted at a
different time and in a different gambling environ-
ment from the other studies. Second, costs for
suicide and government regulation are omitted,
because none of these studies estimated them.
- Third, we did not price adjust the estimates, but
rather took the values as given by the authors.
Last, many studies combined their estimates for
pathological and problem gamblers. We treated
the numbers as if the costs we report apply only to
pathological gamblers. Because costs due to
pathological gamblers are higher than costs due to
problem gamblers, the estimates further under-
estimate the costs connected to pathological
gamblers.
Table 2 shows that the total average social cost
of eight studies is $13,586 per pathological gam-
bler per year. If 1.5% of 196.65 million US adults
were pathological gamblers, this would imply an-
nual social costs of $40.1 billion or $204 per
adult. If pathological gamblers are 1% of the
population, the estimate reduces to $136 per
adult.
Table 3 replicates Table 2 for problem gam-
blers. Only Gerstein et al. (1999) and South
Dakota Legislative Research Council (1998) esti-
mated any separate costs per problem gambler.
These studies estimated only three of the many
cost categories. The average annual cost per prob-
lem gambler by cost category is shown in the last
column. For the same reasons discussed in analyz-
ing the results for pathological gamblers, the
Table 3 total cost estimate of $912 due to problem
gamblers understates the actual cost.
Table 4 applies the information in Tables 2 and
3 to produce annual national social costs per
adult. To test the robustness of these cost esti-
mates, we use the 95% confidence bounds on the
numbers of problem and pathological gamblers
set by Shaffer et al. (1997),39 This confidence
interval sets the fraction of pathological gamblers
between 0.9 and 1.38% of the adult popula-
tion, and the fraction of problem gamblers
between 1.95 and 3.65% of the adult popula-
tion. Based on these lower and upper bounds,
annual national social costs from problem and
Table 3. Annual Social Costs per Problem Gambler
US
Gerstein el ai,
(1999)
($)
SD
S, Dakota,
1998-1999
($)
Row averages:
studies 1994-1999
($)
Crime
Apprehension and increased police costs
Adjudication (criminal and civil justice costs)
Incarceration and supervision costs
Business and employment costs
Lost productivity on job
Lost time and unemployment
Bankruptcy
Suicide
Illness
Social service costs
Therapy/treatment costs
Unemployment and other soc. svc.
(incl. welfare and food stamps)
Government direct regulatory costs
Family costs
Divorce, separation
Abused dollars
200
200
360
155
549
360
352
912
Copyright @ 2001 John Wiley & Sons, Ltd.
Manage, Decis. Econ. 22: 143-162 (2001)
.
.
.
BUSINESS PROFITABILITY VERSUS SOCIAL PROFITABILITY
155
Table 4. National and per Adult Social Costs
NATIONAL COST:
Problem Rate Hi9hl $30.6
Low $27.5
Low
Pathological Rate
BILLIONS of DOLLARS
I ~:~:: I
High
Problem Rate High
Low
Pathological 95% Confidence Bound: LOWER
Pathologicaf!35% Confidence Bound: UPPER
Problem 95% Confidence Bound: LOWER
Problem 95% Confidence Bound: UPPER
0,9000%
1.3800%
1.9500%
3.6500%
pathological gambling range from 27.5 billion to
over $43 billion. On a per adult basis, the num-
bers range from a low of $140 to a high of $221.
Because Shaffer et al. (I997) estimated these con-
fidence bounds based on samples of the nation
before the time of publication including areas
with different degrees of casino gambling they
clearly understate the fractions of the entire US
population that would be identified as pathologi-
calor problem gamblers if casinos were expanded
fully. The costs of Table 4, therefore, also under-
state the associated costs of full gambling expan-
sIon.
IMPLICATIONS FOR FUTURE RESEARCH
This paper makes many contributions to the dis-
cussion of social costs and benefits of casino
gaming, and has numerous implications for future
research in this area. First, we provide the first
theoretical justification of what should be in-
cluded as costs and benefits. This justification is
based on individual utility and distinguishes busi-
ness and social profitability for industries with
externalities. The lack of a clear theoretical basis
has impaired the entire research agenda on this
issue. Much research has examined relatively mi-
nor issues or issues that are not even part of a
properly defined cost-benefit analysis. Con-
versely, there are relatively few estimates of some
of the key components of social costs and bene-
fits. Consequently, a well-grounded theoretical
framework of costs and benefits will make future
research more productive.
Copyright @ 2001 John Wiley & Sons, Ltd.
Second, using this theoretically grounded cost-
benefit analysis we corrected several common
conceptual mistakes prevalent in the casino and
gambling literature. One example of a common
error is the focus on local rather than total social
costs or benefits. On the benefits side, increases in
local profits and taxes are often weighted heavily
while losses in profits and taxes from geographi-
cally distant areas are weighted less or not at all.
Similarly, on the cost side, local crime is often
weighted heavily while there is little discussion
about whether crime was simply moved from
other areas. Another error is the frequent use of
the net export-multiplier modeling of jobs, an
inappropriate method to determine social costs
and benefits. Clearly, identifying these errors will
reduce them in future research.
Third, we used the theory to construct a clear
taxonomy of benefits a.nd costs as applied to the
casino industry. To estimate these costs and bene-
fits we reviewed the available studies that do
original research on this topic. This literature
shows that the extreme upper bound on annual
total social benefits is $75 per adult. The lower
bound for social costs, based on the estimates of
costs associated with prevalence of problem and
pathological gamblers, was between $140-$221
per adult. Consequently, the available research
indicates that when using the highest estimates of
benefits and the lowest estimates of costs, casino
gambling fails a cost-benefit test by a ratio of 1.9
to one or greater.40 Standard Pigouvian corrective
theory for an industry with externalities is that it
should be taxed by an amount equal to the costs
that it imposes on society. Relative to the
Manage. Decis. Econ. 22: 143-162 (2001)
.
156
EL. GRINOLS AND D,B. MUSTARD
revenues for a representative casino of about $230
per adult each year from nearby residents, Pigou-
vian corrective taxes would represent between 61
and 96% of casino revenues.
Fourth, we showed that the available research
indicates there is a lack of quality research on both
the benefit and cost sides of the debate, and that
there is an important need for better research,
There is a need for more uniformity in the manner
in which costs and benefits are treated. Peer-
review-quality studies not funded by the casino
industry or by pro- or anti-gambling groups are
especially needed to refine and improve the cost-
benefit numbers that are currently available. To
further refine the cost- benefit analysis of casino
gaming the following questions must be addressed.
.
What is the Effect of Casinos on the Number and
Gambling Patterns of Problem and Pathological
Gamblers?
Because the social costs of the casino industry are
generated primarily by problem and pathological
gamblers, it is essential to know how casinos affect
problem and pathological gamblers. There is
abundant evidence that increased gambling oppor-
tunities increase problem and pathological gam-
bling. For example, the NGISC reported that the
presence of a casino within 50 miles roughly
doubled the prevalence of problem and pathologi-
cal gambling.4! Other indicators include the
tremendous increase in the numbers of gamblers
seeking help when casinos enter a market, the
increase in gamblers anonymous groups when
gambling enters a state, and the evidence from
survey data on the number of problem and patho-
logical gamblers before and after casino expan-
sIOn.
Casinos may also affect the amount of gambling
by problem and pathological gamblers. An aver-
age adult is expected to lose $200-300 each year in
casinos if they are nearby, while a typical patho-
logical gambler often loses 10-20 times this
amount. Therefore, a small number of pathologi-
cal gamblers accounts for a significant portion of
casino revenues. A related issue is to determine the
share of casino revenues that derive from problem
and pathological gamblers. Does this share differ
by type of gambling? For example, lotteries receive
a smaller portion of their revenues from P&P
gamblers because lottery play attracts a larger
portion of the population.
.
Copyright @ 2001 John Wiley & Sons, Ltd.
How much does an Additional Active Problem or
an Additional Active Pathological Gambler Cost
Society? .
This question is best addressed by studying prob-
lem and pathological gamblers directly. However,
estimates derived from this sample may be biased
because only a small fraction of P&P gamblers
seek formal treatment. If those who seek help
impose the greatest costs on society, our cost
estimates of P&P gamblers would be overstated.
What is the Life Cycle of a Problem and
Pathological Gambler?
For example, when casino gambling becomes
available for the first time, what is the behavioral
time profile for individuals who enter and leave
the states of problem and pathological gambling?
Do individuals begin with a period of increasing
gambling dependence, move through a period of
problem gambling, progress to pathological gam-
bling, seek treatment (or withdraw unilaterally
from the problem), and abstain thereafter? Or are
there relapses and continued problems if treat-
ment is not sought. This information could be
used to predict how many currently active prob-
lem and pathological gamblers to expect for given
population as a function of the availability of
casino gambling.
What Effect do Different Types of Treatment
have on Problem and Pathological Gamblers?
Such information would help people to know how
to efficiently allocate funding resources for treat-
ment interventions.
How can Casino Gambling be Offered to
Minimize its Social Costs?
Quinn (2001) discusses many possible ways of
offering casino gaming to reduce social costs. To
evaluate the effectiveness of these interventions
and their impact on casino benefits one would
need to estimate the dasticity of both P&P and
non-P&P gamblers to such actions.
What are the Net Profit and Tax Benefits of
Increasing Casino Gambling?
Rather than estimating a true social benefit, many
studies estimate only the gross increases in profits
or only weight the increased benefits to local firms
while ignoring lost profits to other firms.
Manage. Decis. Econ. 22: 143-162 (2001)
.
.
BUSINESS PROFITABILITY VERSUS SOCIAL PROFITABILITY
\Vhat are the Distance Benefits of Increasing
Casino Gambling?
To date only one study examines this importan~
question. Testing the robustness of this result will
provide more insight into this understudied area.
Focusing future research questions and
methodologies on a clearly formulated theoretical
foundation will allow us to make our estimates of
both the costs and benefits of casino gaming more
preCise,
APPEI\'DIX A
A Study of the Economic Impact of the Gaming
Industry Through 2005, by The Evans Group: A
Partial Critique
International Game Techno]ogy (. IGT), a man-
ufacturer of computerized casino gaming products
and video gaming machines, and operator of pro-
prietary gaming systems, commissioned The
Evans Group, an econometric consulting firm, to
produce a study of the impact of the gambling
industry in ]996. The 9 September 1996 press
release for the resulting report entitled A Study of
the Economic Impact of the Gaming Industry
through 2005 issued by . IGT reported,
States and localities that permit casino gaming
ha ve improved their overall economic perfor-
Unemploymenl
Role
%
]57
mance . , . The sludy.., reports that wherever
casinp gaming has been implemented, employment
has r.isen, unemployment fallen, and additional tax
revenues have been gt:nerated. (Emphasis a?ded.)
The Evans study describes impacts for individual
states, We will briefly examine the findings related
to Illinois, a state with which the authors are
familiar. On page 4-3 the report states:
Based on these data, it would appear that the
opening of a casino reduced the unemployment rate
in that county in both the year it was opened and
in the following year. The average employment in
these eight counties...implies a total of 37000 ex-
tra jobs. These multiplier figures are much higher
than ordinarily obtained, and employment in these
counties might have risen for other reasons as
well. Nonetheless, the figures do indicate that
casino gaming has been a boon to these counties,
especially those that are more rural. (Emphasis
added.)
Most casinos opened after 1991. The period
1991-1996 covered by the study, therefore, coin-
cided with the nationwide economic expansion
coming out of the recession of 1990-1991. Em-
ployment was rising and unemployment was
falling in many counties, with or without the
introduction of casinos, The authors, therefore,
were right to feel uneasy. Their caution that 'em-
p]oyment in these counties might have risen for
other reasons' shows they knew that simple be-
fore-and-after comparisons finding declining un-
employment and increasing employment proved
~"'''
ko It!:, <St. IY.
l:'.(. i "q,j, Q III
<J'l 0.., "II,.
"Ill"
)q .10 , .~
~p... <-!so. J'J'q
ell ~/eJ'... "
. UnemplOymenl role II' veor 01 coslnos ~ 1994 unemploymenl role
Figure 1. The Evans Group study, reproduced figure 4,.1.
. Copyright (Q 2001 John Wiley & Sons, Ltd. Manage. Decis, Econ, 22: 143-162 (2001)
.
158
E,L. GRINOLS AND D,B. MUSTARD
10110'1
.,...
;c.
'1m ~
6,
5J
,1
I
J'
i
I
I~
~
I
ja1tOi]
~
] J J ..
J :% ;'l .. 5 6 7 8 .9 )0 "
la ,.0'1
.....
2
, t
· I J 4
6 7 8 9 to II
[ii1oOi]
~
l ~'
I'~'''' :
.,-
.
R
~
:1
i
j
!fiiloGl
l~
I
Figure 2. Casino counties are indistinguishable from non-casino counties.
.
nothing about the effects of casinos in a country
recovering from recession. Figure I reproduces
figure 4-1, provided in the original study. The
authors explain that the observed drop in casino
county unemployment rates exceeded the state
average by 0.3 and 0.2 percentage points on aver-
age in the first and second year after introduction.
The authors' conclusions are noted above. The
rest of the story is provided below.
The study gives the impression that counties
that opened casinos experienced better economic
performance than those that did not. However,
Illinois contains 102 counties. We can select other
counties that had the same unemployment rate
(within 0.1 percentage point) as the casino county
in the inital period and compare their peformance
directly. This is done in Figure 2. As shown there,
the unemployment rate dropped in all counties
with similar initial unemployment. Some counties
did better than casino counties, some counties did
worse. From left to right, bottom row first, the
casino counties are numbers 6, I, 3, 2, 2, 3, 7, 3.
Nineteen counties performed better than their
casino cousin, while 19 performed worse.
A statistical test confirms that the drop in
unemployment of casino counties is statistically
insignificantly different from the drop experienced
by the comparable non-casino counties shown in
Figure 2. Let d U denote the change in county
unemployment rate minus the change in state
unemployment rate for the same period, and let
.
Copyright <9 2001 John Wiley & Sons, Ltd.
Casino identify counties that introduced casinos in
the initial period (Casino = 1 if a county intro-
duced a casino, 0 otherwise). Then running the
following regression,
d U = a + b Casino + B
reveals that coefficient b is 2.75 (consistent with
the 0.2 and 0.3 percentage point differences re-
ported by The Evans Group), but with a standard
error of 0.856 implying a p-value of 0.4, Coeffi-
cient b is, therefore, statistically indistinguishable
from 0 at conventional levels. '
NOTES
1. Gambling revenue is the net amount of money that
the gambling operator extracts from patrons. It
,-equals the 'handle' (gross amount wagered-
which may reflect the same chips being bet many
times before it is ultimately retained or lost) less
payouts, prizes, or winnings returned to players.
For example, if players wager $1000000 on out-
comes of a roulette wheel over the course of an
evening, and $880 000 is returned to them as win-
nings (some roulette slots are reserved for the
house), then operator revenue is $120000.
2. According to the Indian Gaming Regulatory Act of
1988, Class I gambling consists of 'social games
solely for prizes of minimal value'. Included in
Class I gambling are traditional Indian games iden-
tified with tribal ceremonies and celebrations, Class
II gambling includes bingo and 'games similar to
bingo'. Class III gambling includes 'all forms of
Manage. Decis. Econ. 22: 143-162 (2001)
.
BUSINESS PROFITABILITY VERSUS SOCIAL PROFITABILITY
159
.
gaming that are not Class I gaming or Class II
gaming', such as blackjack, slot machines, roulette,
and other casino-style games,
3, Kindt (1994), Grinols (1996), Grinols and Omorov
(1996) and Henriksson (1996) discussed a number
of these.
4. Public Law 104-169 of the 104th Congress estab-
lished the NGISC. For more information about its
mission, composition and findings see http://
www.ngisc.gov/.
5, We show below that both concepts are necessary to
a proper cost-benefit assessment of casinos.
6, We will show below that it should be present.
7. We follow throughout the paper standard general
equilibrium accounting conventions for describing
inputs and outputs in consumption and production,
8. That is, $ 100 of utility is defined to be the utility
that can be achieved by optimally spending $100 at
prices p, with nearest casino d miles away,
9. For example, the returns to playing roulette, slot
machines, or a blackjack game are approximately
the same regardless where offered. The framework
could be modified to allow for different qualities of
gambling, In this case the model would deal with
multiple, imperfectly substitutable goods.
10. The transfer of wealth in gambling is generally
from relatively poor to relatively wealthy, There-
fore, if a dollar generates more utility for rich than
poor, our assumption understates the social bene-
fits, If a dollar generates more utility for the poor
than the rich, oui' assumption understates the social
costs of casinos.
11. A positive element of Yj denotes output of a good
or service, and a negative component denotes the
use of an input.
12. Although it is not central to our objective in this
paper, we include z to be consistent with the gen-
eral framework we develop, Excluding z does not
affect the central arguments of this paper, Compo-
nents of z are economy excess demands for traded
goods. A zero denotes a non-traded good, while a
positive entry denotes imports.
13, Use the fact that Li Oij = I.
14, In 1998, gambling revenues were approximately
0,5% of GDP and casino revenues were approxi-
mately 0.25%.
15. It is conceivable, of course, in certain circumstances
that the introduction of casinos could change prices
enough to matter to local residents. For example, if
casinos increased employment and the local popu-
lation, the demand for local housing would in-
crease, thus raising housing prices and creating
capital gains for residents. In such cases, however,
the reduction in demand for residential property
and capital losses in the areas from which the new
residents came would have to be taken into ac-
count. Over time, if new housing responded to the
increased demand, the prices of the existing stock
of housing would decrease, Because gambling
doesn't create new people, but only moves them
from one place to another, a reasonable first ap-
proximation is that the net effect of gambling on
.
Copyright @ 2001 John Wiley & Sons, Ltd.
capital gains and consumer surplus considerations
would be small.
16. We presume that the jobs being compared in two
locations are comparable. Blair el al. (1998) argued
that 'employees in gaming industry occupations are
less satisfied with their jobs than those in other
industries'. If jobs are different in two locations,
then the jobs would appear in the formula as
different because workers would demand compen-
sating wage differentials, and this would affect
profitability, If compensating wage differentials do
not arise, but workers face non-market constraints
that cause them to work hours that are not optimal
given the wages paid, these costs would appear in
the unemployment terms of Equation (5),
17. For industry revenue data, see International Gaming
and Wagering Business (1999, p, 24).
18, The value of tobacco grown each year is $39 bil-
lion. Encarta Encyclopedia, http://encarta.msn.com/
find/Concise.asp?ti = 02A43000 # s 12.
19, The proposal was for West Dundee, Illinois. The
study reported, 'Both Christiansen/Cummings and
Mirage Resorts estimate local gaming demand by
applying gaming win per capita factors to the popu-
lation residing within concentric circles of a gaming
venue, The factors decline as distance increases.
The $200 win per capita applicable to the 0-50 mile
segment was developed jointly by representatives of
Mirage Resorts and Dr Cummings to apply to the
local population in the New Orleans environs in a
1992 evaluation of the New Orleans gaming
market.'
20. See, for example, GLS Research (1994) Clark
County (Las Vegas, Nevada) Residents Study
1993-1994, Even in Las Vegas, one-third of the
population does not gamble.
21. For example, a study of wagers in Minnesota
(Smith and Craig, 1992; Tice, 1995) found that 1%
of gamblers accounted for 50% of wagers, and that
10% accounted for 80%. An Illinois study (Gazel
R, Thompson WN. 1996. Casino gamblers in Illi-
nois: who are they? Manuscript, I -25 (plus data
supplied by the authors)) found that 10% of bettors
accounted for 66% of wagers, Heavy gambling is
not the same as problem and pathological gambling
even though the revenues of P&P gamblers figure
disproportionately among the revenues of the--
highest-gambling segment of the population. When
compared with the population at large, the amount
gambled by P&P gamblers implies that the share of
casino revenues from problem and pathological
gamblers can be as much as 1/4 to 1/2 of casino
revenues (see Grinols and Omorov, 1996), Lesieur
(1998b) reported that 48.7% of casino revenues in
Nova Scotia came from problem gamblers, and
that 55% of revenues for casino cards and dice
games came from problem gamblers in Washing-
ton. In other locations he found that percentages
ranged between 26,7 and 41.4%, In Montana, 37%
of the revenues of video gambling machines was
estimated to come from problem and pathological
gamblers (Polzin et al., 1998), The Productivity
Manage. Decis. Econ. 22: 143-162 (2001)
.
160
E,L. GRINOLS AND D,B. MUSTARD
.
Commission (1999) reported that problem gamblers
account for 2.1 % of the adult population but one-
third of all gambling revenues in Australia. Volberg
et ai, (200 I) also examined the distribution of rev-
enue from different types of gamblers.
22, This description applies in Illinois and many other
midwestem states, In Minnesota, for example, only
American Indians operate casinos, In locations
such as Atlantic City or the Gulf Coast of Missis-
sippi, regulations allow entry to all as long as
certain operating requirements are met. In these
locations competitition drives economic profits to
zero,
23. Non-Indian casinos paid over $2 billion in taxes to
the various states on gaming revenues in 1997. CT's
two Indian casinos paid $236 million to the state
that year. In comparison, states generated revenues
of approximately $10 billion from net proceeds of
lotteries in 1997, or $51.15 per adult.
24, See Economic Report of the President, 1999, Table
B-15, column 8.
25. How should we treat demand derived from addic-
tion? If addiction is not rational then its derived
demand should be treated differently, We, there-
fore, report both figures above, In the lower figure,
we assumed that 32% of casino revenues are from
P&P gambling.
26, A survey of this literature and list of references can
be found in Adam Rose and Associates (1998) and
the NGISC (1999), appendix 5 on Economic
Development.
27. Leven et al. (1998) provide an example of how the
focus on job creation may mislead the unwary or
untrained. They wrote,
'This study seeks to take an objective look at the
economic impact of the gaming industry on the
Missouri economy. Where do the gaming rev-
enues come from? How are they redistributed in
the economy? By how much do state and local
governments benefit? What is the net bottom-line
economic impact? . . . [Gaming] does add spend-
ing, income, and jobs to the Missouri economy.
It should be addressed in this context.'
While the authors do not claim that the answers to
their questions constitute a cost- benefit evaluation,
their plea that gambling adds 'spending, income,
and jobs to the Missouri economy' and that 'it
should be addressed in this context' could easily be
misinterpreted to mean that a calculation of in-
come, jobs, and employment is synonymous with a
cost-benefit evaluation. In their summary (p. 75)
they wrote:
..
'The focus of this study has been the determina-
tion of whether net new output (and jobs and
employment) have been created state-wide in
Missouri as a consequence of casino gaming
operations, and if so how much. . . The 'bottom
line' is that significan t additions to the Missouri
economy have been achieved, As of 1997, almost
Copyright ~ 2001 John Wiley & Sons, Ltd.
18 000 net new jobs, $500 million in added per-
sonal income, and over $750 million of added
output have benefited the state's economy.'
Who would argue with such figures? Or be aware
that regardless of their accuracy, casinos in Mis-
souri might fail to pass a cost- benefit test and thus
be harmful to state welfare?
28. See Grinols et ai, (2000) for a complete review of
this literature, Each of the following crimes has
been alleged in the literature to be associated with
gambling. Index I Violent Crime (Aggravated As-
sault, Robbery, Rape, Murder), Property Crime
(Larceny, Burglary, Auto Theft), and non-Index I
crime such as Embezzlement and Employee Theft,
Loan Fraud, Insurance Fraud, Forgery (including
check forgery), Tax Evasion, Tax Fraud, Con
Games (Swindles, Hustling Cards, Dice or Other
Games), Bookmaking, Working in an Illegal Game,
Pimping, Prostitution, Selling Drugs, and Fencing
Stolen Goods.
29, Lesieur (1992, p. 45) and Executive Office of the
Governor (1994, p, 67).
30, National population data by age cohort are on the
US Census Bureau website http://www.census.gov/
populationjestimatesjnationjintfile2-l.txt. As of 25
August 2000, the US had a population of
275 130000, Moreover, 196649000 were aged 20
or older.
31. See Table 2.
32. See also Frank et ai, (1991),
33, Representative of such cases is the following ac-
count, 'A Florida man who lost about $50000
while gambling here (Atlantic City) during the past
two days died Tuesday after he jumped seven floors
from a Trump Plaza Hotel and Casino roof onto
Columbia Place, officials said'. Brian Hickey, Staff
Writer, 18 August 1999, South Jersey Publishing
Co,
34, The minimum social costs of this category are the
value of the resources spent by those trying to steal
and cover up their offenses and the value of the
resources spent by potential victims to decrease
their likelihood of being victimized. There may be
another component of cost in addition. however.
Social costs can be higher if the original owners of
the property value it more than the offenders do.
For example, if the owners valued their property at
$1000 and the offenders who stole it sold it to
someone who valued it at $300, there would be an
additional social loss of $700, Furthermore, if the
thief is a pathological gambler and spends the
wrongly acquired $300 gambling, his expenditures
may reflect addiction rather than rational choice.
In that case there would be social cost equal to
some or all of the $300 because of his sub-optimal
allocation of resources to the gambling sector. Last,
although there is some debate about whether to
count stolen dollars as costs to all of society (which
includes the thief) because 'the thief gets the
money', it is clear that the non-gambling portion of
society will be made worse off by such actions, and
losses to the rest of society are important in the
Manage. Decis. Econ. 22: 143-162 (2001)
.
BUSINESS PROFITABILITY VERSUS SOcrAL PROFITABILITY
161
.
policy debate because they suggest that all of the
abused dollars represent social costs to the non-
gambling sector. .
35. Westphal et al. (1999) is not used in Table 2, but
supplements Ryan et al. (1999), The South Dakota
Research Council study was completed in 1998, but
addenda were added in 1999. See also Finance and
Administrative Cabinet, Commonwealth of Ken-
tucky (1999), Florida Department of Law Enforce-
ment (1994), Florida Sheriffs Association (1994),
Iowa Racing and Gaming Commission (1995),
Lesieur (1998a).
36. As an alternative way of showing that the differences
in the totals are driven largely by the number of cost
categories estimated, we compared the totals after
'filling the gaps' in each study using the average cost
for a given category from those studies that did
estimate those particular costs, When doing so, the
variance in the totals decreased substantially, The
lowest totals were for South Dakota Legislative
Research Council (1998-1999) and Thompson and
Quinn (1999), $7396 and $8047, respectively. The
largest were $25 742 by the Executive Office of the
Governor (1994) and $18 203 by Thompson et ai,
(1998),
37, In addition to our use of the numbers, some studies,
such as Thompson et ai, (1998) intentionally formed
their original estimates conservatively to understate
costs.
38, Including the nominal value of this study would
increase the cost estimate for three of the four costs
it estimates, Using the values adjusted for 19 years
of price level changes would have significantly in-
creased the estimates of all four costs.
39. See Table 5, p, 34.
40. Our highest estimate of benefits was $75; our lowest
estimate of costs $140. Applying the per adult costs
of $221 from Table 4 to the estimate of benefits
adjusted for P&P gamblers of$34 implies that casino
gambling fails a cost- benefit test by a ratio of 6.5: 1.
41. NGISC (1999, p. 4-4).
REFERENCES
.
Adam Rose and Associates. 1998. The regional economic
impacts of casino gambling: Assessment of the literature
and establishment of a reasearch agenda. Report pre-
pared for the National Gambling Impact Study Com-
mission (August).
Blair BF, Schwer RK, Waddoups Jc. 1998, Gambling
as an economic development strategy: the neglected
issue of job satisfaction and nonpecuniary income, The
Review of Regional Studies 28(1): 47-62,
Deloitte and Touche. 1992. Economic and other impacts
ofa proposed gaming, entertainment and hotel facility.
Prepared for the City of Chicago Gaming Commis-
sion.
Eadington WR. 1999. The economics of casino gam-
bling, Journal of Economic Perspectives 13(3): 173-
192.
Copyright ~ 2001 John Wiley & Sons, Ltd.
The Evans Group. 1996. A study of the economic impact
of the gambling industry through 2005. Study commis-
sioned by International Game Technology, Septem-
ber, iii + 105,
Executive Office of the Governor. 1994. Casinos in
Florida. Office of Planning and Budgeting: The Capi-
tol. Tallahassee, FL.
Finance and Administrative Cabinet, Commonwealth of
Kentucky. 1999, Assessment of the economic and
social impact of expanded gaming in the common-
wealth and neighboring region. December, iii + 182.
Florida Department of Law Enforcement. 1994. The
Question of Casinos in Florida: Increased Crime: Is it
Worth the Gamble? State of Florida: Tallahassee, FL.
Florida Sheriffs Association. 1994. Casinos and crime: is
it worth the gamble? A summary report and position
paper. Florida Sheriffs Association: Tallahassee, FL.
Frank ML, Lester D, Wexler A, 1991. Suicidal behavior
among members of gamblers anonymous. Journal of
Gambling Studies 7: 249-254,
Gerstein D, Murphy S, Toce M, Volberg R, Harwood
H, Tucker A, Christiansen E, Cummings W, Sinclair
S. 1999, Gambling impact and behavior study: report
to the National Gambling Impact Study Commission.
National Gambling Impact Study Commission, (1
April).
GLS Research, 1994. 1993-1994 Clark County Residents
Study. Las Vegas Convention and Visitor's Authority:
Las Vegas, NV,
Grinols EL. 1996, Incentives explain gambling's growth.
Forum for Applied Research and Public Policy 11(2):
119-124.
Grinols EL. 1999. Distance effects in consumption,
Review of Regional Studies 29(1): 63-76,
Grinols EL, Mustard DB, Dilley CH. 2000, Casinos,
Crime, and Community Costs. University of Illinois:
University of Georgia; 1-35,
Grinols EL, Omorov J, 1996. Development or dreamfjeld
delusions?: assessing casino gambling's costs and ben-
efits, The Journal of Law and Commerce 16(1): 49-87.
Henriksson LE. 1996, Hardly a quick fix: casino gam-
bling in Canada. Canadian Public Policy XXII(2):
1 ]6-128.
International Gaming and Wagering Business, 1999.
August.
Iowa Racing and Gaming Commission. 1995. Des
Moines Register. 13 October. Christiansen/Cummings
Associates: Arlington, MA,
Kindt JW, 1994, Increased crime and legalized gambling
operations: the impact on the socio-economics of
business and government. Criminal Law Bulletin, 43:
538-539,
Lesieur H, 1992. Compulsive gambling. Society 29(4):
43-50.
Lesieur H. 1998a, Testimony before the National Gam-
bling Impact Study Commission, Atlantic City, New
Jersey, 22 January 1998, Institute of Problem
Gambling.
Lesieur H, 1998b. Costs and treatment of pathological
gambling. The Annals of the American Academy of
Political and Social Science (Gambling: socioeco-
Manage. Decis. Econ. 22: 143-162 (2001)
.
.
.
162
E,L. GRfNOLS AND D.B, MUSTARD
nomic Impacts and public policy, Frey JH, special
editor) 556: 153-171.
Leven C, Phares D, Louishomme C. 1998, The eco-
nomic impact of gaming in Missouri. Report to Civic
Progress, St. Louis. April.
Maryland Department of Health and Mental Hygiene,
Alcohol and Drug Abuse Administration. 1990. Task
Force on Gambling Addiction in Maryland, Maryland
Department of Health and Mental Hygiene, Alcohol
and Drug Abuse Administration: Baltimore, MD,
McCleary R, Kenneth C, Feng W, Merrill V, Napoli-
tano C, Moles M, Graffeo B. 1998, Suicide and
gambling: An analysis of suicide rates in US counties
and metropolitan areas. Report to the American
Gambling Association. Manuscripts.
Mirage Hotel. 1993, Response of the developers of the
proposed Fox River resort to the West Dundee river-
boat project task force.
National Gambling Impact Study Commission. 1999.
National Gambling Impact Study Commission Final
Report www.ngisc.gov (accessed 18 June 1999].
Phillips DP, Welty WR, Smith MM. 1997, Elevated
suicide levels associated with legalized gambling. Sui-
cide and Life-Threatening Behavior 27: 373-378.
Politzer RM, Morrow JS, Leavey SB. 1981. Report on
the societal cost of pathological gambling and the
cost-benefit/effectiveness of treatment. The Johns
Hopkins Compulsive Gambling Counseling Center,
Presented at Fifth National Conference on Gambling
and Risk-Taking.
Polzin PE, Baldridge J, Doyle D, Sylvester JT, Vol berg
R, Moore WL. 1998. Final Report. Presented to the
Montana Gambling Study Commission, 30 September
1998, Bureau of Business And Economic Research
(University of Montana-Missoula).
Productivity Commission. 1999. Australia's gambling
industries. Final Report, Summary (26 November,
. Report No. 10).
Quinn FL. 2001. First do no harm: What could be done
by casinos to limit pathological gambling, Manage-
rial and Decision Economics 22: 133-142.
Ryan TP, Speyrer JF, with Beal ST, Burckel DV,
Cunningham BR, Kurth MM, Scott LC, Wall lL,
Copyright (Q 2001 John Wiley & Sons, Ltd.
Westphal JR. 1999, Gambling in Louisiana: a benefit/
cost analysis. Prepared for the Louisiana Gaming
Control Board, Louisiana State University Medical
Center (April).
SMR Research Corporation. 1997. The Personal
Bankruptcy Crisis, 1997: Demographics, Causes, Im-
plications & Solutions. SMR: Hackettstown, NJ.
Shaffer HJ, Hall MN, Vander Bilt J, 1997. Estimating
the Prevalence of Disordered Gambling Behavior in the
United States and Canada: A Meta-Analysis. Harvard
Medical School: Boston.
Smith F, Craig WJ. 1992. Whos in for how much?
CURA Reporter. Center for Urban and Regional
Affairs 22: 11-14, 16 (plus data supplied to the
authors),
South Dakota Legislative Research Council. 1998. Eco-
nomic and fiscal impacts of the South Dakota gaming
industry. 1-146 + Attachment: Letter of 7 January
1999 to Mr. Terry Anderson, Director, South Da-
kota Legislative Research Council, Table South Da-
kota Total Estimated Incremental Social Costs.
Thompson WN, Gazel R, Rickman D. 1996. The social
costs of gambling in Wisconsin. Wisconsin Policy
Research Institute Report 9(6): 1-44,
Thompson WN, Gazel R, Rickman D, 1998, Social
costs of gambling: A comparative study of nutmeg
and cheese state gamblers. Twelfth National Confer-
ence on Problem Gambling, NY 1998. Gaming Re-
search and Review Journal 5(1): 1-15.
Thompson WN, Quinn FL. 1999, An economic analy-
sis of machine gambling in South Carolina. The
Education Foundation of the South Carolina Policy
Council, May 1999,
Tice DL. 1995, Why the press hates gaming, Casino
Executive August: 57-58,
Volberg R, Gerstein DR, Christiansen EM. Baldridge
J. 2001. Assessing self-reported expenditures on gam-
bling. Managerial and Decision Economics 22: 77-
96.
Westphal JR, Johnson LJ, Stevens L. 1999, Estimating
the social costs of gambling disorders in Louisiana
for 1998, Study Performed Under Subcontract From
University Of New Orleans, UNO # 313-204-124 (23
March),
Manage. Decis. Econ. 22: 143-162 (2001)
TAB
"4"
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 1 of 8
.
1:1.....
~
~t
~.l.
.~
',-
':'-1
~
!E
I~~
~1'
i:::
=
i::::
t.:,.:..
.c:
t....
,t;;..
rc
~(:
.~
'IS
i]
t~:.
t=:
to:;
.;:1.~
..-
?$
...r~
..c::
~l
("
c1\
. :!'1,
.il.'
~==
1.-'
.1!:.
.'.'
:.:z:
;:i
z::;
~
.....
~
" ~ _, ~iJ~f~~11~.,~~:~,,:~;, i::p\\~>~~-;>~ .
.E A'S Y. ill' 0 LX E \:! .:', ._'~' ',-~~":',t~{:;~~r~:~~;'~. ., rtl0NTUNE
I,nn;,' j..in th.. .li,,'u.~i"ll ,,11,,1 ~j',' chI." n.loI" "I'" ,11I.1 n,l) 1Ilt,'~\'i"I" Ihllt.lil'" . "i1~'b ~ ,tOIl:> ':
,e' ~rl\'~r! (~~r"~ (\ (: .
~,~~~AMBErnG}F:AGTS1 &,,STATS
An excerpt from:
The Business-Economic Impacts of Licensed Casino Gambling in
West Virginia: Short-Term Gain but Long-Term Pain
By John Warren Kindt
John Warren Kindt is a professor at the University of Illinois at
Urbana-Champaign. A faculty member since I 978, he teaches
courses in commerce and legal policy. The text and footnotes
have been updated from the original publication in The National
Impact of Casino Gambling Proliferation: Hearing Before the
House Committee on Small Business, 103rd, Congress, 2nd,
Session 77 (1994). Due to the fonnat of this publication,
substantial footnotes have been deleted.
The Economic Development Argument Exposed
From a business-economic perspective, the main issue involved
in legalizing various fonns of gambling is whether gambling
activities constitute a valid strategy for economic development.
While the dollars invested in various legalized gambling projects
and the jobs initially created are evident, the industry has been
criticized for inflating the positive economic impacts and
trivializing or ignoring the negative impacts (Goodman 1994).
The industry's tendency to focus on specialized factors provides a
distorted view of the localized economic positives, while ignoring
the strategic business-economic costs to the state as a whole
(such as West Virginia) and to different regions of the United
States (California Governor's Office 1992, Kindt 1995). In 1994,
all of the various experts who testified before the U.S. House of
Representatives Committee on Small Business criticized the
impacts that casino-style gambling activities inflict upon the
criminal justice system, the social welfare, system, small
businesses, and the economy (Congressional Hearing 1994).
Utilizing legalized gambling activities as a strategy for economic
development was thoroughly discredited during the hearing.
Florida is the only state which has conducted a comprehensive
statewide analysis of the impacts of legalized gambling activities.
Its report concurred with the congressional hearing's conclusions
(Florida Budgeting Office 1994).
.
Since some issue areas have not received widespread public
attention in West Virginia, this analysis highlights some of the
neglected issue areas as they relate to tax revenues, social-welfare
costs, education, and job creation.
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 2 of 8
EconomicCycles and Gambling's Impact on Tax Revenues
.
From the perspective of U.S. economic history, the United States
has had previous economic cycles with widespread legalized
gambling activities. The most relevant cycle occurred after the
American qvil War and paralleled the post-bellum migration to
the "Wild West." Although gambling proliferated during this
time~frame, within a few years the trend toward prohibiting
gambling activities had begun, and by 1910 there was virtually
no legal gambling in the United States. Gambling activities were
not just prohibited via state statutes and local ordinances, but
more importantly, these prohibitions were incorporated into most
state constitutions. The fact that state constitutional provisions
were utilized to make it as difficult as possible for future
generations to legalize gambling activities (and thereby
experiment once again with a classic "boom and bust" economic
cycle) lends substantial credence to arguments that both
historically and currently, the legalization of gambling activities
eventually causes: (I) increased taxes, (2) a loss of jobs from the
overall region, (3) economic disruption of other businesses, (4)
increased crime and (5) large social-welfare costs for society in
general and govemment agencies in particular. For example, two
studies of the riverboat casinos in Illinois concluded that for
every one job created by the riverboats, most of the surrounding
communities probably lost one or more jobs from pre-existing
businesses (Grinols 1994; Grinols and Omorov 1995).
.
In recent economic history, legalized gambling activities have
been directly and indirectly subsidized by the taxpayers. The
field research throughout the nation indicates that for every dollar
the legalized gambling interests indicate is being contributed in
taxes, it usually costs the taxpayers at least 3 dollars-- and higher
numbers have been calculated (Politzer, Morrow and Leavey
1 981; Better Government Association 1992; Florida Budget
Office 1994). These costs to taxpayers are reflected in: (I)
infrastructure costs, (2) relatively high regulatory costs, (3)
expenses to the criminal justice system, and (4) large social-
welfare costs (Illinois Governor's Office 1992). Accordingly,
several state legislators (e.g., in South Dakota) have called for at
least partially internalizing these external costs by taxing all
legalized gambling activities at a straight 50 percent tax rate.
.
Furthennore, as a matter of good public policy, state officials and
legislators in lllinois have proposed legislation to prohibit
contributions by legalized gambling interests to politicians and
political campaigns. In the case of casinos, New Jersey already
has such prohibitions, but other states have neglected to enact
similar prohibitions. Political scientists have raised concerns that
the newly developing constituencies in the licensed gambling
industry are becoming so widespread that the industry can dictate
economic, social, and tax policies. For example, the industry
drafted a state constitutional referendum in Florida which would
have mandated the introduction of casinos into communities even
if a particular community voted unanimously against a casino
(Dyckman 1994). The industry spent approximately $3 million to
get the Florida referendum on the ballot and $6.5 million to
campaign for the casinos-- more than the combined gubernatorial
campaigns of Governor Lawton Chiles and his challenger Jeb
Bush (Lavelle 1994). In these contexts, an article in the Columbia
Journalism Review cautions the news media "flat out ask
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 3 of 8
.
[experts, academics, and even other reporters) if they make
money off the industry" (Simurda 1994).
Social Welfare Costs
Legalized gambling activities act as a regressive tax on the poor
(Clotfelter and Cook 1989). Specifically, the legalization of
various forms of gambling activities makes "poor people poorer"
and can dramatically intensify many pre-existing social-welfare
problems. Demographic analyses reveal that certain
disadvantaged socioeconomic groups tend to gamble
proportionately greater amounts of their overall income and
marketing efforts, particularly by state lotteries, have allegedly
been directed at these target groups.
.
In a specific example involving casinos, a 1995 Wisconsin report
concluded that "[w)ithout considering the social costs of
compulsive gambling, the 'rest-of-the-state' areas lose- or,
transfer in- $223.94 million to the local gaming areas.
Considering the lowest estimated social costs of problem
gambling, the rest of... [Wisconsin) loses $318.6 I million to
gambling" (Thompson, Gazel, and Rickman 1995). This report
also concluded that without casino gambling, many local citizens
would have increased participation in other "outside" activities.
"More than 10% of the locals would spend more on groceries if it
were not for the casino, while nearly one-fourth would spend
more on clothes. Thirty-seven percent said that their savings had
been reduced since the casino had opened ..." (l1lOmpson, Gazel,
and Rickman 1995),
.
From the business perspective, businesses are not naive. For
example, "in a rare public stand on a controversial political issue,
the Greater Washington Board of Trade's 85-member board voted
unanimously against" Mayor Sharon Pratt Kelly's initiative to
bring casino-style gambling to Washington, D.C. (emphasis
added, Spayd and Woodlee 1993). With the exception of the
cluster services associated with gambling, new businesses tend
not to locate in areas allowing legalized gambling because of one
or more of the aforementioned costs. In areas saturated with
legalized gambling activities, pre-existing businesses face added
pressures that push them toward illiquidity and even bankruptcy.
Although South Dakota does not constitute a saturated gambling
state, this trend has already been reported there. South Dakota
basically had no gambling in 1988 and then instituted casino
gambling and video lottery terminals by the end of 1989. Within
two years legalized gambling a,ctivities constituted one of the
leading causes of business and personal bankruptcies among
South Dakota residents (whereas this cause was virtually
nonexistent in 1989) (Nelson 1993). More subtly, traditional
businesses in communities which initiate legalized gambling
activities can anticipate increased personnel costs due to
increased job absenteeism and declining productivity (Kindt
1994a). The best blue-collar and white-collar workers, the Type-
A personalities, are the most likely to become pathological
gamblers (Kindt 1994b). A business with 1,000 workers can
anticip~te increased personnel costs of $500,000 or more per
year-simply by having various forms of legalized gambling
activities accessible to its workers.
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 4 of 8
.
To some extent businesses must already internalize the societal
costs associated with assisting personnel with drug or alcohol-
related problems. Legalizing various gambling activities
increases the number of problems related to pathological
gambling in the context of the workforce, and these costs are
reflected in increased personnel costs-- such as "rehabilitation
costs..." which can easily range from $3,000 to $20,000 (or more)
per pathological gambler (Kindt I 994b). In the context of the
healthcare debate, the spectre of these unanticipated costs can
raise further concerns to businesses already being asked to bear
certain health care costs.
Education Costs
Gambling activities and the gambling philosophy are directly
opposed to sound business principles and economic development.
Legalized gambling activities also negatively affect education-- -.
both philosophically and fiscally (Better Government Association
1992; Clotfelter and Cook 1989). Adherence to a philosophy of
making a living via gambling activities not only abrogates the
perceived need for an education, but also reinforces economically
unproductive activities (and is statistically impossible since the
"house" always wins eventually). In states with legalized
gambling activities which were initiated allegedly to bolster tax
revenues to "education," the funding in "real dollars;' has almost
uniformly decreased.
.
The Pathological Gambler Problem
States which embrace legalized gambling activities can expect
enormous socioeconomic costs and a decline in the quality of
life. Unlike traditional business activities, legalized gambling
activities cater to a market consisting of addicted and potentially
addicted consumers, and most pre-existing traditional businesses
will fmd it quite difficult to compete for "consumer dollars"
which are being transformed into "gambling dollars." For
example, the field research strongly suggests that the introduction
of widespread legalized gambling in South Dakota, including
casinos and video lottery terminals (VL Ts), over a two-year time
span caused a one-percent increase in the number of problem and
probable pathological gamblers-- a recognized addictive behavior
pursuant to the American Psychiatric Association (Clotfelter and
Cook 1989; Better Government Association 1992). Each newly-
created pathological gambler has been calculated to cost society
from $13,200 to $52,000 per year (Maryland Department of
Health 1994; Better Government Association 1994). These costs
are not just reflected in society as a whole, but impact on all
businesses. In particular, small businesses could easily
experience disproportionate negative impacts, and unlike large
corporations, small businesses would be less likely to have the
asset base necessary to cushion against those negative impacts,
.
Sociologists almost uniformly report that increased gambling
activities which are promoted as sociologically "acceptable" (the
acceptability factor) and which are made "accessible" (the
accessibility factor) to larger numbers of people will increase the
number of pathological gamblers (Goodman 1994: Politzer,
Morrow and Leavey 1981; Better Government Association 1992;
Maryland Department of Health 1994). The baseline of
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 5 of 8
.
pathological gamblers as part of the population begins at .77
percent as reported by the 1976 U.S. Commission on Gambling
(U.S. Commission 1976). Since gambling has been legalized and
made accessible in several states, the range has increased to 1.5
to 5 percent in those states (Alberta Lotteries and Gambling
1994). This phenomenon was specifically confirmed by a 1995
study which concluded that the lifetime probable pathological
and problem gamblers in Iowa increased from 1.7 percent of the
public in 1989 to 5.4 percent in 1995 (Iowa Department of
Human Services 1995, Petroski 1995). Similarly, a limited study
of Native Americans revealed a rate for lifetime probable
pathological and problem gamblers of 14.5 percent in casino
areas (Alberta Lotteries and Gambling 1994). These
developments translate into increases in socioeconomic costs
which must be addressed and absorbed primarily by taxpayers,
but also by businesses, charities, social-welfare organizations and
governmental units.
Negative Impact on Job Creation
.
On a regional level, the combined ranges of these various
socioeconomic costs are so large that they tend to dwarf the
localized economic positives (California Governor's Office
1992). These drains on society could easily translate into a net
loss of jobs on a statewide or regional level. Furthermore, it can
be argued that the combined economic positives and negatives
result in a net negative economic multiplier (Goodman 1994;
Teske and Sur 1991). From the perspective of business-
economics and strategic development, major businesses are and
should be concerned with the trend toward expanding various
fOnTIS of legalized gambling activities. Among other reasons,
nongambling related businesses will not be competing for
consumer dollars or recreational dollars on a "level playing
field," because legalized gambling activities can cater to an
addicted and potentially addicted market segment. Since the U.S.
economy and most state economies are extensive in scope, the
socioeconomic negatives associated with legalized gambling
activities can remain hidden for long periods of time. However,
just because a particular activity is "legalized" by a state
government does not mean that the negative business or societal
impacts have been eliminate-- or even reduced.
Conclusion
.
Increasingly, taxpayers and businesses are beginning to realize
that, as Professor Jack Van Der Slik has summarized for much of
the academic community, state-sponsored gambling "produces no
product, no new wealth, and so it makes no genuine contribution
to economic development" (Van Der Slik 1990). Business-
economic history supports this proposition. The recriminalization
of gambling activities occurred 100 years ago after a brief
gambling boom following the Civil War. Most state legislatures
utilized constitutional provisions to recriminalize gambling,
because lawmakers wanted to make it as difficult as possible for
future generations to experiment with the classic "boom and
bust" cycles and the concomitant socioeconomic negatives
occasioned by legalized gambling activities. To paraphrase
Georg Hegel's common quote, "those who forget the lessons of
economic history are condemned to relive them" (Bartlett 1968).
.
.
.
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 6 of8
References
Alberta Lotteries and Gaming. 1994. Gambling and Problem
Gambling in Alberta. Edmonton, Alberta. January.
American Psychiatric Association. 1994. Diagnostic and
Statistical Manual of Mental Disorders. ("pathological
gambling") Vol. 312.31, pp. 615-618.
Bartlett, John. 1968. Familiar Quotations. 14th. Edition. Boston;
Little Brown.
Better Government Association. 1992. Staff White Paper. Casino
Gambling in Chicago. Chicago, IL.
California Governor's Office of Planning and Research, 1992.
California and Nevada: Subsidy, Monopoly, and Competitive
Effects of Legalized Gambling. ES-I. Sacramento, CA:
California's Governor's Office. December.
Clotfelter, Charles and Phillip Cook. 1989, Selling Hope,
Cambridge, MA: National Bureau for Economic Research,
Harvard University Press.
Congressional Hearing, 1994. The National Impact of Casino
Gambling Proliferation: Hearing Before the House Committee on
Small Business. I03rd Congress, 2nd. Session. Washington, D.C.
Dyckman, Martin. 1994. "Misleading the Public." St. Petersburg
Times. November 1, p. AI3.
Florida Governor's Office of Planning and Budgeting. 1994.
Casinos in Florida: An Analysis of the Economic and Social
Impacts. Tallahassee, FL: Florida Governor's Office.
Goodman, Robert. 1994. Legalized Gambling as a Strategy for
Economic Development. Amherst, MA: University of
Massachusetts-Amherst, Center for Economic Development.
Grinols, Earl. 1994. "Bluff or Winning Hand? Riverboat
Gambling and Regional Employment and Unemployment."
Illinois Business Review (Spring ): 6-11.
Grinols, Earl and 1. Omorov. 1995. "Development or Dreamfield
Delusions?: Assessing Casino Gambling's Costs and Benefits."
University of Illinois, September.
Illinois Governor's Office. 1992. Governor James Edgar, Press
Release. "Governor Warns Land-Based Casinos Could Bring
Crime Surge as well as Overall Loss of Jobs and State
Revenues." September 29 (summarizing several Illinois State
reports ).
Iowa Department of Human Services. 1995. Gambling and
Problem Gambling in Iowa: A Replication Survey. July 28.
Kindt, John. 1995. "U.S. National Security And The Strategic
.
.
.
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 7 of8
Economic Base: The BusinesslEconomic Impacts Of The
Legalization Of Gambling Activities." Saint Louis Law School
Joumal39, p. 567.
Kindt, John. 1994a. "The Negative Impacts Of Legalized
Gambling On Businesses." University of Miami Business Law
Journal 4, p. 93.
Kindt, John. 1994b. "The Economic Impacts Of Legalized
Gambling Activities." Drake Law Review 43, pp. 51, 66 n. 119.
Lavelle, Louis. 1994. "Voters Deal Loss to Casinos; Gambling
Backers Lose Despite $16-5 Million Campaign." Tampa Tribune.
November 9, pp. 1,5.
Maryland Department of Health and Mental Hygiene, Alcohol
and Drug Abuse Administration, 1994. Task Force on Gambling
Addiction in Maryland, Annapolis, Maryland.
Nelson, Todd. 1993. "S.D. bankruptcies down 5 percent: Judge:
Gambling caused most cases." Argus Leader (Sioux Falls, S.D.)
January 15, p. 1.
News Leader (Springfield, Missouri). 1995. "Jeff City rejects
riverboat gambling approved in '92." November 8, p. 1.
Petroski, William, 1995. "Study: More gamblers in jeopardy."
Des Moines Register. August 25, pp. AI, A2.
Politzer, Robert, James Morrow and Sandra Leavey. 1981.
Report on the Societal Cost of Pathological Gambling and the
Cost-Benefit Effectiveness of Treatment. 5th National
Conference on Gambling and Risk Taking.
Teske, Paul and Bela Sur. 1991. "Winners and Losers: Politics,
Casino Gambling, and Development in Atlantic City." Policy
Studies Review 10 (Spring! Summer): 130-137.
Simurda, Stephen F. 1994. "When Gambling Comes To Town."
Columbia Journalism Review (January-February): 36-38.
Spayd, Liz and Yolanda Woodlee. 1993. "Trade Board Rejects
D.C. Casino Plan." Washington Post. September 25, pp. AI, A8.
Thompson, William, Ricardo Gazel, and Dan Rickman 1995. The
Economic Impact of Native American Gambling in Wisconsin.
Wisconsin Policy Research Institute Report.
U.S. Commission on the Review of the National Policy Toward
Gambling. 1976. Gambling in America. Washington, D.C.:
United States Government Printing Office.
Van Der Slik, Jack. 1990. "Legalized gambling: predatory
policy." llIinois Issues (March): 10.
Wartzman, Rick. 1995. "Bayou Backlash: Gambling Is Proving
To Be a Poor Wager For State of Louisiana: Business Is
.
.
.
frontline: easy money: Business-Economic Impacts of Licensed Casino Gambling in Wes... Page 8 of 8
Disappointing And an FBI Graft Probe Roils a Jaded Electorate."
Wall Street Journal. September II, p. I.
Worthington, Rogers. 1995. "Poor get poorer at tribal casinos,
says study of Wisconsin 'garners'." Chicigo Tribune. April 11, p.
I.
ioin the discussion I whal are the odds" I ~amblil1~' pro and con I ~ I ~ I facts & stats I lanes & transcripts I ~
explore FRONTLINE
New Content Copyright Ci 1998 PBS and WGBHlFRONTUNE
mt>
ionline
TAB
"5"
.
GAMBLING IMPACT AND BEHAVIOR STlUDY
Report to the
National Gambling Impact Study Commission
April 1, 1999
.
Submitted By
Dean Gerstein
John Hoffmann
Cindy Larison
Laszlo Engelman
Sally Murphy
Amanda Palmer
Lucian Chuchro
Marianna T oce
Robert Johnson
Tracy Buie
Mary Ann Hill
National Opinion Research Center at the University of Chicago
h tip:/ /www.norc.uchicago.edu
Rachel Volberg
Gemini Research
Northampton, Massachusetts
http://\\'Ww.geminiresearch.com
Henrick Harwood & Adam Tucker
The Lewin Group
Fairfax, Virginia
http://www.lewin.com
.
Eugene Christiansen, Will Cummings, & Sebastian Sinclair
Christiansen/Cummings Associates
New York, New York
.
NORC
...",.
UNlVEnSJn' OF CHIC,.\GO
TABLE OF CONTENTS
LIST OF EXHIBITS ..., ...,.. ....,.,.... .,.............,.,.., ......,........ '...., ...,., ,....................................., ,........ jjj
TABLES. ........,.....,.. ........... .... ,.... , ...............,....., ....., , ,................,........... .,..................,........,......... iii
FIGURES...........,.........,.,...........,............................,......................,...".....................,..........."".... iv
ACKNOWLEDGMENTS......,..........,.,.................,..,.............,......,.,..,.....................................,..... v
PRINCIPAL STAFF ...... ,..... ............., ....,.............., ......... .......................,..... ...., ..,.. ,.. ,. ,..,....... ........... V
NaRC COLLABORATORS ...............,......, .,......................................"....... ..... ..... ,.. ,....,.... ,....,., ...... v
OTHER CONTRIBUTORS .. ..., ,........................,......, .........." ................. .,..... ....... ,..... ........... ,. ,...,.,... vi
HI GHLIG HTS..............................,..,...,................................,....."......,.....,..............,................... viii
BASIS OF FINDINGs.,...,........,..,.....................,.........,.................,.............,.......,.......................... viii
CHANGES IN GAMBLING PARTICIPATION OVER TIME................................................................ viii
PATHOLOGICAL AND PROBLEM GAMBLING................,.............................,................................ viii
YOUTH GAMBLING......,.........,.........,.....,...,..,..,........."...............,.....,............,..,.".".",.,..,..,."..... ix
COMMUNITY IMPACT OF CASINOS".....................................,..........,....,........,............................... x
.
INTRODUCTION .... ........ ...,.. .........,...... ,. ,.................,...,........................... ..,. .,....................... '.." ,. 1
CHAPTER 1. COMPARISON BETWEEN THE RESULTS OF THE 1975 At~D 1998
NATIONAL SURVEYS OF ADULT GAMBLING BEHA VIOR ............................................. 3
METHODS...............",..............,.....",..".......,.......",....,....,.,.."....,...,."..,.............,..... '.,........,......,. 4
OVERALL PREVALENCE RATES ............,..........................,............................................................ 6
DEMOGRAPHICS OF GAMBLERS.... .................., .... ,....,., .... ,.... ,............ ,....... ,., ......... .,........... .,...., ,.. 6
REFERENCES....,...., ...,........... ,.... .., ...,................... .,.........., .... '............ .... ............. '" ......,. ,. .,. ........ 11
CHAPTER 2. THE PREVALENCE AND CORRELATES OF GAMBLING PROBLEMS
AMONG ADULTS .....,........... ,..........,............................... '.... ...........,..... ................ ............ ......... 12
THE SOCIAL CONSTRUCTION OF PSYCHIATRIC TOOLS ............................................................... 12
MEASURING GAMBLING PROBLEMS ........................................................................,.................. 13
Adopting the South Oaks Gambling Screen in population research..................................... 13
Validating the South Oaks Gambling Screen......................................................................., 14
The eclipse oftlIe South Oaks Gambling Screen ..................................................................14
Emergence of a new standard: The DSM-IV....................................................................... 15
DEVELOPMENT OF THE NaRC DSM-IV SCREEN FOR GAMBLING PROBLEMS ("THE NODS").. 17
Validity and reliability of the NODS..................................................................................... 20
The NODS typology ................."........,',..........................................."..".........,..............,...... 20
The role of timeframe,. ,. ...." .....,.."., .... .,......., ...,..,.., .........".." .., ..,.,...." ... ,...."., ..." ,." ..... '" .... 21
PATRON SURVEy..,.......................................,........................,............ ............,.........................,. 22
PREVALENCE RATES..,.,....,..,..,...,..,.......,....................,..".,....,.........................,........,.,.,............. 25
REGIONAL DIFFERENCES AND AVAILABILITy............................................................................ 28
ATIITUDES TOWARD GAMBLING ............................................................................................... 29
CORRELATION WITH OTHER DISORDERS ....................................................................................29
GAMBLING EXPENDITURES ...............................................................,........................................31
ASSESSING PROBLEM AND PATHOLOGICAL GAMBLING IN THE FUTURE..................................... 34
REFERENCES...... ...................................................... ....,..............:.............................................,. 35
.
Gambling Impact and Behavior Study
Page i
.
NORC
..-
UN/VERS/fi OF CHICAGO
CHAPTER 3. ECONOMIC ANALYSIS OF THE CONSEQUENCES OF GAMBLING
PROBLEMS AMONG ADULTS..................................................,............................................, 38
PRIOR STUDIES ON THE COSTS OF GAMB,LfNG ............................................................................ 41
COSTLY CONSEQUENCES OF GAMBLfNG .......,............................................................................ 41
Employment-related impacts .................... .... ,.......,....,...........,.."... ..... ..... ...................... ....... 43
Bankruptcy, debt, unemployment-insurance and welfare .....................................................45
Criminal justice costs.........,....,.......,............,.................,..........,..,..............................,.,....,.. 47
Divorce,........,.,........."..,.....".....".."...........,.,.....,.,......................,..........,.,.............,...,........, 48
Health care,..... ........., ................ ......... ........,...." ............ ...............,.... ......... ........,... ......... ....... 50
Mental health care ............,....,.......,...........".........."..,..."..,.,...,.................... ........ ,............... 51
Treatment for pathological gambling ,.......".........................,..........,.......,...........,................ 51
Total costs of pathological gambling....................................................,..,........,...",,............52
SUMMARY. .................................,..,......"....,....,..........,....,.............................,..,..,............,..,...... 53
ANNEX 1: DESCRIPTION OF OUTCO~[E VARIABLES ................................................................... 55
ANNEX 2: DESCRIPTION OF EXPLANATORY/INDEPENDENT VARIABLES ....................................56
ANNEX 3: METHODOLOGICAL NOTES FOR COSTS ..................................................................... 57
REFERENCES..,.,.........:......,....,....,..,.......,..,........,....,.....,..,.,..........,..,..,......."".,..................,....... 5 8
.
CHAPTER 4. GAMBLING AMONG 16- AND 17-YEAR-OLD yOUTHS.......................... 61
CHAPTER 5. IMPACTS OF CASINO PROXIMITY ON SOCIAL AND ECONOMIC
OUTCOMES, 1980-1997: A MULTILEVEL TIME-SERIES ANALySIS........................... 65
DATA,......,......................,.....,...,......,............".............,.......................,........,.................. ........... 66
METHODS......,....."..............."..,.................,.............................,....................,...................,........ 67
RESULTS...............,.....................,...,.,....,................,...,.,..,..".......................,.........,.................... 70
REFERENCES,...,...,...,.."..,.....,.........,..,...,................,..."....,...,...,..",....,.,......."..,....,..,................. 72
CHAPTER 6. CASE STUDIES OF THE EFFECT ON COMMUNITIES OF INCREASING
ACCESS TO MAJOR GAMBLING FACILITIES ..................................................................73
THE COMMUNITIES.......".,....,.......................,......,..,..,..., ........................,..,........................,.,..... 74
Types of gaming...,..,....,...,......,..,....."....,.................,....,..,...,.,..,........".................,.....,....,.... 74
Economic outcomes ..,.......,.... '.. ,..., ,..... ..,...,.........., ,.,. ,.....,. ,." '..' .,.......... .........., .......,............ 76
Other social benefits and costs ............................................................................................, 78
Problem gambling............."...,..,...""......,.."......"..""...,...,."......,..,........,...,.............,...,....... 78
Public opinion regarding gambling..............................,..,.......................................,............ 79
CASE STUDY ONE: FLORISSANT ......................................,.........................,..".......................,.., 79
Our respondents......,.,..........,.,...,.................,......,.."...........",..,.....,.......,...............,............,. 80
Gaming in Florissant .............................................................:..................."..,...,.......,........" 80
Community changes ..., '.' ,......... ..., .,., ..........,.. ,.....,... ,....... ...,..................... '., ..... .... ......, '... ,..... 80
Current community issues..,...... ,., ......,..,...,........ ...,..............,. .,...... ........... '....., .... .............,." 82
Public views on gaming .....................................................................................,....,............. 83
CASE STUDY Two: HANSEN,..........................................................,..................................;.......84
Our respondents,....... ..,., ,........... ......... ...........".. .........,....... ,...... .....,..... .....,.,......, ....,.. .....,.... 84
Gaming in Hansen ,..,...."..................................................,..................,...................,.......,....84
Community changes,............,.,...,....."........,...........,...."..,...,.........,....,.,..,..........,..,..".......,.,. 85
Current community issues......, ............' ........ ......" ........., ...,..., ..,.,..., '. ..., ..., ,... ....,.,.... ......., ..... 88
Public views on gaming ',.............................,.......................................,....,........................... 88
REFERENCES,.... ............,.......,... ..........,............ ...,.......... .........., .................,....................,........... 89
.
Gambling Impact and Behavior Study
Page ii
NORC
.
......
UNIVERSIn' OF CHICAGO
LIST OF EXHIBITS
Tables
.
Table 1. DSM-IV Criteria for Pathological Gambling ....................................................16
Table 2. DSM-IV Criteria and Matched NODS Lifetime Questions ..............................18
Table 3. Criteria for Classifying Respondents .................................................................21
Table 4. Patron Interviews ...............................................................................................23
Table 5. Key Characteristics in RDD, Patron, and Merged Adult Surveys .....................24
Table 6. Percentage Gambling Types Based on Lifetime and Past- Year NODS
Scores................................................................................................................. 25
Table 7. Lifetime and Past-Year Prevalence of Gambling Problems Among
Demographic Groups, in Percentages................................................................26
Table 8. Attitudes Toward Gambling in RDD+Patron Survey, by Lifetime and
Past- Year Gambler Type.................................................................................... 29
Table 9. Percentage of Lifetime and Past-Year Gambler Types by Health,
Mental Health, Substance Abuse, and Other Problems .....................................30
Table 10. Estimated Annual Amount Ahead, Behind, or Spent (in Millions of Dollars)
in the Past Year, 1998 (from RDD Data).........................................................32
Table 11. Employment Experiences, by Type of Gambler (Lifetime Only)....................44
Table 12. Annual Financial and Job Losses by Problem and Pathological Gamblers .....45
Table 13. Financial Characteristics and Impacts, by Type of Gambler ...........................46
Table 14. Financial Losses, by Type of Gambler.............................................................46
Table 15. Weighted Occurrence of Criminal Justice Consequences, by
Type of Gambler ......... ... ................... ................. ..... ....................... ...... ....... .....48
Table 16. Criminal Justice Losses, by Type of Gambler ................................................48
Table 17. Marital and Health Status, by Type of Gambler ..............................................49
Table 18. Divorce and Health Costs, by Type of Gambler ..............................................51
Table 19. Selected Economic Costs of Pathological and Problem Gambling: Costs per
Pathological and Problem Gambler .................................................................52
Table 20. Economic Impacts of Major Health Problems .................................................53
Table 21. Summary of Comparisons Between Pathological, Problem, and Low-Risk
Gamblers.......................................................................................................... 5 8
Table 22. Casino Proximity Effects in Model 3............................................................... 71
Table 23. Model 3 Parameter Estimates: Selected Outcome Measures .......................... 72
.
Gambling Impact and Behavior Study
Page iii
NORC
.
......
UNiVERSIn' OF CHICAGO
Figures
Figure 1. Past-Year and Lifetime Gaming, 1975 and 1998................................................ 7
Figure la. Past-Year Gambling by Selected Games, 1975 and 1998.................................. 7
Figure 2. Sex of Past-Year Gamblers, 1975 and 1998....................................................... 7
Figure 3. Lifetime Gaming by Sex, 1975 and 1998...........................................................8
Figure 4. Past-Year Gaming by Sex, 1975 and 1998.........................................................8
Figure 5. Lifetime Gaming by Age Group, 1975 and 1998...............................................9
Figure 6. Past-Year Gaming by Age Group, 1975 and 1998.............................................9
Figure 7. Age Distribution of Past-Year Gamblers Versus Age Distribution
of the Adult Population, 1975 and 1998..........................................................10
Figure 8. Past-Year Bingo by Age Group, 1975 and 1998..............................................11
Figure 9. Past-Year Gambling Participation by Type of Game .......................................62
.
.
Gambling Impact and Behavior Study
Page iv
NOR'e
.
....... -
IINIVERS/n' OF CHICAGO
ACKNOWLEDGMENTS
Principal Sta-Ff
Dean Gerstein, NaRC's senior research vice president, held lead scientific responsib:ility
for the design, integrity, and analysis of the Gambling Impact & Behavior Study (GIBS).
In addition, he wrote parts of this report, including the Highlights and Introduction, and
coauthored Chapters 1, 2, and 4. Dr. Gerstein studied at Reed College and Harvard
University. Sally Murphy, senior study director in NaRC's survey operations center and
graduate of the University of Denver, was the GIBS project director, responsible for
overall resource management and the implementation and procedures of data collection.
Marianna Toce, research analyst, led the community case study task, served as
coordinator for development of the questionnaires, and was editor-in-chief of GlBS
reports; she authored Chapters I and 6. She is a graduate of Mount Holyoke College.
.
John Hoffmann, senior research scientist, served as task leader for statistical analysis of
RDD and patron survey data and completion of the community database. Dr Hoffmann
has studied at James Madison University, Emory University, and the State University of
New York at Albany. Amanda Palmer, survey specialist and graduate of the University
of Chicago, served as GIBS assistant project director, preparing, managing, and assUJring
the quality of day-to-day flows of study materials and information. Robert John:;on,
senior survey methodologist, served as statistical analyst for the community database and
designed and performed the merging of the RDD and patron samples; he authored
Chapter 5. Dr. Johnson is a graduate of Oberlin College and the University of Michigan.
Cindy Larison, research analyst and alumna of Old Dominion University and the
University of Maryland, designed and implemented a large proportion of statistical
programming for the patron, youth, and combined adult datasets. Lucian Chuchro, who
studied at the University of Chicago, was manager of the adult and youth telephone
survey operations, including the pilot, validity, and reliability studies. Tracy Buie is a
graduate of the University of Dallas. She was overall supervisor ofGIBS field operati.ons
for the patron survey, including recruitment, training, and development of procedures.
Mary Ann Hill, senior survey methodologist, studied at Pomona College and UCLA; she
was sampling director for the RDD adult and youth surveys. Laszlo Engelmann, senior
statistician, developed the community database sample and conducted initial harvesting
of variables; he studied at Kalamazoo College and the University of Southern California.
NORC Collaborators
.
Rachel V olberg, president of Gemini Research, Ltd., contributed as senior investigator to
all aspects of study design, implementation and analysis. In addition, Dr. Volberg
coauthored Chapters 2 and 4. She is a graduate of the University of California, San
Diego, and University of California, San Francisco. Henrick Harwood, vice president of
The Lewin Group, served as task leader for economic analysis; he studied at Stetson
University and the University of North Carolina, Chapel Hill. Adam Tucker, research
associate at The Lewin Group, contributed extensively to the statistical programming of
the economic analysis; he is a graduate of Kenyon College. Messrs. Harwood and
Tucker authored Chapter 3. Eugene Christiansen, president of Christiansen/Cummings
Gambling Impact ':Ind Behavior Study: Acknowledgments
Page v
NORC
.
...... -
UN/VERS/n' OF CH/C,IGO
Associates (CCA), graduated from the University of California, Berkeley; throughout the
study he provided advice and information pertaining to community and economic aspects
of gambling. William Cummings, general manager of CCA, contributed to the
community database sample design and developed the estimates of community gambling
expenditures; he studied at the Massachusetts Institute of Technology. Sebas1\ian
Sinclair, an associate of CCA, studied at New York University. He assisted in
developing the es.timates of community gambling expenditures.
Other Contributors
The authors wish to acknowledge the important role played by an additional cadre: of
NORC staff, whose dedication was essential to completion of the work. These staff
members were Ann Anderson, Joyce Ashaye, Albert Bard, Haider Baig, Martin Ban-on,
Maureen Bonner, Bradley Bouten, Angeline Bregianes, Angela Brittingham, Sharon
Brown, Dennis Bryson, Jody Dougherty, Nancy Farinella, Lynn Gallagher, Adrian Oil,
Cheryl Gilbert, Gerry Griffin, Angela Hermann, Rita Jena, Ben King, Cynthia Knight,
Heather Kwasigroch, Kathryn Malloy, Michael McNicholas, Bronwyn Nichols, Michael
O'Connell, Albert Pach, Imelda Perez, Ann Ragin, Kenneth Rasinski, Michael Rhea,
Lenora Rodriguez, Annemarie Rosenlund, Alan Sanderson, Sam Schildhaus, Joanna
Small, Patrick Smillie, Howard Speizer, Janel Temple, Faiz Uddin, Cynthia Veldman,
Karen Veldman, Woodly Westbrook, and Kirk Wolter.
.
We also benefited greatly from the work of a technical advisory committee composed of
Drs. Henry Lesie:ur at the Institute for Problem Gambling, Peter Reuter at the University
of Maryland, and William Thompson at the University of Nevada, Las Vegas. Other
advice and information of substantial help us came from James Breiling of the National
Institute of Mental Health, James Colliver at the National Institute on Drug Abuse, David
Fischer of the Department of the Treasury, Curtis Barrett at the University of Louisville,
Philip Cook at Duke University, Sue Fisher at the University of Plymouth, Diane
O'Rourke at the University of Illinois, I. Nelson Rose at Whittier Law School, Richard
Rosenthal at the: University of California Los Angeles, Howard Shaffer at Harvard
Medical School, and Randy Stinchfield at the University of Minnesota.
We appreciate the opportunity afforded us to work with the Commission staff, especially
its executive director, Timothy Kelly, and research director, Doug Seay. We are greatly
in the debt of Research Subcommittee chair Leo McCarthy, for sustaining us through
several long seasons, and to Research Subcommittee members John Wilhelm and James
Dobson, and their respective assistants, Eric Altman and Ron Reno, for ideas and
critiques that were invariably thought-provoking and useful. We thank the Commis:;ion
chair, Kay Coles James, and all other members of the Commission for their support,
encouragement, and constructive criticism.
.
We must note without specific attribution the cooperation given to our pilot and main
patron surveys by members of the commercial gaming industry, ranging from multi-
billion-dollar corporations to mom-and-pop stores, and the assistance to the validation
study of the GIBS questionnaire given by a number of gambling treatment providers. It
is not a light matter to admit survey researchers onto the premises and among the
customers or clients who provide one's livelihood, and we are indebted to all of the
individual owners and corporate officers who were gracious enough to do so. We were
also beneficiaries of the good offices of the American Gaming Association, National
Gambling Impact and Behavior Study: Acknowledgments
Page vi
NORC
.
......
UNIVERS/n' OF CHICAGO
Indian Gaming Association, American Horse Council, several state lottery and gaming
commissions, and the National Council on Compulsive Gambling.
Finally, our greatest debt is to the respondents to our surveys-at home on the telephone,
patronizing gaming facilities, and in their workplaces responding to the community case
study survey. We would like our efforts to be thought of as trying to capture and convey
the stories respondents have tried to tell, while rendering a clear and accurate sense of
what their experiences mean overall for the purpose of informing rational public debate.
Wherever we fell short of reaching that goal, as faithfully as circumstances of the work
and the state of the science permitted, the responsibility rests entirely with the authors.
.
.
Gambling Impact and Behavior Study: Acknowledgments
Page vii
'.
NORC
""""
UNf\lERStTY OF OnC.1GO
HIGHLIGHTS
Basis of findings
The National Opinion Research Center at the University of Chicago, in collaboration
with Gemini Research, The Lewin Group, and Christiansen/Cummings Associates,
collected or assembled and analyzed five new data sets on gambling behavior, problems,
and attitudes. Three data sets were national surveys (2,417 adults at home via telephone,
530 adults intercepted in gaming facilities, and 534 adolescents (16 and 17 years of age)
at home via telephone), and the other two were a 100-conununity statistical data base and
ten community case studies on the effects of casino openings.
.
Changes in Gambling Participation Over Time
· The last national survey of gambling behavior was published in 1976, conducted in
1975, and covered participants' lifetime and past-year behavior, with "past year"
defmed as calendar year 1974.
· Since the 1975 survey, the ratio of adults who have never gambled has dropped from
roughly one out of three to one out of seven, and gambling expenditures have
increased from 0.30 percent of personal income to 0.74 percent of personal income.
· Patterns of adult gambling have changed substantially since 1975:
-Lotteries and casinos are now the most conunon forms of gambling. The
proportion of adults who played the lottery in the past year has doubled to about
one adult in two, and the proportion who gambled in a casino in the past year has
more than doubled, to 29 percent of adults,
-Past-year bingo and horserace betting have declined by two-thirds and about one-
half, respectively.
-Gambling patterns among women have grown more like gambling patterns among
men.
-Proportionately fewer people aged 18 to 44 years are gambling, and
proportionately more people 45 and older are gambling, with the most dramatic
increase among adults 65 and older; however, it is still the case that the proportion
of seniors who gamble is smaller than the proportion of gamblers in younger age
groups.
Pathological and Problem Gambling
.
· Based on criteria developed by the American Psychiatric Association, we estimate
that about 2Y2 million adults are pathological gamblers, and another 3 million adults
should be considered problem gamblers.
· Extending these criteria more broadly, 15 million adults are at risk for problem
gambling, and about 148 million are low-risk gamblers (about 29 million adults have
never gambled).
Gambling Impact and Behavior Study: Highlights
Page viii
.
.
.
NORC
-
UNiVERSI rr' OF CH1CAGO
.
Although the telephone survey results alone did not detect statistically significant
differences between men and women, the combined patron and telephone results
indicate that men are more likely to be pathological, problem, and at-risk gamblers
than women.
.
Pathological, problem, and at-risk gambling are proportionately higher among
African Americans than other ethnic groups, although African Americans still
comprise a minority of all pathological gamblers.
Pathological gambling is present in one out of five of the 1 percent of adults who
consider themselves professional gamblers.
.
.
Pathological gambling is found proportionately less often among people who are over
6S, college graduates, and in households with incomes over $100,000 a year;
however, college graduates are more likely to be at-risk gamblers than those at other
education levels.
.
The availability of a casino within 50 miles (versus SO to 2S0 miles) is associated
with about double the prevalence of problem and pathological gamblers, according to
the combined patron and telephone survey results. This finding is similar to the
difference in the overall level of past-year casino gambling (40 percent of adults
living close to casinos versus 23 percent of adults living SO to 2S0 miles away);
however, these prevalence rates were not different in the telephone survey alone.
· Pathological and problem gamblers are more likely than other gamblers or
nongamblers to have been on welfare, declared bankruptcy, and to have been arrested
or incarcerated.
· Pathological and problems gamblers are much more likely than low-risk gamblers to
gamble for the excitement, to have been troubled by mental or emotional problems
including manic symptoms and depressive episodes, and to have received mental
health care in the past year.
· Pathological and problem gamblers, who comprise about 2.5 percent of adults,
probably accoWlt for IS percent of casino, lottery, and pari-mutuel receipts from the
gamblers who are represented in the surveys.
· Pathological and problem gamblers in the United States cost society approximately
$S billion per year and an additional $40 billion in lifetime costs for productivity
reductions, social services, and creditor losses. However, these calculations are
inadequate to capture the intrafamilial costs of divorce and family disruption
associated with problem and pathological gambling.
Youth Gambling
· Youths 16 and 17 years old gamble less than adults and differently from adults,
primarily betting on private and unlicensed games-especially betting on card games
and sports and buying instant lottery tickets.
· Youthful gamblers tend to bet much smaller amounts of money than adults.
Gambling Impact and Behavior Study: Highlights
Pogeix
.
NORC
-
UNJVEi?S!rr" OF CHICAGO
.
Adjusting for the smaller amounts of money at stake, the rates of pathological and
problem gambling among 16 and 17 year olds are similar to those for adults, and the
rate of at-risk gambling is about double the adult rate.
Community Impact of Casinos
· In communities proximate to newly opened casinos, per capita rates of bankruptcy,
health indicators, and violent crimes are not significantly changed (changes in
nonviolent and minor crime rates could not be analyzed statistically).
· Unemployment rates, welfare outlays, and unemployment msurance ill such
communities decline by about one-seventh.
· Construction, hospitality, transportation, recreation, and amusement earnings rise, but
bar, restaurant, and general merchandise earnings fall, and race tracks are vulnerable
to casino competition.
.
· Per capita income stays the same, indicating the communities reap more jobs, but not
necessarily better jobs. There appears to be more of a shift in the types and locations
of work than a net improvement in the local standard of living.
· There is wide perception among community leaders that indebtedness tends to
increase as does youth crime, forgery and credit card theft, domestic violence, child
neglect, problem gambling, and alcohol/drug offenses.
.
Gambling Impact and Behavior Study: Highlights
Page x
.
NORC
-
UNNc?SJn' OF' CHICAGO
INTRODUCTION
This report covers the backgroUnd, methods, and [mdings of the research program
initiated on behalf of the National Gambling Impact Study Commission by a study team
from the National Opinion Research Center at the University of Chicago (NaRC) and its
partners at Gemini Research, The Lewin Group, and Christiansen/Cummings Associates.
The NaRC team's program of research began with the execution of a contract with the
Commission on April 23, 1998. In the 9 months following, five distinct data collections
were designed, pilot-tested, and completed:
.
· We conducted a nationally representative telephone survey of 2,417 adults (aged 18
and older) regarding their gambling behavior, attitudes, and related factors.
· Using an abbreviated version of the telephone questionnaire, we performed an
intercept survey of 530 adult patrons of 21 gaming facilities (casinos, racetracks,
lottery ticket outlets, and small service establislunents with electronic gaming
devices) in eight states, as a supplement to the adult telephone survey.
· We carried out a national survey of 534 youths aged 16 and 17, using random
sampling and the telephone questionnaire used in the adult telephone survey.
· We built a longitudinal data base (with data points from 1980 to 1996) of social and
economic indicators and estimated gambling expenditures in a randomized national
sample of 100 communities, stratified to represent places near to and distant from
major gaming facilities, as well as states with and without lotteries.
· To complement the statistical analysis of community effects, we conducted case
studies in 10 widely distributed communities regarding the effects of one or more
large-scale casinos opening in close proximity; we based these studies on telephone
interviews with seven to eight key persons in each community.
.
In the first section of this report, we compare the survey methods and key findings on
gambling participation of the 1998 adult telephone survey with the methods and results of
a 1975 national probability survey of adult gambling behavior and attitudes. The 1975
survey was conducted by researchers at the. University of Michigan on behalf of the
previous national commission concerned with gambling policy issues. The second
section of the report describes our survey questionnaire's diagnostic screening approach,
based on standardized psychiatric criteria for problem and pathological gambling, as weIl
as our findings on the prevalence and correlates of pathological and other types of
gambling among the adult population. The third section of the report estimates the
economic costs engendered for the individual, family, and community by the most
severely affected types of adult gamblers. The fourth section turns to the youth survey,
providing our key [mdings concerning youth participation in types of gambling and the
prevalence of gambling problems in the context of findings on these dimensions among
adults. The fifth section reports the findings of a multilevel statistical analysis of the 100-
community database, estimating the effects of casinos on a variety of local economic and
social indicators. The sixth and final section develops the qualitative counterpart to the
statistical analysis of community effects, summarizing the results of 10 community case
studies and including two of these cases in detail. Separately bound from this volume are
three appendices: Appendix A, which includes discussion of the development of the
Gambling Impact and Behavior Study: Introduction
Poge 1
.
NORC
......
UNI'IEnS!TY OF CHICAGO
questionnaires and contains the instruments used in the RDD, Patron Intercept, and Self-
Administered Surveys; Appendix B, which includes discussion of the sampling and
weighting methodologies for the RDD and Patron Surveys and the Community Database; .
Appendix C, which contains our detailed findings for all 10 of the case. study
communities, as well as the questionnaires used for this segment of our study, and
Appendix D, which contains detailed statistical tables.
.
.
Gambling Impact and Behavior Study: Introduction
Page 2
.
NORC.
-
UNjl,'ERS!nO OF CHlCAGG
CHAPTER 1. COMPARISON BETWEEN THE RESULTS OF THE 1975
AND 1998 NATIONAL SURVEYS OF ADULT GAMBLING BEHAVIOR
In 1976, when the Commission on the Review of the National Policy Toward Gambling
issued its final report, only 13 states.had lotteries, 2 states (Nevada and New York) had
approved off-track wagering, and there were no casinos outside of Nevada. The gaming
industry has grown tenfold since the "Review" Commission sponsored this first national
survey on gambling behavior in the United States in 1975. Today, a person can make a
legal wager of some sort in every state except Utah, Tennessee, and Hawaii; 37 states
have lotteries, 21 states have casinos, 37 have lotteries, and slightly more have off-track
betting. Furthermore, between 1976 and 1997, revenues from legal wagering in the
United States grew by nearly 1,600 percent (Cox, Lesieur, Rosenthal, & Volberg 1997;
Christiansen 1998), and gambling expenditures more than doubled as a percentage of
personal income, from 0.30 percent in 1974, to 0.74 in 1997 (Kallick et al. 1976;
Christiansen 1998).
.
Public opinion and the political landscape have changed tremendously since the Review
Commission's report was released. Not only have lawmakers dramatically eased existing
restrictions around the country, but states are aggressively marketing their own games of
chance, as well as marketing themselves to the casino industry. Such changes have
brought not only the opportunity to gamble, but an awareness of the opportunity to
gamble, into the everyday lives of most consumers around the country. One of the
directives of the current Commission is to determine the extent to which these changes
have modified gambling prevalence and behavior in the United States.
Studies on gambling prevalence among the U.S. general population have generally
reported either a "lifetime prevalence rate" (the percentage of respondents who reported
having ever gambled) or a "past-year prevalence rate" (which refers to the percentage of
respondents who have gambled at least once in the past 12 months). The survey results
collected for the Review Commission by a research group at the University of Michigan
(Kallick, Suits, Dielman, & Hybels) showed that residents of the United States had a
lifetime prevalence rate of 68 percent and a past-year prevalence rate of 61 percent. For
the most part, studies conducted since 1976 have only been conducted in individual states
that commissioned studies, usually as a result of concern about the effect of increased
access to gambling opportunities. These studies have found that lifetime prevalence rates
ranged from 64 to 96 percent and past-year prevalence rates ranged from 49 to 89
percent. In 1997, a meta-analysis was conducted of 120 of the 152 available studies in an
effort to establish more precise overall estimates of gambling prevalence in the United
States and Canada. This overview estimated that the lifetime prevalence rate across the
general population was in the vicinity of 81 percent (Shaffer, Hall, & Vander Bilt 1997).
.
While valuable, these results do not provide the kind of detail and comparisons across
time that are needed to inform national policy. In 1998 the National Gambling Impact
Study Commission contracted with NaRC to collect data from a nationally representative
sample of households on gambling behavior and other factors, in order to extend
knowledge about the prevalence and consequences of national changes beyond piecemeal
state and regional levels to a national level. This section is a brief examination of
methods and most notable comparisons of findings that we have been able to make
between the 1975 and 1998 surveys.
Gambling Impact and Behavior Study
Chapter 1. Comparison Between the Results of the 1975 and 1998 Notional Surveys
Page 3
.
NORC
11IM
UNNEIlS!n' OF CHICAGO
Methods
The University of Michigan's national survey of adult gambling attitudes and behavior
took place in the summer of 1975 and involved a three-stage sample design. First, a set
of primary sampling units (counties, large cities, and boroughs of New York) were
selected at random to represent all of the household dwellings in the country.
Approximately 3,250 households were then selected randomly within these primary
sampling units (including an oversample of households in 12 of the largest U.S. cities).
Each selected household was then approached to determine the number of adults (18 or
older) of each sex residing there and to randomly pick one adult to be interviewed (the
within-household selection procedures was designed to achieve a two-to-one oversample
of males). This initial household contact was the "screening" stage, completed in
approximately 2,680 households, or 82.5 percent of those sampled.
Every effort was then made during the field period of the study to complete an interview
with each of the selected individuals. After completion, the survey was weighted so that
each actual individual respondent was calculated to represent a specific number of
persons of the same sex, household type, and geographic category. These weights were
then further adjusted to match the overall sample to other key national characteristics
such as sex, income, race, education, and occupation, using for these corrections the most
contemporary population counts and estimates made by the V.S Bureau of the Census.
.
The Michigan field team completed 1,749 interviews, for a 65.3-percent response rate;
however, due primarily to large differences in response rates between the oversampled
cities and other areas, the weighted response rate was 75.5 percent. The product of the
successful screening rate among households and the successful interview rate among
selected individuals produced the total cooperation rate-53.6 percent of actual
(unweighted) interviews and 62.2 percent of the population after weighting the sample.
The survey of adults (18 and older) performed by NaRC in 1998 was carried~out by
telephone, instead of in person. A random sample of 10-digit telephone numbers was
purchased from Survey Sampling, Inc., a well-known vendor of telephone samples. The
list from which the numbers were drawn included only actual U.S. area codes and
telephone banks (that is, blocks of 1,000 consecutive numbers within these area codes)
that had been determined to contain a threshold number of active residential numbers.
Then each number in the sample of 9,200 numbers acquired by NaRC was called (in
some cases as many as 50 times) to determine whether it was a working residential
number in contrast to a nonworking number, a commercial/business line, a cell phone,
data or fax line, or a nonprimary household telephone. These calls also served to
determine whether there was an English-speaking or Spanish-speaking adult in the
household able to answer interview questions.
.
NaRC staff classified 4,358 numbers as working residential numbers eligible for
interview. NaRC interviewers successfully screened 3,281 of these households to
establish the number of adults of each sex residing there and to select one household
adult (using systematic randomized sampling rules) for interview. Usable interviews
were subsequently completed with 2,417 adults (44 in Spanish), of whom 14 were
completed as self-administered versions of the questionnaire mailed to the respondents at
their preference. The screening completion rate was 75.3 percent, and the post-screener
completion rate was 73.7 percent, for a final cooperation rate of 55.5 percent.
Gambling Impact and Behavior Study
Chapter 1. Comporison Between the Results of the 1975 and 1998 Notional Surveys
Page 4
.
NORC
......
UNlVERS!T'i OF CHICAGO
The respondents to the telephone suryey were weighted by age group, sex, ethnic/racial
group, number of adults in the household, and state (in a few cases, contiguous smaller
states were treated as a block). The weighted numbers and proportions were
approximately equal to those in the general population, according the March 1998
Current Population Survey, and the weights summed to the overall number of adult
residents of the United States, approximately 200 million (more precisely, 197.35
million) persons. On average, each respondent in the 1998 survey represented about
81,650 adults.
The 1975 survey included a supplementary adult survey of 296 persons in three counties
in the State of Nevada, which was then distinguished sharply from other states due to "the
widespread legal availability of gambling casinos, slot machines, bingo, keno, and betting
parlors." This sample was screened to exclude "those who moved to Nevada in order to
gamble," and was meant to "predict what might happen if gambling facilities were
legalized elsewhere" (quotations are from KaUick et a!., 1976, p. 361). Among the
comparisons made between the Nevada and national samples were differences in
opinions about gambling, participation in gambling, and the prevalence of potential and
probable "compulsive gamblers," based on scaling a series of items adapted from a
variety of psychometric measures of personality.
.
We did not need a special survey of Nevada residents in 1998 in order to "predict" the
results of more widespread casino gambling, lotteries, and other forms of gambling,
which had become so much more widely available in the intervening years. There was an
argument to be made, however, for a supplementary survey that would yield an increased
number of problem and pathological gamblers, using much more contemporary measures
than the scales developed in 1975. The approach taken was a supplementary survey of
patrons of gaming facilities. Data from the supplementary sample are described further
and used in later analyses in this report, but not in this chapter.
The 1975 survey and NORC's 1998 survey for the National Gambling Impact Study
Commission were in many respects similar enough to permit ready comparison between
their results. The questionnaire content of the two surveys was also similar in key details.
Both supplemented the demographic and geographic information obtained in the
screening phase with economic and family demographic indicators. Both surveys asked
highly detailed questions about gambling behavior across the respondent's lifetime and in
the past year (or, in the 1975 survey, the 1974 calendar year). Both surveys queried
adverse consequences related to gambling, gambling-related attitudes, and other types of
behavior such as occupation, criminal record, and physical and mental health. The 1998
survey asked a series of standard questions on substance use and dependence similar to
those on the National Household Survey on Drug Abuse.
.
Finally, the 1998 survey included a series of diagnostic questions for the subset of
respondents who reported ever experiencing gambling losses greater than $100 in one
day or across a year. These questions were designed to match the criteria for diagnosing
pathological gambling according to the definition in the Diagnostic and Statistical
Manual of the American Psychiatric Association, Fourth Edition-:-ar, the "DSM-IV
criteria." This series of questions, referred to in this report as the NaRC DSM-IV
Screen for Gambling Problems, or the NODS (to distinguish it from similar screening
instruments such as the SaGS and MAGS), has no close counterpart in the 1975 survey.
Further analysis of the items used in the 1975 survey to assess "probable compulsive
Gombling Impact and Behavior Study
Chapier 1. Comparison Between the Results of the 1975 and 1998 Notional SUlveys
Page 5
.
NORC
-
UN/VcRs!r)' OFCH1CAGG
gambling" and "possible compulsive gambling" might permit us to use some of these
items as stand-in for some of the diagnostic criteria in DSM-IV and thus permit closer
comparisons of diagnostic categories in the two national surveys. However, this
exploration must be deferred to future research.
Overall Prevalence Rates
The survey results published by the Commission in 1976 showed that in 1975, adult
residents of the United States had a lifetime prevalence rate of 68 percent and a past-year
prevalence rate of 61 percent. As can be seen in Figure 1 on the following page, the
proportion of respondents indicating that they have gambled in the past year has not
changed much since 1975, only reported by 63 percent-still considerably below the 78
percent of past-year gamblers in Nevada in 1975. However, the percentage of
respondents who have gambled at least once in their lifetimes has increased noticeably at
the national level, from 68 percent to 86 percent.
.
The change in rates for lifetime gambling is not surprising, since the sheer number of
facilities one can visit to place a wager has exploded since the 1970s. However, it does
appear surprising that the percentage of Americans who gamble each year remains
unchanged. One possible explanation is that persons who play in casinos or buy lottery
tickets tend to gamble more frequently now than before. The high visibility and
controversy surrounding casinos and lotteries may also have Played a role in this regard.
Increasingly more Americans are flol.:king to play these types of games, while the
popularity of the plethora of other games with less visibility and fewer patrons has
declined dramatically. In Figure la on the following page, we display the change in rates
of past-year play for casinos, lottery, bingo, and horse racing between 1975 and 1998.
The percentage of people who reported playing the lottery in the past year has doubled,
and the percentage reporting gambling in a casino in the past year has more than doubled.
Past-year bingo, on the other hand, has decreased by two-thirds, and we found a similar
decline in past-year pari-mutuel betting on horses.
Demographics of Gamblers
Next NaRC examined the data to determine whether the characteristics of various types
of gamblers have changed since the 1976 report. Data from both 1975 and 1998 show
that the sex ratios of lifetime and past-year gamblers has remained constant and is in
accordance with their distribution in the general population (see Figure 2). Of the games
we examined for this overview, past-year casinos patrons most closely fit this overall
finding, with an almost 50-50 division between males and females in both 1975 and
1998. Past-year lottery players did not differ much from casino patrons, except that
males were slightly more likely to have played than females. On the other hand, past-
year bingo players were more likely to be female in both 1975 and 1998, and we found
this relationship to be even stronger today, with women comprising about two-thirds of
adults who have played bingo in the past year.
.
Gambling Impact and Behavior Study
Chapter 1. Comparison Between the Results of the 1975 and 1998 Notional Surveys
Page 6
.
100
80
cu
Ol 60
0
1:
cu
~ 40
cu
II..
20
0
NORC
-
UNiVEnSfrf OF CHICAGO
Figure 1. Past-Year and Lifetime Gaming
1975 and 1998
1975
1998
DPast-Yeor Bettors .Lifetime Bettors
figure 1 a. Past-Year Gambling by Selected
Games, 1975 and 1998
60
GJ 40
CIl
0
..
C
GJ
u
..
GJ 20
ll.
0
Casino
.
cu
en
Cl
'E
cu
u
..
cu
ll.
.
Lottery
Bingo
Horse Racing
01975 -1998
Figure 2. Sex of Past-Year Gamblers
1975 and 1998
100
90
80
70
60
50
40
30
20
10
o
OFemole
.Mole
1975
1998
Gambling Impact and Behavior Study
Chapter 1. Comparison Between the Results of the 1975 and 1998 Notional Surveys
Page 7
.
NORC
-
UN/VERs/n' OF CHICAGO
Despite the equal proportions of males and females who have gambled in their lifetimes,
the actual percentage of all women who have ever gambled has risen by 22 percent, while
for males, the percentage has increased by 13 percent (see Figure 3, below). Similarly,
the percentage of women who have gambled in the past year has risen slightly, but the
percentage of males who have placed a bet in the past year has stayed the same (see
Figure 4).
100
80
GI
Cl 60 -
c
...
c:
GI
u 40
...
GI
a.
20
0
.
80
GI 60
Cl
0
E 40
GI
u
...
GI
a. 20
0
Figure 3. Lifetime Gaming by Sex
1975 and 1998
1975
1998
DAdu/t Moles .Adult Females
Figure 4. Past-Year Gaming by Sex
1975 and 1998
1975
1998
DAdult Males .Adult Females
We also looked at differences between 1975 and 1998 by age group, which revealed
some noteworthy changes. While the percentage of people who have ever gambled has
increased in each age group (see Figure 5 on the following page), most notably among
the population 65 years and older, today we see a more comparable distribution of
lifetime gamblers across age groups. Another fInding worth noting here is that the
proportion of lifetime gamblers among young adults has increased only about 5
percentage points, while this proportion has increased within other age groups between
14 and 45 percentage points.
.
Gambling Impact and Behavior Study
Chapter 1. Comparison Between the Results of the 1975 and 1998 Notional Surveys
Page 8
.
NORC
""""
UNiVEflS!TY OF C.YfCAGO
Figure 5. Lifetime Gaming by Age Group
1975 and 1998
100
BO
<<l
Cl
a 60 -
'E
<<l
u 40
..
<<l
Go
20
0
1 B-2 4
. 25-44
45-64
65+
Age Group
01975 .1998
.
Looking at past-year gambling by age (see Figure 6 below) also reveals some interesting
differences. The proportion of young aduits placing a bet in the past year has decreased
by about 10 percent, while it has increased slightly in the 45 to 64 age group and
dramatically among persons over 65-about doubling. While it may be tempting to
sound an alarm at what may appear to be a gambling epidemic among seniors, such
changes are simply due to the fact that persons age 65 and older had much lower rates of
participation relative to their proportion in the population in 1974. As we show in Figure
7 (on the following page), seniors are still underrepresented among the total population of
past-year gamblers.
Figure 6. Past-Year Gaming by Age Group
1975 and 1998
80
Gl 60
CI
a
..
~ 40
u
..
Gl
Q. 20
o
18-24
25-44
45-64
65+
Age Group
01975.1998
.
Gombling Impact and Behavior Study
Chapter 1. Comparison Between the Results of the 1975 and 1998 Notional Surveys
Page 9
.
NORC
mm
UNJIlEASHY OF CHICAGO
Figure 7. Age Distribution of Past-Year Gamblers Versus Age
Distribution of the Adult Population, 1975 and 1998
50
40
41
CI
c 30
..
C
41
... 20 -
..
41
Q.
10
0
18-24
25-44
45-64
65+
Age Group
c:::JPast-Year 1975 _Past-Year 1998 -0-% of Adults 1974 ___% af Adults 1998
Finally, we examined demographics of players of specific games in 1975 and 1998.
Probably the most striking difference between then and now is the distribution of
individuals who play bingo. When the Commission published its report in 1976, bingo
was far more popular, and one of the reasons ascribed to this was the social acceptability
ofthe game, due to the established stereotype of the bingo player:
.
Bingo is commonly described as a "little old ladies" game. While this does not
imply that only little old ladies play bingo, it clearly indicates that most people
view bingo players as a conservative group, predominantly female, and
somewhat elderly. In addition, they are often perceived as belonging to a low-
income group with a relatively low educational achievement.
However, the data collected in 1975 contradicted this picture. The difference between
the percentages of men and women who played bingo were not "overwhelming" (16
versus 21, respectively), and the group had a significant over-representation of persons
under 25 as well as a significant under-representation of persons 65 and older. They
found also that bingo players come from all educational backgrounds, but with
disproportionately fewer from both extremes (persons without a high school diploma and
college graduates) (Kallick et al. 1976).
Today the stereotype persists, and while it seems to be fulfilling itself, it would appear
that it remains off-mark. The percentage of women who have played bingo in the past
year is double that for men, and the percentage of players without a high school diploma
is triple that of the number of college graduates who play. In addition, the percentage of
players from every age group has fallen off considerably more than it has for players 65
and older (see Figure 8 below). Nevertheless, in 1998, the age distribution of bingo
players is virtually identical to their proportions within the general population.
.
Gambling Impact and Behavior Study
Chapter 1. Comparison Between the Results of the 1975 ond 1998 Notional Surveys
Page 1 0
.
NORC
11IM
UNiVEl?S!fr' OF CHICAGO
Figure 8. Past-Year Bingo by Age Group
1975 and 1998
30
'.2T:
~ 20
o
..
C
ell
'" ~~ '.
..
:. 10
o
18-24
25-44
45-64
65+
Age Group
01975 .1998
References
.
American Psychiatric Association. 1994. Diagnostic and Statistical Manual of the
American Psychiatric Association, Fourth Edition. Washington, DC: author.
Commission on the Review of the National Policy Towards Gambling. 1976.
Gambling in America. Washington, DC: Government Printing Office.
Cox, S., Lesieur, H., Rosenthal, R., and Volberg, R. 1997. Problem and
Pathological Gambling in America: The National Picture. New York: National
Council on Problem Gambling.
Christiansen, E.M. 1998. "The United States Gross Annual Wager: 1997."
International Gaming & Wagering Business, Supplement, August 1998.
Kallick, M., Suits, D., Dielman, T., and Hybels, J. 1976. Survey of American
Gambling Attitudes and Behaviors. USGPO Stock No. 052-003-00254.
Washington, DC: U.S. Government Printing Office.
Shaffer, H., Hall, M., Vander Bilt, J. 1997. Estimating the Prevalence of Disordered
Gambling Behavior in the United States and Canada: A Meta-analysis.
Cambridge, MA: Harvard Medical School Division on Addictions.
.
Gambling Impact and Behavior Study
Chapter 1. Comporison Between the Results of the 1975 and 1998 Notional Surveys
Page 11
.
NORC
11IM
UNI'./EP,51ri OF CH!CAGO
CHAPTER 2. THE PREVALENCE AND CORRELATES OF GAMBLING
PROBLEMS AMONG ADULTS
Legal gambling is now an accepted part of the social landscape in many countries. When
gambling is legalized, the operation and oversight of these activities become part of the
routine processes of government. Gambling commissions are established; revenues are
distributed; and constituencies of customers, workers and organizations develop.
Governments become dependent on revenues from legal gambling to fund essential
services. Many nongambling occupations and businesses also become dependent on
revenues from legal gambling to continue to operate profitably, including convenience
stores, retail operators, restaurants, hotels, social clubs, and charitable organizations.
Ancillary services-including legal, accounting, architectural, public relations and
advertising, security, and financial organizations-expand their activities to provide for
the needs of gambling operations (Volberg 1998a).
.
A further element in the growing legitimacy of gambling has been the "medicalization"
of gambling problems and the professionalization of gambling treatment (Abt &
McGurrin 1991; Rosecrance 1985), in other words, the acceptance of gambling problems
as suitable subjects for disciplines such as psychiatry, clinical psychology, and
epidemiology. A constituency of well-educated treatment professionals has emerged
whose livelihoods involve providing services to governments and gaming operators.
Organizations that provide services to these helping professions-hospitals, clinics,
government health agencies, universities and colleges, the insurance industry-have
growing interests in the development of legal gambling. These organizations are
investing increasing, though still relatively modest, resources in training and certifying
treatment professionals, in educating students, and in covering treatment for pathological
gambling.
The Social Construction of Psychiatric Tools
The tools used to generate numbers are always a reflection of the work that researchers
and others are doing to identify and describe the phenomena in which they are intereste4
(Gerson 1983). Historically, standardized measures and indices have often emerged in
situations where there is, simultaneously, intense controversy and a perceived need for
public action (Porter 1995). Examples include the emergence of measures of "public
utility" in France in the mid-1800s and the development of cost-benefit analysis in the
United States in the mid-1900s.
.
There have been three "generations" of psychiatric research since the turn of the century.
The third, and latest, generation of studies began around 1980 and coincided, as did the
first two generations, with dramatic changes in psychiatric nomenclature (Dohrenwend
1998). The publication of the third edition of the Diagnostic and Statistical Manual
(DSM-III) (American Psychiatric Association 1980), with its systematic approach to
psychiatric diagnoses, led directly to the development of semi-structured interviews and
rating examinations for use by clinicians. These tools were quickly adopted for
epidemiological research despite the lag in research to establish the validity of these case
identification procedures among general population samples (Dohrenwend 1995).
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 12
.
NORC.
-
UNIVEAS/n' OF CH:CAGO
The assumption underlying all of the existing gambling research is that gambling-related
difficulties are a robust phenomenon and that gambling problems exist in the community
and can be measured. Despite agreement among researchers and treatment professionals at
this fundamental level, there is disagreement about the concepts and measurement of
gambling-related difficulties. The ascription of "conceptual and methodological chaos" to
the field (Shaffer, Hall & Vander Bilt 1997:8) may be an overstatyment of the situation
among its experienced researchers, but the presence of competing concepts and methods is
not uncommon among emerging and even mature scientific fields. Nevertheless,
disputation among experts has led to some degree of public confusion and uncertainty
about the impacts of legal gambling on society.
Measuring Gambling Problems
Following the inclusion of the diagnosis of pathological gambling in the DSM-III for the
first time in 1980, a few researchers from a variety of scientific disciplines, including
psychiatry, psychology, and sociology, began to investigate gambling-related difficulties
using various methods from psychiatric epidemiology. At this time, few tools existed to
measure gambling-related difficulties. The only tool that had been rigorously developed
and tested for its performance was the South Oaks Gambling Screen (SaGS). The
SaGS, closely based on the new diagnostic criteria, was originally developed to screen
for gambling problems in clinical populations (Lesieur & Blume 1987).
.
The SaGS is a 20-item scale that includes weighted items to determine if the client is
hiding evidence of gambling, spending more time or money gambling than intended,
arguing with family members over gambling and borrowing money from a variety of
sources to gamble or to pay gambling debts. In developing the SaGS, specific items as
well as the entire screen were tested for reliability and validity with a variety of groups,
including hospital workers, university stUdents, prison inmates, and inpatients in alcohol
and substance abuse treatment programs (Lesieur & Blume 1987).
Adopting the South Oaks Gambling Screen in population research
Like other tools in clinical research, the SOGS was quickly adopted in clinical settings as
well as in epidemiological research. The_SaGS was first used in a prevalence survey in
New York State (Volberg & Steadman 1988). By 1998, the SaGS had been used in
population-based research in more than 45 jurisdictions in the United States, Canada,
Asia and Europe (Shaffer, Hall & Vander Bilt 1997; Volberg & Dickerson 1996; Volberg
& Moore 1999). This widespread use of the SaGS came at least partly from the great
advantage of comparability within and across jurisdictions that came with use of a
standard tool (Walker & Dickerson 1996). Although there were increasingly well-
focused grounds for concern about the performance of the SaGS in non-clinical
environments, this tool remained the de facto standard in the field until the mid-1990s,
when the new DSM-IV criteria were published (American Psychiatric Association 1994;
Volberg & Banks 1990).
.
Like all tools to detect physical and psychological maladies, screening questions to detect
gambling problems can be expected to generate some errors in classification. However,
misclassification has very different consequences in different settings. Misclassification
can occur when an individual without the malady in question is misdiagnosed as having
the malady. This type of classification error is called a "false positive." Misclassification
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 13
.
NORC
..."..
UNiVERSITY OF CHfCAGO
can also occur when an individual with the malady is misdiagnosed as not having the
malady. This type of classification error is called a "false negative." While most screens
to detect psychiatric disorders work well in clinical settings where the prevalence of the
disorder under investigation is predictably high, the accuracy of many psychiatric screens
declines when they are used among populations where prevalence is much lower, such as
the general population (Dohrenwend 1995).
Validating the South Oaks Gambling Screen
A national study in New Zealand in the early 1990s furnished an opportunity to examine
the performance of the SaGS in the general population (Abbott & V olberg 1992, 1996).
This opportunity arose from the two-phase research design employed in the New Zealand
study, which allowed the researchers toidentifY true patholOgical gamblers using face-to-
face interviews with respondents selected from subgroups of respondents in a much larger
telephone survey. These sub-groups included non-problem gamblers, lifetime problem
gamblers, and lifetime probable pathological gamblers, as classified by the SOGS.
Prevalence rates for the national sample were corrected using the "efficiency approach,"
which involved calculating the rate of true pathological gamblers in each group and
dividing this number by the total number of respondents in the sample. The efficiency
approach resulted in a revised current prevalence estimate in New Zealand that was 0.1
percent higher than the uncorrected current prevalence rate.
.
The revised prevalence estimate in New Zealand rested on the conservative assumption that
there were no false negatives among individuals who do not gamble regularly. While error
rates in the sub-groups have an impact on the overall prevalence rate, the size of the error
rate for each group has a different impact because of the different sizes of these groups in
the population. Even if the number of false negatives among respondents who do not
gamble regularly were extremely small, the relatively large size of these groups contributes
to a noticeably higher overall prevalence rate. For example, if the nongambling group in
New Zealand is assumed to include a very small number of pathological gamblers (1
percent), the prevalence estimate increases by 0.7 percent.
The New Zealand researchers concluded that the lifetime SaGS was very good at detecting
pathological gambling among those who would currently meet diagnosis for this disorder.
However, as expected, the SOGS identified pathological gamblers at the expense of
generating a substantial number of false positives. The current SOGS produced fewer false
positives than the lifetime measure but more false negatives. It thus provided a weaker
screen for identifYing pathological gamblers in the clinical sense. However, the greater
efficiency of the current South Oaks Gambling Screen made it a more useful tool for
detecting rates of change in the prevalence of problem gambling over time (Abbott &
Volberg 1996).
.
The eclipse of the South Oaks Gambling Screen
With the rapid expansion of legalized gambling in the early 1990s, state governments
began to establish services for individuals with gambling problems. In establishing these
services in more than 20 states, policy makers and program planners sought answers to
questions about the number of "pathological gamblers" in the general population who
might seek help for their difficulties. These questions required epidemiological research to
identifY the number (or "cases") of pathological gamblers, to ascertain the demographic
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 14
.
NORC
-
UNiVEASlfr' OF CH,'CAGO
characteristics of these individuals, and to determine the likelihood that they would utilize
treatment services if these became available.
Around this same time, a variety of methodological questions were raised about SOGS-
based research in the general population (Culleton 1989; Dickerson 1993; Lesieur 1994;
Volberg 1994; Walker 1992). Some of these issues, such as respondent denial and rising
refusal rates, were common to all survey research. Other questions were related to the issue
of how best to study gambling-related difficulties. These included reservations about the
reliability and validity of the SaGS, as well as challenges to assumptions about the nature
of gambling problems that were built into the original version of this instrument. .
What led to the growing dissatisfaction with the South Oaks Gambling Screen? One
important change was the rapid expansion of legal gambling itself. This expansion led
many people who had never before gambled to try these activities. As legal gambling
expanded into new markets and as new types of gambling were marketed to new groups,
the individuals seeking help for gambling difficulties became increasingly heterogeneous.
In their efforts to discount what they saw as unreasonably high prevalence rate estimates,
representatives of the gaming industries also played a role in the eclipse of the South
Oaks Gambling Screen.
.
Prevalence surveys in the early 1990s suggested that growing numbers of women and
middle-class individuals were developing gambling problems (Volberg 1992; Volberg &
Silver 1993). Several of the specific items included in the SaGS made little sense to
these new groups or to the treatment professionals working with them. Questions about
borrowing fTom loansharks, for example, or cashing in stocks and bonds to get money to
gamble or pay gambling debts were more relevant to the middle-aged, middle-class men
most likely to seek help for gambling problems in the 1970s and early 1980s than to the
young adults and middle-aged women who began to experience gambling problems in the
1990s. Questions about others criticizing one's gambling and feeling guilty about one's
gambling were more likely to receive a positive response from low-income and minority
respondents than others in the population (Volberg & Steadman 1992). Questions about
borrowing from the "household" to get money to gamble would be interpreted differently
by individuals from ethnic groups where "household" may be defined as the entire
extended family.
The need was also growing for tools appropriate to different settings and purposes,
including program evaluation. In 1985, only three states funded services for problem
gamblers, but by 1996, 21 states funded such services (Cox, Lesieur, Rosenthal &
V olberg 1997). Along with these resources came new demands for accountability and
performance. These demands drew further attention to the deficiencies of the SaGS and
increased dissatisfaction with its performance in general population studies.
.
Emergence of a new standard: The DSM-IV
A standard exists when a multiplicity of workers concerned with a phenomenon accept, at
least tacitly, that there is a best available measure to identify that phenomenon, and then
adopt that measure in their daily work (Becker 1960; Dean 1979; Gerson 1983; Volberg
1983). However, the way we look at problem gambling has changed over the past couple
decades, and likewise, the DSM-IV criteria are very different from the diagnostic criteria
adopted in the DSM-III in 1980 (American Psychiatric Association 1994).
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates or Gambling Problems Among Adults
Page 15
.
NORC
mm
UN/VERSln" OF CHICAGO
The changes made to the psychiatric criteria for pathological gambling incorporated
empirical research that linked pathological gambling to other addictive disorders like
alcohol and drug dependence (American Psychiatric Association 1994). In developing
the DSM-IV criteria, 222 self-identified pathological gamblers and 104 substance
abusers who gambled socially tested the individual items (Lesieur & Rosenthal 1991).
Discriminant analysis was used to identify the items that best differentiated between
pathological and non-pathological gamblers. While the results from this sample indicated
that a cutoff of 4 points was appropriate (Lesieur & Rosenthal 1998), the American
Psychiatric Association established a diagnostic cutoff of 5 points. pathological
gambling is now defined as persistent and recurrent maladaptive gambling behavior as
indicated by five or more criteria (listed in the table below), with the reservation that the
behavior is not better accounted for by manic episodes-a reservation added somewhat as
an afterthought, as it was not part of the underlying research on which the DSM-IV
criteria were based.
.
Table 1. DSM-IV Criteria for Pathological Gambling
I Preoccupation 'Iils preoccupied with gambling (e.g., preoccupied with reliving past
I gambling experiences, handicapping or planning the next venture, or
, I
I !thinking of ways to get money with which to gamble)
ITolerance INeeds to gamble with increasing amounts of money in order to .
i iachieve the desired excitement !
~_....._."-""......"".__..............+_............,,...._.......................-....-............................".........-.-........-"...."".....""......."....-".......-....-.....-......................".'.".-.........-..-......-.--......"..........."...................!
iWithdrawal lis restless or irritable when attempting to cut down or stop gambling i
1"-.........--..................."........".-.....-.......... ...........-...............-..-............-.."........"..-.----..............-....-................-".......................-...-."...........-.........,,,,,,,,,.,,.,.,,,,,,,,,.,,,,,-,,,,,-,,,,,..,,,,,-,,,,,.,,,,,,,,,,,,,,.,,-,,,,''''''''''''.'-''''''i
iEscape [Gambles as a way of escaping from problems or relieving dysphoric i
! Imood (e.g., feelings of helplessness, guilt, anxiety, or depression) !
IChasing iAfter losing money gambling, often returns another day in order to get I
l___._.._.__~~~_n_..Lc.b..asi n..~~~~~ses~L___..__...____._.._...._.....___.......__.._____........___..______1
I Lying ! Lies to family members, therapists, or others to conceal the extent of !
I iinvolvement with gambling I
I Loss of control i Has made repeated unsuccessful efforts to control, cut back, or stop I
! [gambling i
i--...........--.-....--.............-..-..---...... -.-'-""'''-..''..' ".....;.....-....--:-".".....--..--..."....""........--........--.......-............-.-...-"..........-..................-...........-................'.-."''''''''''''''-'''-''''''''''.''.1
illlegal acts IHas committed ~llegal acts (~.g., forgery, ~raud, theft, or I
! iembezzlement) In order to finance gambling i
IRisked significant iHas jeopardized or lost a significant relationship, job, or educational I
[~:~~i~...ble.-...-....-.j~ra~g~e~:l~;;i~;f:\~~C~;~~ O~~~;y~;~~liev;~-~f;;per~t~...fi~~-;,~i.~Tl
i iSltuatlon caused by gambling i
Most researchers and treatment professionals working with gambling problems have
expressed satisfaction with the new DSM-IV criteria. At two recent international
meetings of gambling researchers and treatment professionals, I the consensus was that
the field needed to move fully into the new "DSM-IV era." Internationally, researchers
and treatment professionals have adopted the DSM-IV criteria as the new standard. For
.
I
The first meeting took place in conjunction with the Twelfth National Conference on Problem Gambling in
June 1998 in Las Vegas, hosted by Trimeridian, Inc. Invited participants included researchers and treatment
professionals from Australia, Canada, Great Britain, Spain, and the United States. The second meeting took
place in September 1998 in Malta at the 42nd ICAA Intemational Institute on the Prevention and Treatment of
Dependencies; this meeting included members of the newly-organized ICAA Gambling Section from the
countries of Canada, Denmark, Great Britain, Italy, the Netherlands, Spain, Sweden, and the United States.
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 1 6
.
NORC
11IM
UNIVERS!TY OF CHICAGO
all we have yet to learn about pathological gambling, the DSM-IV criteria are now the
measure against which the performance of other instruments must be demonstrated.
At the end of the 1990s, one finds a rapidly growing community of researchers and
treatment professionals active in the gambling field and a growing number of tools to
measure gambling problems for different purposes. Until 1990, only three screens
existed to identify individuals with gambling problems, including the ISR screen used in
the last national study; the CCSM; and the SaGS (Culleton 1989; Kallick et al. 1975;
Lesieur & Blume 1987). Since 1990, in contrast, nine screens for adults and three
screens for adolescents have been developed, including two based on the SOGS and at
least four based on the DSM-IV criteria.
Despite this proliferation, the psychometric properties of these new tools remain
unexamined. Even more significantly, few of these new screens have been tested for
their differential performance in clinical settings, population research, and program
evaluation. Another concern is how to calibrate the performance of these new screens
with the results of more than a decade of SOGS-based research.
.
Development of the NORC DSM-IV Screen for Gambling Problems
(lithe NODSII)
The guidelines put forth by the National Gambling Impact Study Commission specified
that the DSM-IV criteria be used to identify problem and pathological gamblers in the
general population. This meant that the SOGS could not be used, since this instrument is
based on the outdated DSM-III criteria. In developing the questionnaire for the research
to be conducted for the Commission, the NORC team identified three screens based on
the DSM-IV criteria that had been used in population research. These included the
Fisher DSM-IV Screen (Fisher 1996), the Diagnostic Interview Schedule (DIS;
Cunningham- Williams et al. 1998), and the Diagnostic Interview for Gambling Severity
(DIGS; Winters, Specker & Stinchfield 1997).2
Careful consideration was given to all three of these possible tools for identifying
individuals with gambling-related difficulties. Our initial decision was to use the DIGS
rather than the Fisher DSM-IV Screen or the DIS. This decision was based on the fact
that only the DIGS had been tested for its performance with non-clinical groups
(Stinchfield 1997). This decision was further based on the high internal consistency of
this screen (Winters, Specker & Stinchfield 1997). However, examination of the
individual items that make up the DIGS raised several doubts, especially about the
varying timeframes associated with different items and about the forced splitting of some
of the DSM-IV criteria into two items.
Accordingly, the research team elected to develop a new instrument based on the DSM-
IV criteria. We have called the new instrument the NODS illQRC ;QSM ~creen for
Gambling Problems). The specific items that make up the NODS and the DSM-IV
criteria to which they relate are shown in Table 2 below.
.
A fourth screen based on the DSM-IV criteria, the Massachusetts Gambling Screen (MAGS; Shaffer,
LaBrie, Scanlan & Cummings 1994) has never been used in adult population research.
Gombling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 17
.
NORC
Im!Il
UNJVERSlfr' OF CHICAGO
.
Table 2. DSM-IV Criteria and Matched NODS Lifetime Questions
[P~;~-~~-p~iio~----Tl----F~-~~;iT;~-~~~v~--;:-b;~-n--p;ri~-d;-i~~ti ng-2~~E- or 1~~g~~-.~~Al~-;--'-l
i iYoU spe~t a lot of time thinki~g about your gambling experiences i
I l___._.J~!:..e!Qil.l1.!~g...Q.~Li~tu~~.~ mbJ!.':.1.g._~~~t.~.':~QL!?~.t~L.g_R__.______~
1 12 IHave there ever been periods lasting 2 weeks or longer when i
L_________.___.J___l~~_~~;.~i~hl?o~~~:~~in~~~~_~~:.~ays::_~~~~~_~_~.:~~~_.::_.J
ITolerance 13 Have there ever been periods when you needed to gamble with 1
. ! increasing amounts of money or with larger bets than before in I
~Withd~~-;r.------k-.----I~~~~~~;~:~~~~T;;Pj~~T~~~~&fJ~~~~~~-~!-~t~~Ty.~~~-g~-~bii~g-d
, 15 On one or more of the times when yow tried to stop, cut down, or!
L______________J__.____J~~~.r.Q.Lyg.~C_9.~1!.lJ?l!!.!_gL w~_~_YE_l!..r.~.~Jle~~_or jrriLCI.~~l.______.___.J
jLoss of control IU'6 IHave you ~ver tried but n~t succeeded in stopping, cutting down, !
! ! or controlling your gambling? I
! ! 7 IIf so, has this happened three or more times? i
i----~~1~~~~=~~~;~~~~u;~
L__._____.______..J_____...jHl}j~_'_<:!n.l<if?!YLh~.!pl.~?~.f?~?.L.Q~_c}~P!.f?~~!g.nL___._.._______________.___._____J
iChasing ! 1 0 IHas there ever been a period when, if you lost money gambling I
l-cyi~-g--.---.----------r.f,-.--I~~~cl~~~Y;~~~T~~j~~1~1i~~l-~1~~;-f~~~~~:~;!.1!t-h~-~-~b~~.t----l
! I ,how much you gamble or how much money you lost on j
I L____..JH9_C!:l_~.!i.~g.L___._.____________,,____,,______.___-.---.----------________~_.______.____J
I 112 If so, has this happened three or more times? i
Pllegal acts 13 Have you ever written a bad check or taken money that didn't
I belong to you from family members or anyone eise in order to
I I a for our amblin ?
iRisked significant 14 Has your gambling ever caused serious or repeated problems in
irelationship your relationships with any of your family members or friends?
I OR
i 1 5 ASK ONLY IF R IS IN SCHOOL Has your gambling caused you
I any problems in school, such as missing classes or days of
I school or our rades dro in? OR
16 IHas your gambling ever caused you to lose a job, have trouble
Iwith your job, or miss out on an important job or career
o ortunit?
iBailout i 17 Have you ever needed to ~sk fa~ily members or anyone else to I
I I loan you money or otherwise ball you out of a desperate money i
L_.____._________..____.L___J?_i!.l,l atig0._!h...9-!.._~~~._!~~9..E:!_lY-~(J.!:l.~~ci._9.y.y~!:!t.lJ~~~.!.i~9.L.._..____...___.___j
.
The NODS is composed of 17 lifetime items and 17 corresponding past-year items,
compared to the 20 lifetime items and 20 past-year items that make up the SaGS, and the
20 items (19 items in the field test) that make up the DIGS. Like the revised South Oaks
Gambling Screen (SOGS-R) used in most of the epidemiological research on gambling
since 1991, the past-year item is asked for each lifetime NODS item that receives a
positive response. The maximum score on the NODS is 10, compared to 20 for the
SaGs. Although there are fewer items in the NODS, and the maximum score is lower,
the NODS is designed to be 'more demanding and restrictive in assessing problematic
behaviors than the SaGS or other screens based on the DSM-IV criteria.
Gambling Impact and Behavior Study
Chapter 2: The Prevolence ond Correlates of Gambling Problems Among Adults
Page 1 8
.
NORC
11IM
UN/VERs/n' OF CHICAGO
Several complications needed to be overcome in developing the NODS. For example, a
number of the DSM-IV criteria are difficult to establish with a single question. In
assessing these criteria (preoccupation, escape, and risking a significant relationship), we
used two or three questions, and respondents received a single point if they gave a
positive response to any of the questions assessing that criterion. Another complication
in constructing the NODS is that two of the DSM-IV criteria (withdrawal and loss of
control) assume that the questioner already knows that the individual has tried to "stop,
cut down, or control" her or his gambling. Therefore, we obtained this information first
before asking whether the respondent had felt restless or irritable during these times (i.e.,
withdrawal); we then assessed whether the respondent had succeeded in doing so (or,
experienced loss of control).
Our final decision in developing the NODS was to place definite limits on several of the
criteria, in keeping with the approach taken in alcohol and drug abuse research. For
example, in assessing "preoccupation," the NODS asks if the periods when respondents
spent a lot of time thinking about gambling or about getting money to gamble have lasted
2 weeks or longer. Similarly, the NODS asks whether respondents have tned to control
their gambling three or more times without~success (loss of control). We also ask
respondents if they have lied to others about their gambling three or more times (lying).
Only a positive response to the latter questions contributes to the respondent's score on
the NODS.
.
The greater specificity of the NODS was adopted by the research team in response to
concerns about misclassification. As noted above, research on the performance of the
SaGS has shown that the lifetime screen is very good at detecting pathological gambling
among those who currently experience the disorder. However, the lifetime SaGS
accurately identifies at-risk individuals at the expense of generating higher numbers of
false positives. Although more research is needed, it is likely that the lifetime NODS will
prove more effective than the lifetime SaGS at detecting pathological gambling in a
variety of populations.
In the national survey, NaRC chose to administer the NODS only to those respondents
who acknowledged ever losing $100 or more in a single day of gambling, as well as to
those respondents who denied this, but acknowledged that they had been behind at least
$100 across an entire year of gambling at some point in their lives. We chose to use
these "filter" questions in the national survey after our pretesting indicated that
nongamblers and very infrequent gamblers grew impatient with repeated questions about
gambling-related p.roblems. Moreover, our review of previous surveys indicated that
persons who had never experienced significant losses were not those who reported
problems related to gambling (Volberg 1997a, 1997b). We believe that this approach
captured virtually all of the respondents within the survey's respondent population who
would report three or more problems.) Further analyses will be needed to estimate the
capture percentage for gamblers who would have reported one or two problems.
.
) A bias may exist toward under-registration of problem gambling among lottery and bingo players. Such players
tend to lose smaller amounts on any given day of gambling, which may accumulate to substantial sums; however,
such players may not consider themselves "behind" in the sense that most gamblers would. Filtering based on
expenditures or frequency of play rather than losses is an alternative approach that may yield some differences in
survey-based estimation.
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 19
.
NORC
......
UNIVERSITY OF CHICAGO
Validity and reliability of the NODS
In developing our instrument for identifying individuals with gambling-related
difficulties, we received substantial assistance from the gambling treatment community.
Most significantly, we were able to field-test the NODS and examine its performance in a
clinical sample prior to adopting it in the national survey.
A sample of 40 individuals in outpatient problem gambling treatment programs
throughout the United States responded by telephone to an abbreviated version of the
questionnaire used in the national survey. Thirty-eight of these individuals (95 percent)
scored five or more points on the lifetime NODS. A diagnosis of pathological gambling
requires an individual to meet five or more of the DSM-IV criteria. The other two cases
scored four points on the lifetime NODS. Lesieur and Rosenthal (1991, 1998) have
argued persuasively that meeting 4 of the 10 DSM-IV criteria constitutes an appropriate
threshold for a diagnosis of pathological gambling.
.
Scores for the past-year NODS were somewhat lower than lifetime scores in the clinical
sample. Thirty of the forty individuals (75 percent) scored five or more points on the
past-year NODS; five of these individuals (12 percent) scored three or four points; and
the remaining five (12 percent) scored zero to two points. The significance of lower
scores on the past-year than the lifetime scale may differ depending on the status of the
client. As discussed below, the DSM-IV criteria are meant to accumulate or apply across
as many years as the individual has gambled-five - criteria are not required to. appear
within the confines of a single year in order to establish the diagnosis, firmly identifying
the individual as a pathological gambler. Lower past-year scores may also result when an
individual has been in treatment for an extended period (more than 1 year) or entered
treatment in order to prevent an impending relapse.
The test-retest reliability of the NODS across a period of 2 to 4 weeks was tested in 44
cases, including some of the clinical cases discussed above. Both the lifetime and past-
year scores on the NODS were highly reliable. The lifetime test statistic (r=0.99) and the
past-year test statistic (r=0.98) were well above the 0.80 considered desirable for overall
test-retest agreement. Our conclusion based on the field test was that the- NODS has
strong internal consistency and retest reliability. The lifetime NODS appears to have
strong validity as well in identifying clinically confirmed pathological gamblers. In this
respect, the past-year NODS does not perform quite as well. We report the past-year data
here to permit comparison with results of other surveys which use the 12-month time
frame, but we consider the lifetime NODS scores to be the superior instrument for the
purpose of estimating prevalence rates and investigating correlates. We report the past-
year data here to permit comparison with results of other surveys which use the 12-month
time frame, but we consider the lifetime NODS scores to be the superior instrument for
the purpose of estimating prevalence rates and investigating correlates.
.
The NODS typology
Numerous terms have been adopted or proposed in the field of gambling research to
identify individuals who experience difficulties related to their gambling. The terms
"compulsive" and "addicted" are popular with the public and the media; however, the
psychiatric term "pathological gambler" is more widely used in the gambling treatment
and research communities. The terms "problem," "at risk," "potential pathological,"
"sub-clinical," and "in transition" have all been proposed by gambling researchers or
Gombling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 20
.
NORC
""""
UNlVEf?SIr,"OF CHICAGO
treatment professionals to identify individuals who do not meet the psychiatric criteria for
a gambling disorder but who nevertheless appear to experience substantial difficulties
related to their gambling. One recent term, "disordered" gambling, was proposed as a
way to describe the continuum of problems from less to more severe levels, noting the
similarities and differences among troubled gamblers as observed in a multiplicity of
studies (Shaffer, Hall & Vander BiIt 1997).
In discussing the results of the national survey, we have adopted the following
terminology to correspond to the problem levels determined by the survey questions:
Table 3. Criteria for Classifying Respondents
,---.---.-",-----..---.----..---------.--..-.-.------------.--...---..----------.-..-......---,
-iNongambler INever gambled .
r,-.--.--.----..--.---..-..-.--..---....-.-....-------.-.-...."-..---..-..-----.--.......-".-.--...-...-.-.,,..-."-..--.--..-."..-...""..--...-.-"-".-"--.--.-.-.-.,,-,,.-...-,,,,-......"1
i Low-risk gambler I Gambled, but never lost more than $100 in a single day or year
i lOR !
! i Lost more than $100 in a single day or year but reported no I
!.._____.___.______._.__ ______.l Q..~~=.!~_s:~.i~!j9._____~__ _____._...._"..__._._..._.._._.____...___,,_.____._"__..--.........".--..-.......1
!Lost more than $100 in a single day or year AND reported: i
~~rislga~~~r ____I On~~~~t;:<? Dj_M-IV ..s!l!~r.~97--=::=-~===::=-~===--=-:==:==~-~~~==--J
lf~blem_g.9..~ bier -.----lIb.r:.~Q!-~..l!r. DSM=I~S.!:!~C~___.._______.____._______.___.___._,,__.l
LP.ilth 0 I~~l.g a I"!!..~I~I. f.~~~_ or m...9...r::.~_._Q. SM=!'{..s..r:~~E!..rl9..______.._____.______._____".______.______.._.........J
.
The role of timeframe
The DSM-IV describes pathological gambling in the following terms:
Pathological gambling is persistent and reCUlTent maladaptive gambling
behavior...that disrupts personal, family, or vocational pursuits.... Although a
few individuals are "hooked" with their very first bet, for most the course is
more insidious. There may be years of social gambling followed by an abrupt
onset.... The gambling pattern must be regular or episodic, and the course of
the disorder is typically chronic. (AP A 1994, pp. 615-17)
In the study of clinical disorders, pathological gambling is considered a chronic rather
than an acute disorder. Acute disorders, like influenza, wounds, or broken bones, may be
healed and leave no further mark or susceptibility. Chronic disorders; like pathological
gambling, alcoholism, and manic depression, are quite different. Once fully developed,
chronic disorders strongly tend to recur, constituting a lifelong vulnerability; even in
periods of remission or relative quiescence, the disorder may yield a continuing stream of
disabilities. This vulnerability to relapse may be effectively treated and kept in check.
However, a period in which the individual is relatively free of symptoms does not
indicate that the person is free of the disorder.
From the perspective of measuring prevalence, the strongest emphasis belongs on
determining whether pathological gambling has developed, rather than on whether
symptoms are recent or current. The DSM-IV criteria clearly reflect this, in their focus
on the accumulation of discrete symptoms across a lifetime. These criteria do not require
that specific symptoms be clustered tightly together in time (e.g., during the past year).
.
The field test conducted prior to the national survey demonstrated that the sensitivity of
the lifetime NODS in a clinical population was higher than the past year NODS. One
would expect this if pathological gambling were appropriately conceptualized as a
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 21
.
NORC
"""'"
UNIVEi1Slfr' OF CHfCAGO
chronic disorder. It remains to be seen how well the past-year NODS criteria map onto
clinical assessments of pathological gambling. Based on how the NODS is constructed,
as well as our findings in the general population, we believe that the specificity of the
NODS items is very good, reducing the rate of lifetime false positives; in this respect, we
believe the performance of the NODS exceeds that of the SOGS.
Patron Survey
It was expected (and the results below confirm) that the adult RDD survey would yield a
relatively small number of cases of pathological and problem gamblers. In anticipation
of this limitation, NaRC was charged with conducting a second survey to generate
additional problem and pathological gamblers. An intercept survey of patrons of gaming
facilities was selected as the most promising approach-in other words, to go where
gamblers are, and especially where more frequent gamblers would be found in
concentrated numbers. The research design called for 500 patron interviews to be
collected from 5 major facility types in approximate proportion to their estimated share in
overall gaming revenues. This distribution was targeted as follows: 170 interviews in
lottery ticket outlets (not including locations with video lottery terminals only), 125 in
Nevada and New Jersey casinos, 65 in riverboat casinos, 65 in Indian reservation casinos,
40 in pari-mutuel locations, and 40 in locations with video lottery terminals.
.
NORC first carried out a pilot study, which comprised 86 interviews with randomly
selected patrons at three destination-style casinos in Wisconsin and Nevada. This type of
facility was viewed as the most difficult in which to successfully conduct such a survey.
(These casinos agreed to participate in the pilot study through the offices of the National
Iridian Gaming Association and the American Gaming Association.)
The pilot survey provided experience with the process of recruiting sites to participate in
the survey, as well as with some of the specific methodological features of randomly
intercepting and recruiting patrons in these settings (generally in the stream of foot traffic
exiting through access doors or corridors), including the feasibility of completing an
interview of this length. We shortened the RDD survey instrument by about one-third, to
191 items, to take on average 18 minutes to administer. In addition, the order of
questions was revised somewhat for ease of administration in a paper-and-pencil in-
person format. The field-test interviews were administered by NORC field interviewers.
Finally, in addition to testing the methods of the patron intercept survey, the pilot study
provided confirmation of the expectation that proportionately greater numbers of
pathological and problem gamblers might be obtained though this type of survey.
Although the pilot survey was too small to confirm this point with counts of these two
categories alone, the proportion of at-risk, problem, and pathological gamblers combined
was nearly 36 percent in the pilot survey-a much richer concentration of such gamblers
than was obtained in the pilot RDD survey.
.
The design for the patron intercept survey originally called for rotating the periods of data
collection throughout business hours, picking 32 data collection sites with 16 interviews
per site. The tight schedule for completion of the patron intercept survey, once
Commission authorization was obtained, as well as the need to deploy interview
resources efficiently, necessitated a revised design. Therefore, we selected fewer sites,
took a greater number of cases per site, and standardized the time frames to the busiest
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Con'elates of Gambling Problems Among Adults
Page 22
.
NORC
""'"
UNiVERSITY OF CHICAGO
hours of mid-afternoon and mid-evening. At the conclusion of the patron intercept
survey, we had completed 530 interviews in 21 facilities (see Table 4).
Table 4. Pa1ron In1erviews
,-----.---..-----.-.---......--..----.-.--.,---.....----.--.--......---r-.---.------.-..-----..-,--...------..--.-..--'---.-.-.--...--.-----1
i , I i Response i
I Type of Facility I Targeted I Attempted I Completed j Rate i
i Casinos in NV & NJ 125 I 313 150 48% I
I Riverboat casinos 65 I 119 64 54%
, I
I Tribal casinos 65 98 67 68%
I Lottery (traditional & VL T)! 210 I 313 I 193 j 62% i
rp~_;:;:;-~~~T--.--...-...----.-----..J----4o--.T-------~25.------;---.-.-----_-56.-----r------260/0------;
i TOTAL i 505 ! 1,068 530 r 50%
The sample facilities were in 7 states from all regions of the country: 4 in the Northeast,
with 106 completed interviews; 7 in the North Central region, with 160 interviews; 4 in
the South/Southwest, with 110 interviews; and 6 on the West Coast, with 145 interviews.
.
The patron intercept data were intended as a supplement to the adult telephone survey.
Due to the constraints of sample selection and size, the intention was not to view these
cases in isolation but to analyze them, to the extent possible, together with the telephone
cases, improving the overall precision of our information about frequent players and
problem and pathological gamblers. After carefully studying the composition of the
patron intercept sample, we arrived at a procedure to combine the samples and re-weight
the resulting larger file to accurately reflect the "dual-frame" origin of the respondents
(that is, we viewed all adults as having two opportunities to be represented in the
sample-to be contacted at home via telephone, and to be intercepted while visiting a
gaming facility). We combined the samples by creating a file that included all of the
more frequent past-year lottery or casino players from both surveys (intercept patrons not
interviewed in casino or lottery sites were included here if they met either the lottery or
casino participation criteria in their questionnaire responses). This "players" sample
contained about 1,226 individuals (450 from the patron intercept survey and the
remainder from the telephone survey) representing about 64 million players.
We then sorted the players from both surveys into 23 groups or ;'adjustment cells"
(described in more detail in Appendix B); each cell included respondents who reported
similar frequencies of casino and lottery play and were similar in age. We then took the
population estimated to have the characteristics of each of these cells according to the
telephone data alone and divided that population number by the number of patron AND
telephone cases in the cell. We then assigned this average weight to each of the patron
cases, and finally readjusted all the weights to add up once again to the cell's population.
In other words, we had the intercepted patrons share the sample weights assigned initially
to the telephone cases whom they most resembled in terms of age and past-year gambling
behavior. Finally, we recombined these re-weighted cases with all of the telephone cases
who were not in the "players" file; none of these other caseweights (adding up to 133
million persons) were changed.
.
The following table indicates key characteristics of the weighted RDD file, the original
unweighted patron file, and the combined, reweighted patron+RDD file.
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 23
.
NORC
......
UNNEJ,Slrf OF CHICAGO
.
Table 5. Key Characteristics in ROD, Patron, and Merged Adult Surveys
~~~~=~~~~=Ni~l~~~-f~~~
I Female I 51.9% i 43.2% I 51.5% I
r-w------.--..-.-w..-----~---------.----.---.-_M!--.-.--..------.-----~-i.-.--.-------.--.-.--..-..---....----..------..------.------/
I Male I 48.1 i 56.8 I 48.5 :
[~a~~7E!~j1i~LiX~~~~-.~:..~=~==__=~~:::::=_~~~=_~~~=:==~~.::~~=-==~~~:~:=~:===::_=~~==:=:il
iWhitel 71.5 I 71.1 i 71.4
!---..-...-.---..-.-.......--...-.......-.--.--..-......--.--....-.-.....-..-....-..-.--..--..........-.-..--r.-.-.----.-....------....-.-.-+-..-........-.-..----..-...----...,
I Black ! 11.1 ; 20.0 i 12.2 i
[@l~~~~-~~=~~=:-.-~==-..==-~=~~~~~=_=_I_=~~~===~~l~:--:~=-E=:~===J1:~~=::==E=~1~f==~-=!
! Age I
r-- ---.---.---...---.-......-.....--.----.-.--.-.r-.---.-----------r--.--.--.----~...T-...----.-.---I
I 18-29 I 22.5 I 11.6 I 22.3
~+~:~~~~~E~~f=-~
! 65+ : 16.2 j 20.5 j . 16.-i . -j
l~~uca!~ii=:_==-~=__=_====____:::.::==~_=:==~===__:__;=..=====:=:..:=T--::=:=~..-=.::=:-~===.=l
~ ~~~~h t:;~:;g.b~~-~~~~l.-.--.-.---..-I-.---Jt1--.-.I.----..-~ ~ ~~--..-....+.--.----H~~-..-.----j
~~~~-= ~~ i~!~~~~~lil
! Less than $24,000 I 34.4 ! 28.5 I 32.8 '
--.--.-------.--.-.........-...-......--.-...-.--..-.......-.--.......-......-...-----.........--.-.-.---.,..-.-.----.--.---.-.-.-.-........+--.--. --.--...-.- ---- "" I
i $24,000-49,999 .! 31.2 ! 31.9 i . 31.6.- .-1
r $ ~cI.Qg:Q;99~ 992=~=~===~==L_-=~]=!..~-=~C~~::-. 2 8~7-=:..~~t"=: 2 6]~:=J
l-if~;~~~i;a-~-ent.~rSt~t.us --.--L----.-.2.:.1_._____L___-1Q:.~_____L____._2:~___1
r-.-:.--.--.....-..--..-.--....---~----..---..---..1-.----.-----.--...--[--.-------1
I' ~i~~~;~djSe "~~~-t~d----.---------t_.----~~-~--...-.-t_-----{-;: ~ --.----1----.-4 ~ :~-)
r:-:-..--~e..--..--------.-.----.-...----.-.__;--...--..-.-------t--------.---.-j
I Never married I' 24.7 I 21.7 ! 25.0 i
1--.--.--..-.--.--.--..---.----.....-.--.-.--.--..---..-...--.....-.-...----.--.----...---.---------.-.----.-..--.-1
! Other marital status i 7.4 i 6.3 ! 6.6 .
~i~~J:~~=~~~T2~~~~j
I_~..l:!.~..@_~t Full-tirl2~__~l..~y.ment .---.J.-..--.--.~---J-.-----2LL.____i-.--2~~Q.----_J
f-~~q~~~T!.;J~l~.~0.!-------..-.-..--.j---..}}:.~--..--.t---~.~i-.+--..--. ~6:~ ---'1
'--.---P-.~.--.-------_.--'_._____..________'____.._.___.___....J._'."__'___._-j
l.~~~:~:_::~~=I~--- ~l .?..__~-T___'--- 60.E==-~-~.I==~_ 2-4~4-----~j
!51-250miles i 64.1 i 37.5 I 61.7 I
r-~~~=~~~~te =-=--=.:===l~~~-~q: i =r=- 1 o~~i==~~~::~== ~!:~=--=.~
~_._._:!y__.._.___... -..--.----.---.-------..-r--..---..-.-.---~.-.-----.----.._.1
L~~rofessionaLga '!'bler'~____.L.__.___J..:.tL..___L_____~~Q....__._L..___.___I~__I
i...!!~9Ue!!9.'_ of ~!~y._------'_"'------__--r____________.__--.--__.___.____I
i~_!..~a st Y!.~~.kI.y.-'''O-'!.E:!'Y-________._...-1____.___.J..?..:;3_+-..!!:j--_-.J-".---__!~:..?--.J
l AtJ~ast m~.nthly..9.!.~~~__99!!1bl~g._.L_~ .6_.___-1____ 49 .8 _l~1..:..9__._1
.
The patron group was on the whole somewhat more likely than the RDD sample to be
male, African American, older than 50, less than college educated, divorced, not
employed, not an active parent, and living close to a major casino and in a lottery state.
Some of these characteristics are also likely to be associated to some extent with under-
representation in a telephone sample. More to the point of carrying out the patron survey,
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 24
.
NORC
""'"
UNf'.'ERSlr, OF CHfCAGO
the patrons were three to four times more likely to play the lottery at least once a week,
gamble in other venues at least once a month, and (albeit only a small fraction) consider
themselves to be "professional" gamblers.
Prevalence Rates
Prevalence rates are based on the proportion of respondents who score on increasing
numbers of items that make up the lifetime scale used in the survey. Table 6 presents
information about the proportion of respondents who scored at particular levels on the
lifetime NODS screen in the RDD survey, the patron survey, and the combined sample
(which pools past-year casino and lottery players from the patron survey). The
classifications, as discussed above, are nongamblers, low-risk gamblers (limited
gambling losses or zero DSM-IV problem criteria), at-risk gamblers (affirmed one or two
criteria), problem gamblers (affirmed three or four criteria), or pathological gamblers
(affirmed five or more criteria).
.
As we display in Table 6, about one in seven (or 29 million) adults have never gambled,
and about 148 million adults are low-risk gamblers. At the other end of the spectrum are
pathological gamblers, who comprise aboutO.8 percent of the adult population based on
the RDD sample alone. Problem gamblers comprise another 1.3 percent of the adult
population, based on the RDD sample, and 1.5 percent based on the combined sample.
Our best estimate based on the combined sample is that there are about 2 Y2 million
pathological gamblers, 3 million problem gamblers, and 15 million at-risk gamblers in
the United States.
Table 6. Percentage Gambling Types Based on Lifetime
and Past-Year NODS Scores
r-.-----------.-:---.....---..-------.-.-------.---...-.T--..-----.----..-.--------..r---..----Co-m b i ne-cr--'-"--l
! I ROD Survey I Patron Survey II (Patron+RDD)~
I oliN ~ 01 I N II 01 ! N I
L/O , II/a! 11 10 I .
, ---,.-----.t-----r---------...--.---.-,.-.----;--------.1---1'...--.---.-------1
I L.ife- I Past i L.ife- i Past n L.ife- ,I Past! L.ife- I Past 1 L.ife- Past I L.ife- I Past I
tIme i Year' tIme! Year tIme Year i tIme IYeor ftme Year! ftme Year
ITOTAL 100.0IlDO.012 41712,417 100.0IlDO.O! 5301530100_0100.02,867 2,867!
jNonga;"bler i 14.4! 36.71 3421 898' 0.6: 2.8; 3; 15; 14.4 36.7: 342! 898i
~.ow:. Ri~___-1_Z?&L.2.Q:~!..I.!_~.~JiL:!5Jlr{&:.~.LZ~:-QL~-9Jj...l~-~II--L~~.IJ-?-?-~LI:.l.!-.L~?11J2~.ill
, . k '79' 31 831 5 'i 7 9' 1 31 9 I 7 I I
t~~~~;~:.~~t:=L~t::---~~j.t=.:.=~~t-.--:.:~~~~~~:~~.~t:=t~:I::=:J~t~=~:~jl=t;t:::~~~:=~~.~~lr[:=~-]
IPathological I 0.8! 0.1! 211 3;1 7.9! 5.31 421 28,1 1.21 0.61 57 271
The higher rates of at-risk, problem, and pathological gambling in the patron survey
confirm the expectations of the patron pilot survey, although that survey was carried out
only in casinos.4
.
4
The division of gambling types among patrons at particular types of gambling facilities was a particular
interest of tne Commission. As the table below indicates, the sample sizes, as well as the small munber of
sites for each detailed facility type within the patron database, render hazardous any attempt to generalize
from the patron data alone. On a self-weighted basis, 13.2 percent of the patrons interviewed were problem
or pathological gamblers, and 17.9 percent were at-risk gamblers. The pari-mutuel patrons at the three race-
Gambling Impact and Behavior Study
Chapter 2: The Prevalence ond Correlates of Gambling Problems Among Adults
Page 25
.
NORC
11IM
UNlI/ERSIr,' OF CHICAGO
As in other surveys, prevalence rates in the national survey are different among the various
subgroups of the population. Table 7 shows lifetime prevalence of gambling type by
demographic characteristics in both the RDD and combined surveys.
jDemOgraPhiC
Characteristic
,Gender
I Male
I Female
IRace
l__ Wh it~-.:.______L~~_~_1.j__l_:U9~~t-_.p.6_L52J.i -__~~8_L1:ZI_}:.4_LQ:_~f_..J:Q_LQ:~j
I Black I 8.1 /2.9, 2.3/1.21 1.9 / 0.0~_.._?_:-~.L4~1L12:_L.!.:ZL_~:-:?.Ll2~
LJ::fuJ?_~ni~_____JJ}.J.1~:E.L9-.:.~.Q:~_.9.:U 0.0'}I]_L~:ZLJ?~cz.l921__0.5I..QJJ
[--Other i 9.6/ 1.41 1.1 /0.51 0.6/ 0.3ij 8.8/1.81 1.2/0.51 0.9/0.4:
IA~ I
1- 18-29------lT6"j-.T4-.-3Tl-:9TO--:8l---i~/O:~ 1-13.912.1 / 1 .0L__l:~i._Qli
I 30-39 _ I 6.9 / 1.4.1 1. 0 /_ 0.4 i_J?2LQ]! 6. ~_L?:..lL!2..LQ~L_LQL O. 6J
Table 7. Lifetime and Past-Year Prevalence of Gambling Problems Among
Demographic Groups, in Percentages
ROO Surve . % __.R~O:!"!.'_'!!!.E.,-,~u'Y_~}:'.J%L_J
At-Risk I Problem I Path. At-Risk i Problem I Path. I
I , I I I
I (n=183) I (n=30) '(n=21) (n=267) I (n=56) I (n=67) I
----t'"7':";.-.- -.-- .-------..,,-.---- .--."
Life!Year I Life!Yeor Life!Yeor Life!Yeor! Life!Year ! Life!Year i
I
J
I 9.6 / 3.2,'1 .6 / 0.4 To. 9 / 0.1 i =?=~L~:?L~CQ.L9_:_?L._LiT(J.8 i
I 6.3/ 1.61 1.0/0.41 0.7 /0.2i 6.0/2.01 1.1 /0.61 0.8/0.31
~ .!
.
tracks visited by interviewers differed significantly from the other five types in their distribution of gambling
types; there were no statistically significant differences among the first five facility types. The past-year
NODS distributions were very similar to the lifetime but at lower levels: 10.2 percent of all patrons were
problem or pathological gamblers, and 14.3 percent were at-risk gamblers; across the six facility types from
Nevada/Atlantic City casinos to pari-mutuel, pathological and problem gamblers were 6.7, 14.1, 10.5,6.1,
6.7, and 28.5 percent of patrons, respectively.
On a more generalizable basis, patterns of lifetime prevalence were calculated using the combined
RDD+Patron data file for past-year patrons (more than one visit) of Nevada and Atlantic City casinos,
riverboats, tribal casinos, traditional lottery outlets, storeslbars restaurants with VL Ts or other electronic
devices, and pari-mutuel racetracks. The percentages of pathological and problem gamblers at these sites
were, respectively, 8.5, 9.6, 7.7, 3.8, 5.3, and 15. I percent. These results indicate that, among the most
common gambling venues, multi-visit lonery patrons in general have the lowest prevalence of pathological
and problem gambling; casino patrons have higher prevalence rates, with small differences by type of casino;
and pari-mutuel patrons have the highest prevalence rates.
.
Table a. Percentage of G~mbllng Types Based on NODS Lifetime Score, by
Gambling Venue--Patron Data Only
rTyp~-;;r..----c:.jijyZ~C.-~~i.~i.:.I~i~::~-iiFe~~;..]~=~T~Id~-~'f~~I~i.I~~:::.l
iGambler i % i N I % i N J % I N -l
I TOTAL i 1 cici:'O'%T"' 149 I 100.0% i 64 I 100:0% I 67 I
! Non ambler I 0.7 1 O! 0 0 - 0
,low-Risk 68.4 102 67.2 43 73.1 49
At-Risk 22.1 33 15.6 10 16.4 11
. Problem 3.4 5 6.3 4 6.0 4
!'.9J.b.~.29iS.91_L...2,L._._. __.__..._..._B._L__19"?.._._1___.LL___.~2.____ __.___1._
[TYP;;~-f--"r_.!:~!!~rio~.!I~!:S::(~)"::ti=_ VL Uoc~tlans_~LLP~ri:M~ucl-(~":=l
Gambler I % ~ N %! _ N I % I ~
!TOTAl i 100.0%! 164! 100.0% i 30 i 100.0%! 56 i
!~ongombler _~_____Q.___~____~j.___.__..L_!____Lj__~~_~___~__1
rff~if~-==t=}~':~-~=l:==:~~..;,t.t~:i~j=-~~t~..::::=~lj-.~ti=l==lij
~~;~-tic;;r..--t"--H-"--I'-----+i---1:~-----f-----}-+-"}t.8----r-----~-j
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 26
.
NORC
l!IIl:m
UNI'./ERSIT)O C~ CHICAGO
.
Table 7. Lifetime and Past-Year Prevalence of Gambling Problems Among
Demographic Groups, in Percentages
I i ROO Sl.!rvey (%) !~Q.O+':'at!:on Surve}'_(~L_J
I I At-Risk I Problem! Path. ~ At-Risk ! Problem! Path. I
IDemographic U~~283Lll~~30L.lJn=2~lt~_==267)j__l~=5~LL~=67) J
;Characteristic I Life/Year I Life/Year' Life/Year. Life/Year! LifelYear ! LifelYeor i
i 40-4 9 ._____~:..U.~1ll;5 / _Qdj'__Q. 9 / 0-.:.3.j~.2.L~:~lJ.JJ?2_1'~~LfZQ.]J
I 50-64 1 5.3/ 2.3! 1.7 / 0.0 1.1 /0.01 6.1 /3.61 1.2/0.3 2.2/ 0.9[
U~___L!>..: 9 / 1 .3LQ. 2 / O. IL9.:.!LQ:Q! 6. 1 L.!:ZLQ":z--LQ:9.L::9~~7o:-2'1
IEducation I
LessthanHS 110.7/1.611.4/0.911.2/0.010.0/2.411.7/1.212.1/1.01
I HS gradu~te ---r8~-13~2Tl~Y/(f.3-r--'.O-:9-/0.3: 8:'o-Ti~r2T2TlJ:=I~?=2I=~
L__?.E~e:. col~g~____L_J. 81.. 2. ~L-l..:...'!.L~zL.o. 9.1QJ.I_..z.:CZ.L~L~LL~j 0 .811-...J..:..]...L.9-:..~~
i Colle e raduate ! 6.1/1.4 0.7/0.0! 0.3/0.0: 6.4/2.01 0.8/0.2, 0.5/0.11
Ilncome ,
[-------..-----T--------.;-;:;-:;r------~...------T--.---------r-------.------r.------.-.-j
, Less than $24,000 1 8.0/2.71 1.0/0.31 1.0/ 0.2i 7.3/2.6 1.6/ 0.71__J..2.LQ.:2J
I $ 2 4 ,000-4 9, 999_J_Z1L 2. 71__1:..!jg_~LJl.JU_ 0 ..1_,. __~~2fl-.:.?.Ll.: 8 I.Q:~L_]_:~LLQ:_~J
$50,000-99,999 I 7.6/1.7 1.0/0.4' 0.7/0.1 8.0/2.51 1.3/0.71 0.9/0.21
1 'L I I
L~O,OOQi:......___tJl~~_Lf..:l_LQ:2.L~~_ _Q~E.i...9.:QJ_~!L~.:2L.!:4 / Q~j.LQ2LQ:.?.1
'Marital status' I
I Married ____L._6.0Ll.61 0.910.~_~8 / 0~5.9 /1.91 1.0/ Q..&Ll.:Q1..Q.:1j
Divorced/se orate 8.9/2.0 1.2/0.0 1.0/0.0; 9.9/4.71 1.7/0.9 3.0/1.71
L-~:~:rb~-~~~~---rk:-f7 ~:!-f-~:~~~~: ~:~ ~ ~~6 J~~~ ~..Hlf:i->~~~I.--.6-i-~-b~~
Widowed 8.9/1.91 1.5/0.01 0.:0/0.0: 7.3/1.71 0.5/ O.O! 0.0/ O.OJ
Minor children 1
None 7.9/2.411.3/0.310.4/0.0117.7/3.1
One or more 7.7/2.3: 1.2/0.61 1.4 / 0.21 7.8/2.71
[Employment i
I Full-time -'I8.5/25[...l.:.U 0,51 0.9 mrs:-S- /3.01 1.5/ 0.7! 1.~ / 0.61
I ~:~~i:;loyed I ;:~ ~ ~:;I ~:~ ~ ~:~l b:~ ~ ~:~l ~:; ~ ~:~I ~:~ ~ ~:~I---.~~O+~':.~I
IRe~~~heast ---rlO""T/3.41"""]~4/ 0.0 -Q5/6~B'-S:812.9r--o.8 / 0.01 o.~
~~~~est ~:~ ~ ~:~. ~:~ ~ ~:;: ~:~ ~ ~:~I"-+:l;-~:'61
West 0.5/0.0' 12.1 /4.31 2.3/1.41 1.4 / 0.6[
Lotte state I
No 4.5/ 2.8! 1.4 /0.81 1.4 / 0.1 4.6/2.9 1.4 / O.7I___L~ Lq]J
~.- Yes _________..L?2!.2:llJ.~LQ~.3 i 0.7 / o.ll-..?..:.u1.:~j_.J_~_LQ:Z.LL~.LQ.:~i
IDistance to casino !
r-~~~~--~-~+t}I-.+~~+~~~---~-~~i-rt7.~jj--i~~+~~H-~~;~~i~~~~
L..--1~Q.~__________.L.9..:Ql_~1Ll:.9.1..9..:}J...J..1.L_Q:iL5. 5 / 2~L...!..:..?..! o. 3_L!.d.LQ:i1
Professional ambler j
No 7.8/2.31 1.3/0.3. 0.7 / 0.1' 7.6/2.9, 1.5/0.71 1.1 /0.5!
---------------- ----1' I r-..--i---.----.........-j
. Yes 14.7/5.9' 5.9/11.8 11.8/0.0 19.2/11.4 2.9/7.2,19.8/10.4:
.
Several interesting observations can be made based on this table and associated tests of
statistical significance (every specific difference in our findings that is noted in the
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 27
.
NORC
......
UNIVERS!rr" OF CHICAGO
following discussion has a 5 percent or less likelihood of arising by chance). First, with
regard to sex, we found that prevalence rates of problem and pathological gambling
tended to be higher among men than women in the RDD survey, but not to statistical
significance. However, significantly more at-risk male gamblers were present in this
sample than at-risk female gamblers. When we increased the sample size by merging the
RDD and patron survey data, analysis revealed that rates of at-risk, problem, and
pathological gambling were all higher among men.
When we examined differences by age, we found that persons 65 years and older were
substantially less likely to be at-risk, problem, or pathological gamblers than those in
younger age groups. The prevalence rates of at-risk, problem, and pathological gambling
are also higher among African Americans than whites in the combined survey (the
difference in the RDD survey reaches significance only for pathological gamblers). In --
the combined survey data, the prevalence of at-risk and problem gambling is higher
among the never married than those who are married, and divorced people have higher
prevalence rates of pathological, problem, and at-risk gambling than married people.
Prevalence is higher among respondents living with minor children in the household than
among those without minor children in the household, but this may be due to the different
age and sex profile of people living with minors (more of whom are women and under
65) versus those who are not.
.
Finally, about 1 percent of respondents in the telephone survey identified themselves as
"professional gamblers." Problem and pathological gambling are both present at elevated
rates in this group. The DSM-IV states that professional gamblers, who limit their risk-
taking and gamble in a "disciplined" way may relate somewhat differently to the
screening items than other gamblers. However, the DSM-IV does not suggest that
professionals are immune to gambling problems, any more than professional bartenders
are immune to alcohol problems; nor does the DSM-IV recommend (or have a basis to
recommend) that the screening criteria be modified for this subgroup. The evidence of
our survey suggests that persons who consider themselves professional gamblers do not
necessarily earn their entire living or even a significant part of it by gambling.
Regional Differences and Availability
In summarizing the results of a large number of prevalence studies conducted throughout
the United States, Cox, Lesieur, Rosenthal, and Volberg (1997) noted that prevalence
rates tend to be lowest in the Midwest and higher in the Northeast. This cross-
jurisdictional analysis also showed that prevalence rates are highest in southern states like
Louisiana and Mississippi, where the availability of legal gambling has increased rapidly,
where the population is ethnically diverse, and where socioeconomic levels are relatively
low. In the national survey, prevalence rates of pathological and problem gambling are
lowest in the Northeast and highest in the West.
.
The availability of a state-owned lottery has a statistically significant association with the
prevalence of at-risk gambling (which increases about 80 percent, in parallel with the
generally greater past-year lottery play in these states, 55 percent versus 33 percent).
However, the difference in prevalence of problem and pathological gamblers hovers at
the edge of statistical significance, and is in the opposite direction. The availability of a
casino within 50 miles (versus 50-250 miles) is associated with a higher prevalence
(about double) of problem and pathological gambling in the combined survey results,
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adul1s
Page 28
.
NORC
11IM
UNlVERSlfr' OF CHIC,1GG
parallel with the general difference in levels of past-year casino gambling (40 percent
among adults within 50 miles, vs. 23 percent of adults 50-150 miles). However, we
found little difference in the prevalence of at-risk gambling in the combined survey, and
differences in prevalence were not statistically significant in the RDD survey.
Attitudes Toward Gambling
It is interesting to examine general attitudes toward gambling among problem and
pathological gamblers. One might assume that despite their gambling-related difficulties,
problem and pathological gamblers enjoy gambling and believe that it is generally a good
thing for society. However, Table 8 shows that nearly half of all pathological gamblers,
as identified by lifetime NODS scores, believe that the overall effect of legalized
gambling on society is either bad or very bad; these attitudes are more negative than for
any other group of gamblers. When the smaller groups with positive past-year NODS
scores are examined, gamblers at every problem level are less negative about the impact
of gambling (conversely, the group that did not gamble in the past year is much larger
than just the lifetime nongamblers, and generally more negative about gambling).
.
Respondents in the national survey were also asked about their reasons for gambling.
Table 8 shows that the majority of at-risk, problem, and pathological gamblers gamble
for excitement or challenge, and in this respect are quite different from low-risk
gamblers. A great majority of at-risk, problem, and pathological gamblers also gamble in
order to win money, and in this respect they also differ from low-risk gamblers. Finally,
we found no statistically significant differences among these groups in the extent to
which they gamble with friends or family, except that pathological gamblers exceed
others. All of these results are the same whether the past-year or lifetime NODS is used.
Table 8. Attitudes Toward Gambling in RDD+Patron Survey, by Lifetime and
Past-Year Gambler Type
r--------.--.-.-.-..--..---.---...-------r--.-.----.-....-...-,---.-------.---. -.-.--.-.-.-....---,-----.--.----,
ill Problem I Path. I
I 1 Low Risk 1 At Risk Gamblersl Gamblers I
I r--.-.---.-..-.~-..-.f.-....-.-.-._-..-_..-~ -.--.--.-...---......---. ---.---.--.--....
IAttitude Toward Gamblin ~ Life/Year I Life/Year Life/Year LifelYear I
Overall impact is bad/very bad 32/24%1 21 / 11 % 27 / 18% 49/ 19%1
-' f~~;;:.~sn~~~;~~~~~~~.~;;pf.-.---~i~-~.j-~---.-~~ ~ -~- --~~7-~~-.-\---~.~7~~_.1
~--.-..---.-...---..-.--.-.-.-.-..-..--..--.---.-..----..-..-.---..---.-.'-.------..--..-..-..--T---.-..-.-..--....-.----.-...- --.-..-..--...-....-..-....-......--......-.-...-...--.....-......-,.,
I Us~a IlrJ;J.9.~b~~ith .iriends~_ f0'21J!LL_~~~~---L-20 / ~4 6_~L?_!_L_~...!_L~.~__J
Correlation with Other Disorders
.
Finally, it is useful to compare problem and pathological gamblers to others in the
national survey in terms of physical and psychological disorders and other kinds of
troubles in life. Table 9 shows the percentages of gamblers and nongamblers who have
experienced some of these problems. Lifetime pathological gamblers are twice as likely
as other gamblers (31 percent versus about 15 percent, with nongamblers, an older group,
falling in between) to describe their general health over the past 12 months as fair or
poor. Lifetime pathological and problem gamblers are twice as likely as all other groups
(13 percent versus 6 to 7 percent) to have sought professional help for emotional or
mental health problems in the past year. Lifetime pathological and problem e;amblers are
more likely than at-risk gamblers (42 percent, versus 27 percent) to acknowledge being
Gombling Impact and Behovior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 29
.
NORC
......
UN/vERs/n' OF CHICAGO
somewhat or very troubled by their emotions, nerves or mental health; lifetime at-risk
gamblers are in turn more likely than lifetime low-risk gamblers (16 percent), who are
more likely than persons who have never gambled (11 percent) to affirm this.
.
Table 9. Percentage of Lifetime and Past-Year Gambler Types by Health, Mental
Health, Substance Abuse, and Other Problems
r---.-.---.-.-.-----...--.--.--..-.-.----...---.....-..-.'-.------.--..--..---.-..-........---...-........-:--.-.-.--r...____._._..._..~---m-.-..--...--.-.-..........--............--"'-'1"-"--'-"'-'-'-'.'--'-'''-''--''--''
i I' Non- Low-Risk i At-Risk I Problem Path. I
I i am~lers Gamblers i Gamblers I Gamblers I Gamblers i
I I L.ife- I Past L.ife-! Past I L.ife- I Past I L.ife- i Past I L.ife- i Past I
I ~~:~~ep:;r /fai'~~p-;~t-y~-~--I-~~-~8t.-~el:~Lt;1~61-~~2~; P;~~7-r-~~~~r~~e3~-;~~l!~~1 ! ~~~i i
iM~nt~Tly-t~~-b'i;d'-{~~~~~~-iiyi-'-'l"-'-lo:i! -"'i4':6-["--15'~9'r'-i7:1'r26:5'r"'-28:'5'i..42~'3-1-"'24':2r-"41~-9'11' --66.5;
T"ROO onl I,'! I ',i ! i i
iMental health tx, past year ! 5.1 j 6.91 6.81 6.31 6.41 10.1 i 12.8! 5.4 13.31 12.9:
.----.---.-----......--.-..---.--.-.-..-....-.---.-.-......-...----..-,.--..-----.-.+---.---...1---.--..--1.-----....t.-.--------r.--.--..-.-,.-.-..-.--...--.-----.;--.----~
I Emotionally harmful for:nily I NA! 0.5 0.11 0.31 0.81 6.81 15.8! 10.5 53.11 65.61
!.9_~g.lJ_I1}~_~~~Q~~g~~~I!_r:'.g.____L._._____L____________~...----.---L-.-J---..--t--..------L-.-..---. -.-...--..-..--I--.----...i
IMonic symptoms, ever I NA! 0.7 NA! 1.61 11.3 17.6! 16.81 13.41 32.5! 40.11
l~6~e~~~:_~Pi ~~-~:...:~~~-.----.t_-~.~L~~~ -..--~~L.~~~t--~~~ -...~~~~--~~-~I--~:t.~~.:-l.L~.~~)
IAlcohol/drug dependent, i 1.11 0.9 1.31 1.8i 5.6 13.31 12.4/13.9 9.91 20.01
lever ROD onl ! I i I I!
IOruguse5+doys,postyear! 2.0! 2.4 4.2! 5.1! 9.213.5,16.8116.1 8.1'13.91
i~_r:.t.L?~_~~~~_E':I.~_l~.~~_______L?~,_.__~:_~ ..--.-~.:?L---~..:.~L-?:.? __?JL_1..9._:~I.._Q:9_ ___~_~..:~I._._~.?...:.Q.I
I Bankruptcy, ever ! 3.9! 3.3 5.51 6.41 4.6 10.91 10.31 13.81 19.21 10.71
. c------------------.-----~-------...,.......- .___..L._._____.____..__ ...---.----+-....---...,.-----_j.___.__+___,
!Arrested, ever I 4.0: 7.0 10.0t' 11.91 21.1 25.71 36.3! 25.0! 32.31 26.41
;-.-----------.-.-.---.----.-,...--.--..-.--.'.----..-r-..-----.--------.-.........--.--l-..--..-..-...-r-....--...-.'..----.-i-...---,--r-----~
ilncarcerated, ever (ROD only)! 0.41 -I 3.7 -I 7.8: -I 10.41 -I 21.4: -I
The survey questiolll1aire includes screens for manic or depressive episodes, that is,
questions asking whether a respondent ever displayed certain symptoms that are strongly
indicative of manic or depressive episodes. The depression items led, if answered
positively, to a full series of DSM-IV diagnostic questions, but this series was only used
with respondents scoring one or more points on the NODS; other national surveys
estimate general prevalence with the same questions. The manic screening items were
asked only of respondents who scored one or more points on the NODS. This helps shed
further light on the "manic episode" exclusion for the psychiatric disorder of pathological
gambling, but we do not attempt to implement this exclusion. It has not been
implemented in previous studies using the DSM-IV criteria, nor is there an underlying
research base to indicate how it would be implemented, and therefore no research to
validate an exclusion procedure (Lesieur and Rosenthal, 1998).
.
Table 9 shows that lifetime and past year pathological gamblers are significantly more
likely than other risk respondents to have symptoms associated with manic disorder. The
lifetime prevalence of major depressive episode among problem and pathological
gamblers is significantly higher than that observed in the general population in other
surveys, and a definite downward trend is noted from those with the most to least severe
gambling problems in our sample. Table 9 also shows that both lifetime and past-year
respondents reporting at-risk, problem, and pathological gambling are more likely than
low-risk or nongamblers to have ever been alcohol or drug-dependent and to have used
illicit drugs in the past 12 months. Lifetime, as well as to have ever been arrested or
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 30
.
NORC
11IM
UNIVERSITY OF CHICAGO
incarcerated. Finally, pathological and problem gamblers are more likely than any other
group to have lost ajob in the past year and to have ever declared bankruptcy.
Gambling Expenditures
One expectation of our survey work was that it would enable us to estimate the
proportion of gaming revenues associated with problem and pathological gamblers.
There are two principal obstacles to this enterprise. First, a certain fraction of gaming
revenues, particularly in destination-style casinos but also in certain high-stakes lotteries,
have historically been derived from a relatively small number of high-end players, many
of whom are not U.S. residents. Therefore, estimates based on a survey that does not
sample from this special stratum must restrict its scope of generalization to exclude
reference to these very wealthy players. Because these players are so few in number,
determination of the population prevalence and correlates of problem and pathological
gambling are not affected by their absence from the sUl'vey. However, due to the amount
of money that these individuals put into play at casinos (and to an evidently much lesser
extent in other games), any denomination of gambling in monetary units will be missing
this component.
.
The second problem is the weakness in individuals' reports of gambling winnings and
losses. Virtually none of the survey data on the reported amounts "ahead" or "behind"
(won or lost) appears to be accurate at face value, when compared with official statistical
data on regulated games. An exception is lottery play, for which we were able to
reconcile the survey data quite well with officially counted sales receipts. This exception
is probably due to the more routinized purchase patterns of most lottery play, compared
with the way that betting and payoffs take place in other games with faster, more
complicated, and more interactive formats. But here as well, the net win/loss data vary
appreciably from what ticket buyers are known to spend and not recover through winning
tickets. Instead of a careful, computer-like accounting for gaming dollars, individuals
tend to understate their net losses and exaggerate their net wins, particularly when
accounting for expenditures in private settings.
Table 10, which is based on the RDD questionnaire data (which covered gambling wins
and losses much more extensively than the patron questionnaire; but see footnote 4
below), displays gambling win, loss, and expenditure totals for five types of gambling, in
total and by type of gambler. The table provides numerous instances of non-credible
overall results-for example, the results of adding up reports of lottery ticket purchases,
on the one hand (expenditure data), and on the other, how much the survey respondents
thought they were ahead (won) or behind (lost) over the course of a year. The annual
information is calculated separately from two kinds of questions-items about the last
day the respondent gambled, which are summed up taking account of the reported
number of days gambling each year, and a direct question about past-year gambling wins
or losses.
.
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 31
.
NORC
-
UNNERSJf'r'OF CHICAGO
.
Table 10. Estimated Annual Amount Ahead, Behind, or Spent (in Millions of
Dollars) in the Past Year, 1998 (from RDD Data)
r-------------r..------.--------.-.-Caslno-.----.--...---.----r---..-------..---.-.Tr~-Cl;-------..--_.------...-l
! Fk~~~J~y~~~~-l-Ah-!~~!~j:-~i~~~~f~~~~~y~~~~~I.~~-!~~!~!Yj~~;~l
;---..--------.--.i----.--.--..-.[.--.-----.---,--.----..r--.--.-..-.-:-:j--.-.-----..~---..-...-.-..-. .-------t-----..-l
iTotal i $35,5551 $30,460, $9,461[ $6, 13~ $9,589 $3,855 $2,903: $1,003,
, , 100.0 i 100.1 ! 100.0 100.0: 1.00.01 100.0 100.0! 100.0'
!Gambler !)'pe reportin amount !
!Low-Risk I 28,059 12,751 6,99 3,2651 8,4311 3,033 2,39~ 6721
~i~R;,k--~.-~~!I~-1\~5l~~o~~-f}---~3rJ7~fi-l~1~-~~
~~ble~----.--j.--.---V;-t~~----~~~~i)f-.....--l~~6~1-.j~-4:~Oil--.--..JT~~j.....--...J..l..?-6~l..-...-m--(lj-9~r---(2+f~1
fpatho,OgiZ~r+---(~7Q~.-.-.--\~~.~t--EE.31--..---[j-5~~----jfi;lr-.-ll:~~-..--.-J~~)j.--..j 1 23lJ.
I I (0.8)1 (3.3): (0.3): (12.3)! (1.2)/ I (0.1)1 ~
, Private I Lottery I Unlicensed J
, " I
i i Last-Do Sumsl Past Year Past Year Past Year
I IAheadlBehindlAheadlBehind Ahead Behind Total SIAheadBehindi
t~_~:~===_.J_~.t~;-[~~}lQ~:~1..~~l~~::.Q~l_J~l~1_.~l~~1)1~~t6;~~~r~~=~~~~~f.;~;~~~~[..Ll~~~j
i~_~.~E!~...!l.~._..r:~p~_r.:!.!~_9..._~'.!!~_~..!.1_!.._._..::J.--------.._____________._________..__.._....____....-_______..--.---___._1
ILow-Risk 1 18,99d 1,5641 2,1651 241' 1,3241 3,95 18,670, 1,855! 2211
i I 79.61 45.8! 52.91 71.7 93.21 64.9 73.9! 76.71 49.7 i
!At-Risk I 4,5281 117j 26511. 8 81il 1,64 4,560" 2761 1731
! ! 19.0; 6.5! 6.5 25.3 5.7 27.0 18.0j 11.41 (39.0))
iProblem I 3421 1,7321 1 ,65~ 11 151 411 1,7421 2881 3~
f------.---T"--JL4l--(~Q..:~l..-.--~.:..?1-l<L~J---(-LJ..)I.-...-(~ZJ _.____....19:?Jj__..JlL~);-----j~~)!
1~:~.~o'o.=~~:'_l_._._.._____t...._______j..._m._JQ;..}~l_._._J~AJl..._...___.._..j____j_l:4.t.__.._....(f:~l..._._..__..j__..__J~:l~1
The total spending estimate for lottery tickets using the RDD data only is 25.5 billion.
This figure is approximately 20 percent below the national lottery sales figures for 1998
sales cited by Clotfelter, Cook, Edell, and Moore (1999) in their analysis of lottery
gambling for the Commission.5 However, to be consistent, the same players in the
survey, when asked to estimate their net receipts, should have reported losses of
.
5
The combined RDD+patron data on lottery expenditures, in contrast to other data in Table 10, are more
complete than the RDD alone. Using the combined data set, we calculated the total lottery expenditures in
the past year to be $31,5 billion, which is very similar to the figures cited by Clotfelter and colleagues for
1998 national lottery sales. However, Clotfelter and colleagues arrived at their own survey-based estimate
for total national expenditures on lottery tickets (based on the same combined data set) by using a more
complex summing algorithm to quantify the frequency-of-play response categories in the NORC
questionnaire, as well as a series of post hoc adjustment factors designed to match the survey estimates for
each major lottery type with the FYI998 sales figures published by LaFleur (www.lafleurs.com). When we
used the same program code (Malme, private communication) to calculate the distribution of expenditures by
type of gambler, we calculated the percentage of the $3\.9 billion total expenditure by low-risk, at-risk,
problem, and pathological gamblers at 67.5 percent, 18.0 percent, 7.9 percent, and 6.6 percent, respectively-
statistics which are very close to our estimates based on the combined data (67.0, 18.8, 7.6, and 6.6) , which
are similar to those in Table 10 except that the RDD data ascribe a much lower percentage of lottery
expenditures to pathological gamblers.. Further analytic explorations of these data, as well as further
methodological research on how to elicit the most accurate expenditure information, will undoubtedly prove
useful to students oflottery play.
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 32
.
NORC
"""'"
UNIVERs!r( OF CHICAGO
approximately $14 billion, reflecting the percentage of lottery expenditures not returned
to ticket-holders. Instead, the data equate to a net loss of $4.7 billion, reflecting $1.4
billion in winnings (claimed by about 8 percent of all the past-year players) and $6.1
billion in losses (contributed by about 85 percent of players); the remaining 7 percent
"broke even." This loss is about one-third what it should have been based on the survey-
expenditures captured in the RDD data, and about one-fourth the actual amount lost
according to the official sales data.
Undercounting of losses and/or overcounting of winnings are also evident for other forms
of gambling, both for wins and losses across the past year as well as on the last day
respondents gambled. The balance of past-year casino wins and losses for last-day and
past-year items shows patrons ending up with a $5 billion or $3 billion windfall, instead
of leaving more than $20 billion at tables and machines-the rev-enues reported by the
casino industry. The same reversals hold for tracks and for unlicensed betting, comprised
largely of sports books.
.
Most revealing of the rosiness of the collective view of gambling results is private
gaming, largely at cards, in which there is no "house" or commercial intermediary to
remove money from players' wins and losses. In private bets, all of the wins and losses
should balance. However, the last-day-based and past-year aggregates from the survey
show the amounts won exceeding the amounts lost by factors of seven and twelve,
respectively. Unlike lottery play, in which 85 percent of buyers consider themselves net
losers and 8 percent net winners (a 1: 10 ratio), in private games, on the last day of play
there were 3 self-reported winners for each loser (3:1), and over a year's time, five
overall winners for every three losers (5:3). While these ratios are not inherently
impossible, since they might imply that each losers' money was spread out across a larger
number of (smaller) winners, the individual amounts reported as won and lost by each
group actually greatly exaggerate, rather than reduce, this disparity in numbers of winners
and losers.
Despite the lack of realism in the overall estimates of monetary wins and losses, there is
some degree of information in the extent to which problem and pathological gamblers
account for the amounts in both the win and loss columns (and in the case oflotteries, the
expenditure column). Discounting the impossible sums of conjured winnings in private
games, most of the money actually and reportedly changing hands is in lotteries, casinos,
and pari-mutuel betting, and these are the estimates for which the number of respondents
reporting win/loss data are the largest.
In lottery play, problem and pathological gamblers account for 8 percent of total
expenditures (but 14.2 percent in the combined survey data for this measure), 8.1 percent
of past-year losses, and 1.1 percent of past-year winnings. In casino play, problem and
pathological gamblers account for 22.1 percent of past-year losses, 14.3. percent oflast-
day-based losses, 16.8 percent of past-year winnings, and 6.4 percent of last-day-based
gains. In pari-mutuel betting, problem and pathological gamblers account for 15.5
percent of past-year losses, 3.3 percent of past-year winnings, none of the last-day-based
losses, and 1.2 percent of last-day-based winnings..
.
Overall, when we sum up these data, similar information on the less frequently played
games, and parallel monetary items such as the amount of money taken to gamble with or
how much the person was willing to lose, these sums converge on the estimate that about
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 33
.
NORC
......
UN/vERs/n' OF CHICAGO
15 percent of the dollars lost gambling are lost by problem and pathological gamblers.
These figures well exceed the percentage of problem and pathological gamblers in the
general population, but not by so much as to dominate the economics of gambling.
Perhaps a more general finding from these data is that gamblers, whether or not they are
classifiable as problem or pathological, seem accustomed to a fairly high level of wishful
thinking about the economics of the games they play.
.
Assessing Problem and Pathological Gambling in the Future
The issues surrounding legal gambling have become far more complex than they were
when the last Commission published its report in 1976. Policy makers, government
agencies, gamb!ing -regulators, and gaming operators are concerned about the likely impacts
of changing mixes of legal gambling on the gambling behavior of broad segments of the
population, as well as on the prevalence of gambling-related difficulties. Public health
researchers and social scientists are concerned with minimizing the risks of legal gambling
to particular subgroups in the population. Economists, financial institutions, and law
enforcement professionals are concerned about the relationship between legal gambling and
bankruptcies, gambling and crime, and the reliance of the gaming industries on problem
gamblers for revenues. Treatment professionals, government agencies, and not-for-profit
organizations are concerned about how to allocate scarce resources for the prevention and
treatment of gambling problems (Volberg 1998b). Finally, groups opposed to the
expansion of legal gambling are now~ working to prevent the further expansion of legal
gambling and to repeal existing activitfes.
Like much of science, measurement is a developmental process. Instrumentation is
always a reflection of the work that researchers are doing to identify and describe the
phenomena in which they are interested. As research on problem gambling continues,
our systems change for classifying problem gamblers. The SaGS represents a culturally
and historically situated consensus about the nature of gambling problems. As research
continues and as the definitions of problem gambling change, new instruments and new
methods for estimating prevalence in the general population and for testing models of
gambling behavior will continue to emerge. These emerging methods must be tested
against each other and against the SOGS in order to advance the field of problem
gambling research in an orderly manner, ensuring the relevance of past work as well as
work in the future.
There are several areas for which we would recommend future research investments
Much more work needs to done with the patron intercept methodology as a way to
capture frequent players. This type of survey work requires cooperation between
researchers and the gaming industry, which will undoubtedly increase as the importance
and value of onsite research findings becomes more widely understood. Research is also
needed on the efficacy of treatment for gambling problems, both through voluntary
support groups and professional channels. Finally, longitudinal studies are needed that
can extrapolate across 20-year spans or longer on the "careers" or "natural history" of
gambling and related disorders.
.
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 34
.
NORC
11IM
UNNERSfrYOFCHICAGO
References
.
Abbott, M.W. & R.A. Volberg. 1992. Frequent Gamblers and Problem Gamblers in
New Zealand: Report on Phase Two of the National Survey. Research Series
No. 14. Wellington: New Zealand Department of Internal Affairs.
Abbott, M.W. & R.A. Volberg. 1996. "The New Zealand National Survey of Problem
and Pathological Gambling." Journal of Gambling Studies, 12 (2): 143-160.
Abt, V. & M.C. McGurrin. 1991. "The Politics of Problem Gambling: Issues in the
Professionalization of Addiction Counseling." In Gambling and Public Policy:
International Perspectives, W.R. Eadington & l.A. Cornelius (eds). Reno:
University of Nevada Press. Pp. 657-670.
American Psychiatlic Association. 1980. Diagnostic and Statistical Manual of
Mental Disorders, Third Edition. Washington, DC: American Psychiatric
Association.
American Psychiatric Association. 1994. Diagnostic and Statistical Manual of
Mental Disorders, Fourth Edition. Washington, DC: American Psychiatric
Association.
Becker, H.S. 1960. ''Notes on the Concept of Commitment," American Journal of
Sociology 66: 32-40.
Cox, S., H. Lesieur, R. Rosenthal & R. Volberg. 1997. Problem and Pathological
Gambling in America: The National Picture. Report prepared by the Research
and Public Policy Committees of the National Council on Problem Gambling.
Christiansen, E.M. 1998. "The United States Gross Annual Wager: 1997."
International Gaming & Wagering Business, Supplement, August 1998.
Clotfelter, C.T., Cook, P.l.,. Edell, lA, & Moore, M. 1999. "State Lotteries at the
Turn of the Century: Report to the National Gambling Impact Study
Commission." Durham, N.C.: Duke University.
Culleton, R.P. 1989. "The Prevalence Rates of Pathological Gambling: A Look at
Methods," Journal of Gambling Behavior 5: 22-41.
Cunningham-Williams, R.M., L.B. CottIer, W.M. Compton & E.L. Spitznagel. 1998.
"Taking Chances: Problem Gamblers and Mental Health Disorders: Results from
the St. Louis Catchment Area (ECA) Study," American Journal of Public Health
88 (7): l093-1096.
Dean, 1. 1979. "Controversy Over Classification: A Case Study From the History
of Botany." In Natural Order: Historical Studies of Scientific Culture, B.
Barnes & S. Shapin (eds). Beverly Hills, CA: Sage Publications. Dickerson,
M.G. 1993. "A Preliminary Exploration of a Two-Stage Methodology in the
Assessment of the Extent and Degree of Gambling-Related Problems in the
Australian Population." In Gambling Behavior and Problem Gambling. W.R.
Eadington & JA. Cornelius (eds). Reno: University of Nevada Press.
Dohrenwend, B.P. 1995. "The Problem of Validity in Field Studies of Psychological
Disorders, Revisited." In Ming, T. Tsuang, Mauricio Tohen & Gwendolyn E.P.
Zahner (eds), Textbook in Psychiatric Epidemiology. New York: Wiley-Liss.
Pp. 3-20.
Dohrenwend, B.P. 1998. "A Psychosocial Perspective on the Past and Future of
Psychiatric Epidemiology," American Jour of Epidemiology 147 (3): 222-231.
.
Gambling Impact and Behavior Study
Chapier 2: The Prevalence and Con'elates of Gambling Problems Among Adults
Page 35
.
NORC
...".
UNNERS!n' OF CHICAGO
.
Fisher, S.E. 1996. Gambling and Problem Gambling Among Casino Patrons.
Report to the British Casino Industry Consortium.
Gerson, E.M. 1983. "Scientific Work and Social Worlds," Knowledge 4: 357-379.
Kallick, M., Suits, D. Dielman, T. & l Hybels. 1976. Survey of American Gambling
Attitudes and Behavior. Research Report Series, Survey Research Center,
Institute for Social Research. Ann Arbor: University of Michigan Press.
Lesieur, H.R. 1994. "Epidemiological Surveys of Pathological Gambling: Critique
and Suggestions for Modification," Journal of Gambling Studies 10 (4): 385-
398.
Lesieur, H.R. & S.B. Blume. 1987. "The South Oaks Gambling Screen (SaGS): A
New Instrument for the Identification of Pathological Gamblers," American
Journal of Psychiatry 144: 1184-1188.
Lesieur, H.R. & S.B. Blume. 1991. "When Lady Luck Loses: Women and
Compulsive Gambling." In Feminist Perspectives on Addictions, Nan van den
Bergh (ed.). New York: Springer. Pp. 181-197.
Lesieur, H.R. & R.I. Rosenthal. 1991. "Pathological Gambling: A Review of the
Literature (prepared for the American Psychiatric Association Task Force on
DSM-IV Committee on Disorders of Impulse Control Not Elsewhere
Classified)," Journal of Gambling Studies 7: 5-40.
Lesieur, H.R. & R.I. Rosenthal. 1998. "Analysis of Pathological Gambling." In
TA Widiger, A.l Francis, H.A. Pincus, R. Ross, M.B. First, W. Davis & M.
Kline (eds.). DSM-IV Sourcebook. Volume 4. Washington, D.C.: American
Psychiatric Association, pp. 393-401.
Polzin, P.E., l Baldridge, D. Doyle, IT. Sylvester, R.A. Volberg & W.L. Moore.
1998. "From Convenience Stores to Casinos: Gambling-Montana Style."
Montana Business Quarterly 36 (4): 2-14.
Porter, T.M. 1995. Tn/st in Numbers: The Pursuit of Objectivity in Science and
Public Life. Princeton, NJ: Princeton University Press.
Rosecrance, J. 1985. "Compulsive Gambling and the Medicalization of Deviance,"
Social Problems 32: 275-284.
Shaffer, H.J., M.N. Hall & J. Vander Bill. 1997. Estimating the Prevalence of
Disordered Gambling Behavior in the United States and Canada: A Meta-
analysis. Boston, MA: Harvard Medical School Division on Addictions.
Stinchfield, R. 1997. Reliability, Validity and Classification Accuracy of the South
Oaks Gambling Screen (SOGS). Paper presented at the 10th International
Conference on Gambling and Risk Taking. Montreal, Canada.
Volberg, R.A. 1983. Constraints and Commitments in the Development of American
Botany, 1880-1920. Ph.D. Dissertation. University of California, San Francisco.
V olberg, R.A. 1992. Gambling Involvement and Problem Gambling in Montana.
Report to the Montana Department of Corrections and Human Services.
V olberg, R.A. 1994. "Assessing Problem and Pathological Gambling in the General
Population: A Methodological Review." In Gambling in Canada: The Bottom
Line, Colin S. Campbell (ed.). Vancouver: Simon Fraser University Press. Pp.
137-146.
.
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 36
.
NORC
""'"
UNlVER$ln" D.!: CH:C~GO
.
Volberg, R.A. 1997a. Gambling and Problem Gambling in Colorado. Report to the
Colorado Department of Revenue.
Volberg, R.A. 1997b. Gambling and Problem Gambling in Oregon. Report to the
Oregon Gambling Addiction Treatment Foundation.
Volberg, R.A. 1998a. The Evolution of Gambling in North America. Keynote
Address, Alberta Lotteries and Gaming Summit. Medicine Hat, Alberta.
Volberg, R.A. 1998b. Testimony: Research on Problem Gambling. Testimony
before the National Gambling Impact Study Commission. Expert Panel on
Pathological Gambling. (January 22, 1998).
Volberg, R.A. & S.M. Banks. 1990. "A Review of Two Measures of Pathological
Gambling in the United States," Journal of Gambling Behavior 6(2): 153-__
163.Volberg, R.A. & M.G. Dickerson (eds.). 1996. Journal of Gambling
Studies, Special Issue on International Prevalence StUdies and Related Treatment
Developments. Vol. 12, No.2.
Volberg, R.A. & W.L. Moore. 1999. Gambling and Problem Gambling in Sweden.
Report to the International Gambling Research Team of Sweden.
Volberg, R.A. & E. Silver. 1993. Gambling and Problem Gambling in North
Dakota. Report to the North Dakota Department of Human Services, Division of
Mental Health.
Volberg, R.A. & H.J. Steadman. 1988. "Refining Prevalence Estimates of
Pathological Gambling," American Journal of Psychiatry 145: 502-505.
V olberg, R.A. & H.J. Steadman. 1992. "Accurately Depicting Pathological
Gamblers: Policy and Treatment Implications." Journal of Gambling Studies 8
(4): 401-412.
Volberg, R.A. & R.M. Stuefen. 1991. Gambling and Problem Gambling in South
Dakota. Report to the Governor's Office of South Dakota.Walker, M.B. 1992.
The Psychology of Gambling. Oxford: Pergamon Press.
Walker, M.B. & M.G. Dickerson. 1996. "The Prevalence of Problem and
Pathological Gambling: A Critical Analysis," Journal of Gambling Studies 12
(2): 233-249.
Winters, K.C., S. Specker & R. Stinchfield. 1997. Brief Manual for Use of the
Diagnostic Interview for Gambling Severity. Minneapolis, MN: University of
Minnesota Medical School.
.
Gambling Impact and Behavior Study
Chapter 2: The Prevalence and Correlates of Gambling Problems Among Adults
Page 37
.
NORC
......
UN/\"ERs!r( OF CHICAGO
CHAPTER 3. ECONOMIC ANALYSIS OF THE CONSEQUENCES OF
GAMBLlN~ PROBLEMS AMONG ADULTS
Problem and pathological gamblers, as defined in previous sections of this report,
experience excessive rates of adverse consequences that have tangible economic costs.
Further consequences experienced by th~se gamblers that are quite real (e.g., broken
relationships and families), although not readily amenable to having price tags attached,
are often termed "intangible" costs. Another dimension of gambling consequences is that
their impact is usually spread across an entire community. While costs begin with the
gambler, they spill over to the household, other family members, friends, employers,
creditors, and the community as a whole. -
.
The focus of this analysis is on the tangible economic value of gamblers' problems or
consequences that have been identified and analyzed in the literature on problem and
pathological gambling (see, e.g., Lesieur 1998; V olberg et al. 1998). The earliest studies
examined these phenomena through indepth interviews and surveys of persons who
sought help to control their gambling. This approach has been effective in identifying the
breadth and types of consequences that pathological gamblers experience and in
exploring alternative ways that economic values can be attached to some of these
problems. The analysis in this study in large measure builds upon the prior studies, but
refines the earlier methods in order to generate estimates of t,he impacts of problem
gambling among problem and pathological gamblers who can be identified from surveys
of the general population.
The estimates in this study diverge from reported impacts and costs among the very small
proportion of problem and pathological gamblers who have sought help (believed to be
only about 3 percent; V olberg 1998). Since it is often the severity and accumulation of
problems well beyond the threshold of clinical concern that may drive gamblers to seek
treatment (Lesieur 1998), we expect that the average costs and impacts from the general
population will be significantly lower than the estimates from treatment populations.
.
A challenge is posed for this study based on the fact that "denial" is considered a clinical
characteristic of addictive disorders, including gambling. This has led to some concern
among researchers (see, e.g., Chapter 6, by W. Thompson. in WEF A Group 1998) that
surveys of the general population will not elicit acknowledgement or valid responses
from pathological and problem gamblers that are selected into the samples. While this
study has asked for specific attribution of adverse consequences/outcomes by respondents
to gambling problems, the primary emphasis has been upon consequences that are
experienced by the entire population, which are also susceptible to being affected by
gambling problems. It is believed that asking about general problems without requiring
adverse outcomes to be attributed to gambling (although allowing it) should elicit
plausible responses. While validation surveys would be useful in the future, the findings
in the rest of this section demonstrate that pathological and problem gamblers are indeed
willing to acknowledge adverse outcomes, and at rates in excess of low-risk gamblers and
nongamblers; indeed, our study found that the very low proportions that experience
adverse consequences tend to attribute them to problem and pathological gambling.
Examples of such consequences include job and fmancial problems, divorce, poor health,
and criminal justice involvement.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 38
.
NORC
......
UNiVERSITY OF CHIC.'\GO
In this analysis, our basic strategy is to compare rates (and costs) of specific adverse
consequences associated with problem and pathological gambling for each of our
designated gambling types. For example, problem and pathological gamblers (and
perhaps those considered at risk as well) are believed to experience higher rates of
personal bankruptcy (primarily attributed to their problems with gambling) than persons
who are otherwise similar but do not gamble or at lower risk gamblers. Obviously, there
are reasons unrelated to gambling for individuals to experience bankruptcy.
The analysis thus attempts to ascertain whether the bankruptcy rates (and other negative
consequences) of problem and pathological gamblers are greater than bankruptcy rates of
other gambling types who are otherwise similar, and to determine whether the difference
is larger than might be expected due to chance. The bankruptcy cost attributed to
problem and pathological gambling adjusts for "expected" rates of bankruptcy. Thus, the
estimates are of "excessive" costs (be it for bankruptcy, job loss, health problems, etc.)
experienced by problem and pathological gamblers.
In attempting to assess the postulated impacts, the survey incorporated questions that
explicitly examined behaviors and problems that prior research on problem and
pathological gambling has suggested are disproportionately experienced by this
population. A large number of questions ask whether respondents attributed specific
aspects of such problems directly to gambling.
.
Costs that could be measured on an annualized, present-value basis (poor physical and
mental health, job losses/unemployment) sum to about $1,200 and $700 for each
pathological and problem gambler, respectively. Other costs are infrequent (e.g., divorce,
bankruptcy, arrest, incarceration), and in the absence of a very large study sample, they
are more readily observed and measured on a lifetime basis (e.g., "Have you ever been
divorced?" vs. "Have you gotten divorced in the past year?"). We estimate these
"lifetime" costs (which are additive with the "annual" costs when the latter have been
translated to a lifetime basis) at about $10,500 and $5,100 per pathological and problem
gambler, respectively. About 25 percent of these estimates are costs generally termed
"transfers." Under standard economic theory, transfers are not treated as costs because
they represent a loss to the "donors" (generally taxpayers) and a gain to the recipients (in
this case, problem and pathological gamblers). Thus, the total cost (including transfers)
might be thought of as the cost to those who are not problem or pathological gamblers.
In this analysis we estimate costs per person by gambler type, with a particular focus on
problem and pathological. We combined the cases from the supplemental survey of
patrons with cases from the adult telephone survey. As explained in Chapter 2, we re-
weighted these groups in order to make the weighted samples generally equivalent to the
age and gender distribution of the general population of 197 million adults age 18 and
older in 1998. This strategy is designed to maximize information about the problem and
pathological gamblers who were relatively rare in the telephone survey and much more
numerous in the patron survey. It is possible to convert these to aggregate or total
national costs-that is, to sum our economic quantities to represent the entire population
by combining the estimates of the number of problem and pathological gamblers with our
estimates of the costs per problem and pathological gambler (per year and across
lifetime).
.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 39
.
NORC
""""
unIVERSITY OF CHICAGO
Other dimensions of the analysis concern the time period over which gamblers have
experienced symptoms of pathological and problem gambling and the component of the
data from which estimates have been derived. This concerns whether the individual is
classified as a pathological or problem gambler for the past year (number of problems
reported in the past 12 months), or on a lifetime basis. The second issue concerns
estimates for the random digit dial (RDD), patron survey, or combined data set.
We believe that the best estimates for the purpose of understanding the economic impacts
of pathological and problem gambling come from use of the combined survey and
lifetime measures. The estimates in this section of the report are based ;largely on these
calculations. However, we have also examined the potential impact of using the different
survey components and the past year versus lifetime in a series of tabulations that appear
in the appendices.
.
When the data are reassesed by the noncombined survey components and the past-year
classification, the general patterns reported in this chapter are supported. However,
relatively few observations are available for many of the values calculated, and small
numbers generate unstable estimates. For example, the impact of past year pathological
gambling sometimes is and sometimes is not estimated to be greater than the impact of
lifetime-but-not-past-year gambling. We observe this for problem gambling as well.
These instabilities do not invalidate the estimates but simply reflect random variation
around the main effect in relatively small samples. Furthermore, the study has not
attempted to identify when patterns of problems were initiated or stopped, or their
duration. Pathological and problem gambling are often long-term, with the adverse
impact(s) building up over time. The adverse consequences may take some time to abate,
and may never completely do so. A person with a problematic work history (or criminal
justice record) carries such a record forward the rest of their life.
Tabulations appearing in the appendix also compare the rates from the RDD and patron
survey. The patron survey has a higher concentration of pathological and problem
gamblers (as a share of all persons interviewed); this is the principal reason this
supplemental survey was undertaken. In both components of the survey we find that
pathological gamblers generally have comparable, if not higher or more severe problems
than problem gamblers. These two groups certainly have more severe impacts than other
types of gamblers (and non-gamblers). Pathological gamblers in the two respective
components of the study (RDD and patron) are generally more like each other in terms of
problems and impacts than they are like "problem," "at-risk" or "no problem" gamblers.
The same statement is true for the "problem" gamblers in each study component-they
are generally more like each other than like other types of gamblers.
.
We believe that the combined sample provides the best technical results for the purpose
of the analyses done in this chapter, and we have used them for the main analysis and for
the preparatory analyses with a few exceptions. In compressing the length of the RDD
interview so as to better suit the patron-intercept mode of data collection, some questions
in the RDD survey pertinent to the analysis in this chapter were omitted from the patron
questionnaire. In these instances we have used the RDD instead of the combined data in
order to develop estimates of problem prevalence from which costs were calculated. All
such instances are identified in the applicable table. Appendix C includes disaggregated
estimates of all of the fundamental descriptive values used in developing the cost
estimates.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 40
.
NORC
""""
UNIi/cnS!rr' OF CH!C,1GO
Prior Studies on the Costs of Gambling
There have been several prior efforts at describing the economic impacts of problem and
pathological gambling (e.g., Lesieur and Anderson 1995, Thompson, Gazel, and Rickman
1996; WEFA Group 1997; Westphal, Rush, and Stevens 1998; also see reviews by
Lesieur 1998; V olberg et al. 1998). The critical contribution of these studies has been the
identification of consequences and impacts of problem and pathological gambling that
have economic implications, and the efforts made to develop estimates of these costs.
Among the obvious financial consequences these studies have examined are gambling-
attributed bankruptcy, dissipation of assets, debt, and theft. Other impacts studied are
missed work or lateness to work, lost employment, stress and impaired physical and
mental health, suicidal ideation, and alcohol- and drug-related disorders. Families and
personal relationships usually are adversely affected, with associated conflict and strife,
with divorce frequently the result.
For the most part, the existing body of research examines persons in treatment for a
gambling disorder. This approach has had obvious advantages in developing and
understanding the phenomenon. Persons in treatment have generally initiated treatment
because they experienced severe consequences from their gambling. Patients enrolled in
treatment based on the self-help tradition are generally encouraged to be forthright about
the problems they have caused or encountered related to their disorder, whether it is for
gambling, alcohol, drugs, or some other problem. It is possible to compile a picture of
the problems of pathological gamblers that go for treatment by reviewing these studies.
.
Still, the objective of this study is to attempt to portray the consequences and economic
costs of typical or average problem and pathological gamblers. Data on gamblers in
treatment probably describe the most severely impacted individuals-the tail of the
distribution in terms of severity and number of impacts. We expect that the general
population survey will identify individuals who have not reached this extreme level of
severity.
Our survey instrument asked about most of the impacts that the existing literature
indicates are likely to be experienced (or imposed) by problem and pathological
gamblers. The following sections will attempt to give some indication as to how
comparable the measured impacts are to prior estimates derived from individuals in
treatment or Gamblers Anonymous (GA).
.
Costly Consequences of Gambling
The NODS survey undertook to examine a number of different types of impacts of
problem and pathological gambling. These included family impacts, job impacts,
financial problems, and criminal/legal problems. While there are many facets and
dimensions to such problems, there are certain issues that are more conducive to both
measurement and valuation. Therefore, this analysis focuses on a small number of
tangible consequences. The consequences of concern were selected both because a body
of literature already exists that strongly suggests that problem and pathological gambling
may'cause such outcomes (e.g., Lesieur and Anderson 1995, Thompson, Gazel and
Rickman 1996, V olberg 1998), and because it is possible to measure certain economic
values that are associated with them. Such consequences include the following:
Gambling Impact and Behavior Study
Chapter 3. Economic Anclysis of Consequences of Gombling Problems Among Adults
Page 41
.
NORC
......
UNI'lEnSITYOF CHlCAGO
.
Divorce;
.
Poor health and mental health problems;
Job loss and lost wages from unemployment;
Bankruptcy; and
.
.
.
Arrest and incarceration.
Based on the existing research literature, it is expected that gamblers with higher counts
of gambling symptoms will have higher rates of problems. Since the problems often
attributed to problem and pathological gambling are also experienced by many people
whether or .not they gamble, we adjusts for whether a problem or pathological gambler
has other characteristics or behaviors that might contribute to the consequence in
question. For example, if those who gamble also have alcohol and drug problems,
ignoring these other problems might result in attributing an inaccurately high
consequence rate to problem and pathological gambling.
.
Our analysis used logistical regression to control for the following sociodemographic
factors: age, gender, ethnicity, educational attainment, residence with one's children, and
use/abuse of alcohol and illicit drugs. In general, these factors were generally strongly
predictive of whether individuals had experienced the costly consequences identified
above. Ignoring these control factors would result in attributing a larger proportion of the
consequences to gambling than if the controls were applied. The specifications of the
variables used and the primary results are presented in the annexes to this chapter.
As we state above, it is important to note that many of the costs often associated with
problem and pathological gambling are not unique to persons who gamble or who might
need help for gambling problems. Thus, our analysis examines the following questions:
. To what extent did the problem and pathological gamblers surveyed experience a
certain consequence?
. To what extent did they attribute the consequence to their gambling?
. . What plausible -economic costs can be associated with higher than expected rates of
this consequence?
Based on these questions, we concluded that the major findings are as follows:
. Problem and pathological gamblers have significantly higher rates of costly
consequences than otherwise similar persons do.
. Problem and path.ological gamblers expenence or Impose thousands of dollars of
economic costs per year on society.
. Problem and pathological gamblers rarely directly attributed these costly problems to
their gambling behaviors or difficulties.
.
The next section presents our fmdings about the extent to which consequences with
tangible economic costs are associated with different types of gamblers. These patterns
are analyzed in order to determine whether problem and pathological gamblers have other
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gombling Problems Among Adults
Poge 42
.
NORC
-
UNIVERs,n' OF OftCAGO
characteristics or behaviors that may be the cause of their higher rates of consequences,
such as gender or age. The section thereafter presents our findings on selected economic
impacts experienced or imposed by problem and pathological gamblers, adjusting for the
effects of other factors.
Employment-related impacts
Adverse financial consequences are the crux of the issue for problem and pathological
gambling. While there are obviously other manifestations and consequences that can and
often do arise, the financial problems are generally thought to underlie these in some
way. One potential mechanism through which gambling might bring adverse
consequences is for the gambler to lose too much money relative to her or his earning
capacity and/or wealth. Problem and pathological gamblers in this study display a pattern
of higher rates of certain types of financial problems relative to other gamblers (with no
or few problems) and to nongamblers. While this finding is almost tautological
(attributing financial problems to gambling contributes to a determination of gambling
type), this is exactly the pattern of problems that contributes to other sorts of
consequences (e.g., family, legal, and health problems).
Another mechanism for adverse consequences is for one to engage in gambling at times
and places that are inappropriate given one's responsibilities; adverse outcomes could
include a decline in job performance and additional costs to employers, job loss, lost
wages, and reliance on Unemployment Insurance and/or other social welfare programs.
.
Studies of pathological gamblers in treatment have looked at a variety of the potential
impacts on the workplace, but they have been limited by not having comparison
populations. Such studies have examined narrow aspects such as lateness or missing
work in order to gamble as well as gambling while on the job, while broader impacts
have included job loss and unemployment. While it is possible to develop cost estimates
from such data, they may present an inaccurate picture, since workers in general are
sometimes late and miss work, or use work time for personal purposes.
Lesieur (1998) found in his review of the cost literature that between 69 and 76 percent of
pathological gamblers have missed work at some point in order to gamble. Various
studies in his review found that from 21 to 36 percent of gamblers in treatment have
attributed a lost job to their gambling problems. A survey in Wisconsin of 98 GA
respondents found that 66 percent had missed work due to gambling, and 21 percent had
lost or quit their jobs due to gambling (Thompson et al. 1996). A general population
telephone survey found that problem gamblers miss slightly more work (2.7 days and 1.7
days for "level 2 and 3" gamblers, respectively) than low-risk and nongamblers (0.9
days) (Westphal et al. 1998).
.
Our model attempts to reflect these issues by building on the standard model of labor
markets. This model decomposes the employment experience into labor market
participation/employment, amount of employment, and wage rate. In the standard model
of the labor market, the wage rate represents the perceived/actual value of the employees'
productivity to their employers. Divergence of an employee's perceived/actual
productivity from their expected rate will result in an increase or decrease in their wage
or salary and/or termination (in the case of under-performance). In an economic model
of the l.abor market, we hypothesize that to the extent problem and pathological gamblers'
behaviors in the workplace impact their overall or average performance, employers will
Gombling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 43
.
NORC
11IM
UNNfflSlri OF CHICAGO
generally recognize and reward the worker with continued employment and increased
wages, or penalize the worker with lower wages and/or termination of employment.
The data reveal somewhat complex patterns regarding employment. For example,
pathological gamblers had relatively high employment (76.3 percent) at the time ~f the
survey. However, among those that had worked in the past year, we found a slightly
higher (but not statistically significant) rate of working less than a full year (about 26.6
percent, versus 18.6 percent for low-risk gamblers). Still, pathological gamblers who had
worked in the prior 12 months were significantly more likely to have 10sUbeen fired from
a job (13.8 percent versus 4 percent for low-risk gamblers). However, they .were not
significantly more likely to have been earning a wage below $10 per hour than others.
The mean household income for pathological gamblers was about 15 percent lower than
for low-risk gamblers, but this difference was not statistically significant.
.
Table 11. Employment Experiences, by Type of Gambler (lifetime Only)
I ! Gamblin ~ e .
I I Non- Low I I Problem I Path. I
!-I:ree of Char_~cte~ist~L9.~m.?ler_ _._ Risk _L~!_~sk_L G.!!!!!...l?.le!:......L~~_"-!..l?le.!_]1
Lful,pJ.oyed curre~.t1Y.._______L~~} **:.. ___Z~d..._--L_ll~__._J___~_~_:_~.::J __...1 6 :~_
I Any employment past ! 64.4 78.8 I 80.3 I 77.2 82.3 I
! year ! i' :
i Amon Those Workin Past Year...
! An unem 10 ment 21.5 12.7 17.7 23.8 15.9
. Months unem 10 ed J 1.6 I 0.9 ; 1.3 1.8 1.3 I
!__~~~!_._<:lj~!?L!i re_~J~~~LY.~<:lLL..____.~_:9..____L____~.:9___.__l._____~~__._--L.__.___]_g~.~.P_._____.___.._l.:3._:~*.._._j
L!:!ou ~_IJ.'._~9.9.~J~.QD 9!1Jy.L__L._E.i. 6Q_.J_.__~Jl!J.Q___1._~L~J..Q__.__J____.!..!"~_:Q..Q.___ .__.._._~_L?..:Jg_._._i
Statistical significance of differences between groups tested using multivariate logistical regression, with
control variables for age, gender, ethnicity, education, child in household, and alcohol and drug use/abuse.
Gamblers with no problems were used as the base group.
Significance tests: Problem and pathological types tested separately; statistically significant at the: ***
=0.01 level; ** = 0.05 level. * = 0.10 level. Problem and pathological types were combined for significance
testing; statistically significant at the: Ij>Ij>Ij> =0.0 I level; Ij>Ij> = 0.05 level.lj> = 0.10 level.
Problem gamblers, in contrast, were significantly more likely to have been unemployed
or at least not working at the time of their interview (58.9 percent, versus 73.3 percent for
low-risk gamblers). However, those who did work were employed for as much of the
year as low-risk gamblers. Their rate of having lost or been fired from a job was also
higher (10.8 percent compared to 2.6 percent for nongamblers). Wage rates did not
appear to be impaired in this group.
.
Employers' losses
The most unambiguous measure of employer dissatisfaction with employee performance
(productivity) is to fire an employee. As noted above, both problem and pathological
gamblers have higher rates of job loss than low-risk or nongamblers-l0.8 and 13.8
percent, respectively (compared to the expected rates of 5.8 and 5.5 percent). Employers
incur search and training costs assumed equal to 10 percent of the annual salary for each
employee replaced. Frazis et al. (1998) estimated that 4 percent of an employee's hours
go into training; we are assuming employer costs equivalent to an additional 6 percent of
an employee's time is invested in recruiting and initially training a replacement hire.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 44
.
NORC
11IM
UNJl.,'cRSln' OF CHICAGO
Since pathological gamblers in our sample earned about $18 per hour, or $40,000 per
year, firing an employee costs an employer an average of $4,000. Since pathological
gamblers had ajob loss rate of 13.8 percent, versus the expected rate of 5.8 percent, their
"excess" rate of job loss was 8 percent. Therefore, the average pathological gambler cost
his or her employer 8 percent of $4,000, or about $320. The cost of excess job loss for
each problem gambler was $200.
Table 12. Annual Financial and Job Losses by Problem and Pathological
Gamblers
Who Pays Problem Path.
! the Cost Gamblers i Gamblers
F~!~:i~~!\~~~1~~1
.
Employees' loss of earnings
Even though problem and pathological gamblers have elevated rates of job loss, there is
no systematic indication that they earned less than otherwise sirnilar individuals due to
either excess unemployment or lower wages. While problem gamblers had a lower
employment rate than expected at the time of the interview, we elected not to incorporate
this in our cost estimates, because the estimate does not appear to be supported by other
related measures. For example, problem gamblers were virtually identical to low-risk
and nongamblers in the proportion that had less than a full year of employment. Also,
their average hourly wage rate was virtually the same. This is not necessarily
inconsistent with an elevated rate of job loss. If problem and pathological gamblers are
less likely to voluntarily quit or leave jobs, their aggregate rate of unemployment could
be the same even though they are more likely to be fired or laid off. .
Bankruptcy, debt, unemployment insurance and welfare
Previous studies of GA and treatment populations have given a good deal of attention to
other [mancial impacts. Such studies have found, for example, that pathological
gamblers have high levels of debt and declare bankruptcy at higher rates than other types
of gamblers (and nongamblers). Research on gambling treatment populations found that
gambling-attributed current debt (as opposed to lifetime borrowing) was $39,000 in-'
Wisconsin and $114,000 in Illinois (Thompson, Gazel and Rickman 1996; Lesieur and
Anderson 1995). These studies found that in the GA/treatment populations, between 18
and 28 percent of males and 8 percent of females had declared bankruptcy.
However, debt per se is not unexpected or an indication of unusual problems, as many
individuals buy residences, automobiles, and other large purchases on credit. What is
unusual is when an individual declares bankruptcy, based on an inability to repay debt
when compared to the income of the individual or the household. When bankruptcy
occurs, some fraction of the debt may be never repaid, and it is this fraction of debt and
borrowmg that constitutes a loss to creditors (rather than the magnitude of borrowing or
indebtedness). However, this loss is theoretically considered a transfer, and generally is
not be included in "cost" estimates.
.
Pathological gamblers have clearly elevated rates of indebtedness, both in an absolute
sense and relative to their income. Indebtedness per person is 25-percent greater than
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 45
.
NORC
Ilml
UNtVERsiri OF CHICAGO
that of low-risk gamblers and about 120-percent greater than that of nongamblers.
However, the disparity is even greater when debt is compared to income: pathological
gamblers owe $1.20 for every dollar of annual income, while low-risk and nongamblers
only owe $0.80 and $0.60, respectively. In accord with their higher debt, pathological
gamblers have significantly elevated rates of having ever declared bankruptcy: 19.2
percent, versus 5.5 percent and 4.2 percent for low-risk and nongamblers.
Again, for problem gamblers the story is not as clear. Their average level of
indebtedness is actually the lowest of any type of gambler; however, they still have an
elevated rate of bankruptcy (10.3 percent), but this is only marginally statistically
significant when compared to the rate among nongamblers.
.
Table 13. Financial Characteristics and Impacts, by Type of Gambler
i-----.-.----.-.--..------.-F-~:~~~~~=~l~~~:r::~f~l=-~~~~ff!i~~:r~~r~~--T=-p~t=h~~=l
~~~=r:~~~~~ ~:;;~5;;~
1~~::~Jl;;;;~~';,lT ~136.000hX7;oooj$4a;OoO I $4S,OOOh46;oool
; mos. ROD ,I ;
! Household debt, I $22,000 I $38,000 I $37,000 I $14,000 $48,000 II
I current (ROD . ! I I I
I Filed bankruptcy, ever I 4.2 I 5.5 ! 4.7 ! 10.3<1> 19.2* I
Statistical significance of differences between groups tested using multivariate logistical regression, with
control variables for age, gender, ethnicity, education, child in household, and alcohol and drug use/abuse.
Gamblers with no problems were used as the base group.
Significance tests: pathological and problem types tested separately; statistically significant at the: *** =
0.0 I level; ** = 0.05 level. * = 0.10 level. Pathological and problem types combined for significance testing;
statistically significant at the: 4>4>4> =0.0 I level; 4>4> = 0.05 level. 4> = 0.10 level.
On average, excess lifetime losses involved with bankruptcy are about $3,300 for
pathological gamblers and $1,600 for problem gamblers. Almost 19 percent of
pathological gamblers have ever declared-bankruptcy, versus an expected 10.8 percent,
given their personal characteristics. For problem gamblers, their lO-percent rate
compares to an expected rate of 6.3 percent. Personal bankruptcies result in an average
of $39,000 in losses to creditors (WEFA Group, 1998; Gropp et aI., 1997), although one
should keep in mind that there are major differences between Chapter 7 and 13 filings.
Table 14. financial Losses, by Type of Gambler
~~~nei~i~~iir:~:j~~l~
L_____.____._E._9'__......__._.____.L__.._._.___________j_._________.._____ ._____.___._____..__..__.__..___j
.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 46
.
NORC
""'"
UN/'/Ep.Slri OF CHIC,1GO
Criminal justice costs
Pathological and problem gamblers in treatment populations often reveal that they have
stolen money or other valuables in order to gamble or pay for gambling debts (Lesieur
1998). Nearly half (46 percent) of GA participants in Wisconsin reported they had ever
stolen something to gamble, and 39 percent had been arrested (Thompson et al. 1996).
The GA survey in Illinois found that 56 percent had stolen to gamble (Lesieur and
Anderson 1995).
Although we asked study participants if they had ever stolen money in order to gamble or
pay a gambling debt, the reported frequency was too low to measure, or at least report in
this study. However, it was possible to obtain information about the frequency with
which respondents reported ever being_ arrested and/or serving time in jailor prison
(unfortunately, past-year rates were too low for analysis). However, these are only
indirect measures of the underlying issue that we would like to measure. Still, to the
extent that problem and pathological gamblers have rates of arrest and imprisonment that
are greater than low-risk gamblers and nongamblers, it is possible to infer that the
difference may be related to gambling behaviors and problems (although the direction of
causality may be open to debate).
.
Table 15 below shows that those with more gambling symptoms have much higher rates
of lifetime arrests and imprisonment. About one-third of problem and pathological
gamblers reported having been arrested, compared to 10 percent oflow-risk gamblers and
only 4 percent of nongamblers. About 23 percent of pathological gamblers and 13
percent of problem gamblers have ever been imprisoned. Again, these rates are much
higher than rates for low-risk gamblers and nongamblers (4 and 0.3 percent,
respectively).
For this analysis, we performed tests to establish the probability that these differences
were not primarily associated with other characteristics of the respective gambler types
(e.g., age, gender, alcohol and drug problems) and were not observed due to chance. The
arrest and imprisonment rates of problem and pathological gamblers were highly
significant.
Arrests
Pathological and problem gamblers account for about $1,000 each ($1,250 and $960,
respectively) in excess lifetime police costs. Almost one-third of each group has been
arrested or detained by the police at some time in their life (their expected rates are about
19 and 15 percent, respectively). Based on the survey, pathological and problem
gamblers had been arrested about 3.3 and 1.6 times, if they had ever been arrested. In
1992 (the most recent national data available), police spent $41.3 billion to make 14
million arrests (about $2,900 per arrest; U.S. Dept. of Commerce 1998). Thus, the 32
percent of pathological gamblers with arrest histories had about $10,000 in lifetime arrest
costs. However, the $10,000 must be prorated across all pathological gamblers, and
further adjusted for the 19-percent expected rate of arrest in this population. Thus, the
average cost per pathological gambler is $10,000 x (32% - 19 %), which equals $1,250.
.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 47
.
NORC
Ilml
UNr....EilSH). OF CHICAGO
Table 15. Weighted Occurrence of Criminal Justice Consequences, by Type of
Gambler
i-;;~e-:-----T~~;;~::;=-:L~~'~~~-:r::;a~
I c.~!~~!---r!!"'~.~~- At ~~-T~!'~1!>~L~'e,-j
~;;~t~~~il~~1t==I=it~~E~f~~J-tj
Statistical significance of differences between groups tested using multivariate logistical regression, with
control variables for age, gender, ethnicity, education, child in household, and alcohol and drug use/abuse.
Gamblers with no problems were used as the base group.
Significance tests: problem and pathological tested against low-risk gamblers; statistically significant at the:
*** =0.01 level; ** = 0.05 level. * = 0.10 level. Pathological and problem gamblers tested against
nongamblers for significance test; statistically significant at the: $iJ>iJ> = 0.0 I level; $$ = 0.05 level. iJ> = 0.10
level.
.
I ncarcetation
More than one-fifth of pathological gamblers have ever been incarcerated in a prison or
jail in their lifetimes (the survey did not ask about number of incarcerations). The
simulation indicates an expected rate of about 6 percent. The cost of incarcerations has
been estimated based on the ratio of national police and corrections spending. The most
recent survey of criminal justice spending found that total corrections costs were about
one-quarter smaller than total police spending (USDOJ 1996). This ratio has been
applied to estimate the lifetime incarceration costs for problem and pathological
gamblers. However, a further adjustment has been made to account for the fact that
pathological gamblers are much more likely to have been incarcerated, if ever arrested
(this is consistent with the findings that pathological gamblers have 3.3 arrests, if ever
arrested, compared to 2.1 for low-risk gamblers). Thus, pathological gamblers are
estimated to have $1,700 in lifetime corrections costs, with problem gamblers having
$670 in costs (see Table 16).
Table 16. Criminal Justice Losses, by Type of Gambler
[~~~~~-~;-~~_:':~-~I__':~:~~:;~~~~-~r~~~:E~~~~~;~::::::[~~;~~~~~;;~_:_~I~~~~~~~;:~l
I Arrests I Government Lifetime I $960 i $1,250 i
i Corrections i Government Lifetime ! $670 $1,700 i
Divorce
Family problems are one of the primary concerns associated with problem and
pathological gambling. Lesieur (1998) reports that between 26 and 30 percent of GA
members attribute divorces or separations to their gambling difficulties. While this type
of consequence is difficult to measure and to assign value to, the number of resulting
divorces can be measured, and legal fees can be estimated. One measure of gambling as
a factor in divorce is that respondents representing about 400,000 adults pointed to their
own gambling as a cause or factor in a past divorce, and respondents representing 2
million adults identified a spouse's gambling as a significant factor in a prior divorce.
.
The analysis estimates that the average pathological gambler has accumulated $4,300
more than expected for legal fees involved with excess divorces (measured rate of 53.5
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 48
.
NORC
......
UNiVERSlfr' OF CHICAGO
percent, versus an expected rate of 33.4 percent). Low-risk gamblers and nongamblers
have lifetime divorce rates of 30 and 18 percent, respectively. Problem gamblers have
losses of $1,950 in lifetime excess divorce legal fees. Their reported divorce rate was
39.5 percent, compared to a rate of 31 percent expected for persons otherwise similar
without gambling problems. Legal fees per divorce average $20,000 (Wilson). The costs
per problem and pathological gambler were developed by multiplying the average
number of divorces per gambler times $20,000 to get legal costs per gambler ever
divorced. This total was averaged over all pathological gamblers and adjusted down to
account for the difference between reported and predicted divorce rates.
Table 17. Marital and Health Status, by Type of Gambler
! I Lifetime Gambling Behavior
1._~t_C!t..~~.___..._...._._....J.__g~;~~~~.J.._.~ls.~......I.._~!.:..gi~~_I...~~:~~j;~.J..~:.~i~~_...j
f~~~~~~'~:=1===i!f-==F=litt=1*-::~:-~~~t=::ft}~-:1
_____....._.__....____..____._..__......__..__._._........___L.___._____....,..__.___._......_._..._.._L.._.___._____.____-'-____._.._..___.__.__._.__.___.J
Statistical significance of differences between groups tested using multivariate logistical regression, with
control variables for age, gender, ethnicity, education, child in household, and alcohol and drug use/abuse.
Gamblers with no problems were used as the base group.
.
Significance tests: Problem gamblers against pathological; statistically significant at the: *** = 0.01 level;
** = 0.05 level. * = 0.10 level. Problem and pathological tested against nongamblers A for significance test;
statistically significant at the: <P<P<P = 0.0 I level; <P<P = 0.05 level. <p = 0.10 level.
The economic consequences of divorce are actually much greater than the direct value of
the associated legal costs. The major economic conclusion from the divorce literature
(Everett 1991) is that the economic well-being of children and the mother usually
significantly falls, while that of males increases materially. Thus, there is a tragic
winner-loser scenario, where the values are somewhat offsetting.
These costs are clearly to be differentiated from the emotional cost that is borne by all of
those involved. The ability to calculate these economic costs in the present study is
limited, however, because the costs are quite complicated. They involve interpersonal
losses and gains by the adults and the children involved, and entail detailed information
about the timing and duration of marriage, divorce, and any remarriage. The current
study was not designed to perform such analyses, as it would be necessary to collect
equivalent and extensive data for the two parties to the divorce. However, it is possible
to describe the kinds and relative magnitudes of the economic impacts that prior research
on divorce has identified.
.
One study estimated that women with minor children suffered a 73-percent reduction in
their standard ofliving in the first year after divorce (Weitzman 1985). National statistics
show that married couples had a median household income of $47,000 in 1995, compared
to $21,000 for a female-headed household with absent husband (U.S. Department of
Commerce 1998). This results from a combination of factors, such as the infrequency of
awards of alimony (less than 20 percent of divorced women) and a scarce majority of
women receiving child support (U.S. Department of Commerce 1986). This factor is
further exacerbated by the fact that a significant fraction of child support and alimony
payments are never m'ade, and when women make recourse to courts, the legal costs can
easily consume a significant share of the payments in arrears.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 49
.
NORC
-
UflNEI1S1T'i OF CHfCAGG
Another prominent aspect of the economIC impact is that many mothers with small
children do not work or work only part time. Leaving the workforce (generally because
marriage makes this economically feasible), whether entirely or partially, impairs future
earning ability through a loss of valuable work experience (Mincer and Polachek 1978).
Also, part-time jobs generally entail lower skilled occupations with limited opportunities
for career and earnings growth.
Paradoxically, when a previously unemployed mother returns to the workplace due to
economic hardship associated with divorce, she experienced an increase in earnings; this
increase is considered to offset the loss in income contributed by the absent spouse.
However, this movement of a mother into the workplace in order to earn more constitutes
a net loss (in an economic sense) of her contribution to the household, since she can
spend less time engaged in child care and other household activities. One study-'
estimated the difference in the value of these services at about $12,000 per year (adjusted
for inflation; Paringer and Berk 1977). This is the value of services the mother can no
longer contribute to the operation of the household, because time is spent outside of the
home engaged in paid employment.
.
Probably the longest term and potentially the greatest economic cost is associated with
impacts on children. Several studies have found that children from divorced households
have lower academic and/or occupational achievement (Krein 1986; Cassetty and
Douthitt 1985). These costs would last for most of the lifetimes of the affected children,
and when discounted to their present value, could be in the tens of thousands of dollars
per child. Again, the economic costs of divorce are quite substantial, however they are
measured. This study has only represented a small-although very tangible-component
of such costs, because the study was not designed to undertake the level of sophisticated
analysis needed to make such estimates.
Health care
Several studies have suggested that pathological and problem gambling is correlated with
a decline in health and elevated rates of illness-either physical or mental (Lesieur 1998).
We did not identify research that examined personal health care utilization and
expenditures, or health status (generally the strongest predictor of health expenditures
within age and gender groups). It is unclear how gambling problems would cause
adverse impacts on health, although such impacts are believed to be a function of stress
and strain. In our survey, 33.8 percent of pathological gamblers reported that they were
in poor or only fair health, while only about 14 percent of low-risk gamblers reported
poor or fair health. We estimated that annual health care expenditures were elevated by
about $750 for pathological gamblers, with an estimated annual expenditure of about
$3,800 per capita. Based on their other characteristics, absent the effect of gambling, we
expected significantly fewer pathological gamblers to be in poor or fair health-about 17
percent, with personal health expenditures of about $3,000 per capita.6
.
This calculation used analyses from the National Medical Expenditure Survey and The Lewin Group's
Health Benefits Simulation Model to produce estiI:11ates of expected annual expenditures for population
groups defined on self-reported health status, gender and age (all available from NODS).
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 50
.
NORC
l!mI.
UN/VEr1S!TY OF O..;:c:.,GO
Table 18. Divorce and Health Costs, by Type of Gambler
I I ! Time Period of Problem I Path. i
L!~I?_~_~t~~!!__...__J_Y.'J1_~_~_~Y~__~.~~!J__.__.~_~!L,!,a!_~_.____.__ _<?_~_~PJ.~!._._..L~~_~bl_~L_._1
l,r-S~~EE~=-~:~~~=~-.~=~L~~~~~~~~.:::~=~_~~l~~i{~~~-.-.::-.:=.=::=~:_-.=:-.~:~~~~f-':+::_::_:=~~1[~~::-j
, , I I '
L Me~l_C!.Lt!.~.9-IJ.~_____.J._~s.~.!:.c.:f.1_~~__________._Ll.9-s.!_Y~~.~_______._______________~_;3_~Q_.__L______.._____~.;3.~Q.j
Mental health care
Pathological and problem gamblers had annual mental health expenditures about $330
and $360 greater than expected, respectively. About 13 percent of these two groups
reported past-year use of mental health services, while our analyses projected use by only
about 6 percent. Utilization of mental health services was just under 7 percent for low-
risk and nongamblers. In 1996, about $50 billion was spent on mental health care
(excluding psychiatric hospitals and residential treatment centers for children) to treat
about 10 million adults (about $5,000 per person receiving care; Mark et al. 1998).
Therefore, an excess of 7 percent of problem and pathological gamblers had mental
health problems, at an average cost of $5,000 per year, which yields the estimated cost
per problem and pathological gambler of about $350 per year.
.
Treatment for pathological gambling
There have been no national studies on the issue of pathological gambling treatment.
Volberg (1998) estimates that only about 3 percent of current pathological gamblers
obtain professional treatment in a given year (not including participation in self-help
groups like GA). This rate of treatment access is much lower than rates for persons with
current drug addiction (about one-third), alcoholism (about 15 to 20 percent), and other
mental disorders (ranging from about 40 to 80 percent). In Oregon, Volberg found that
public clinics had about 600 documented patients and/or affected family members per
year, compared to a current estimated prevalence of about 20,000 pathological gamblers.
Similarly, no substantial data exist regarding costs for treating pathological gambling.
Inpatient treatment facilities generally keep patients for several weeks, at a cost of up to
$10,000; outpatient providers treat patients for several months or more, often taking on
-' patients after they leave 24-hour care. V olberg (1998) reports that in Oregon, patients
generally receive care for up to 6 months in outpatient addiction treatment centers
(similar to the course of treatment for alcohol and drug addiction), although due to client
dropout, the average duration of treatment per patient is about 3 months. Costs in these
centers runs about $70 per week (Mark et al. 1998), suggesting average costs per patient
of between $900 and $1000.
In sum, about 3 percent of pathological gamblers seek care in a given year, with an
average cost per person of $1 ,000. If one uses these data to estimate the cost of treatment
in a year, then the annual treatment cost per pathological gambler is about $30. It is
assumed that most problem gamblers do not seek treatment unless or until they advance
to pathological. While in a given year a pathological gambler may have a 3 percent
probability of entering treatment, over a period of pathological gambling there may be a
greater probability that they will seek care.
.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adulis
Page 51
.
NORC
-
UNIVERSITY OF CHICAGO
Total costs of pathological gambling
The total costs estimated above are summarized in Table 19. Costs are shown for both
past year and lifetime, since certain costs could only be calculated in these respective
units. Annual costs of lifetime pathological gamblers are estimated at $1,195, compared
to $715 for lifetime problem gamblers. However, substantial additional costs are present
that can only be estimated on a lifetime basis, as they did not occur frequently enough in
the past year to be estimated with the current sample size. Lifetime impacts were
$10,550 and $5,130 for pathological and problem gamblers.
.
Table 19. Selected Economic Costs of Pathological and Problem Gambling:
Costs per Pathological and Problem Gambler
~'----."--'_._-..'.'-'~------'.-------'--"-._-.--'-_..-.--.------.--...--.-.-..--.~----....-......--.--------.---.-1
'I' ! I Problem I Pathological j
I I .,
I i I Gambler Costs Gambler Costs I
I I Who Pays I-.-...-----..-r-.p~.s-t-..-.-t-------..---..--.r.-..p~st---'1
~f~~~:~~1~~g=~~~=-~=I~~~9=-~1t1
i benefits: I I , j i
..-......--..--.--......--....--..--..-..-----....-...--.--.---.-....-.-......--............-__......._.._.__...__1......_..._.........-....-.-...r...----..-.-.-..-...-...-.-....--...-.....-..-.--....,.."
I Welfare benefits I Government i n.e. i $90 n.e. I $60 i
L____~___._._._._._.__._________._.___.___._._~.__._._--.-.-...........----...--.-.-----l.--.-.---.-u-.-....-..--.l...._._____....._..______......___. ._...._....._._............!
f~,~~E~~t90+~;~~;~~~i--j~~~g-I-~:tiH~~+--~:~
t~:~~;~~l~&&1~g~f~~i;~~tl
i Poor mental i Health insurance I n.e. I $360<91 n.e. I ($330@@ll
i health Iii I I i
D~g:~I:.Ei~i~ii~iI~9.-~i~~iiiQL::~::.:r:::::~:::~:~[I~,:::::~::g::t:::::::::~~i:+::J1ii-:-~J
~~~~~~;k~:~~~~~+~?~~;~-JI;gZ~
l T ra nsf~~.Q...g.2.~!?1 er~._____..._..._________._L_$..l.t~:?-9...__L_..._$ 1 5 5____.._._._$..~!~_QQ_.___.E~_j
@This is a net increase in cost. @@rhis is a part of tot a! health. n.e..: not able to be estimated in this survey.
We believe that the annual costs should be increased to incorporate some contribution
from the lifetime costs. However, the basis for making such an allocation is weak at the
present time. This study has found that past-year prevalence rates are about one-half of
that for lifetime prevalence, indicating that pathological and problem gambling is a
chronic problem for many, with the disorder going into remission and later recurring.
Future studies should consider collecting data about the course of gambling problems,
including the age of onset and the ebb and flow of gambling problems, in order to
ascertain the period of time over which costs are incurred.
.
The sums indicated in the table include several types of costs that are termed "transfers"
in the formal economic sense; they should be treated differently depending on the
calculation one is making. Transfers are those costs that mainly represent a shifting of
resources from one individual to another, with one person gaining what the other loses.
The transfers of concern in this model are the costs of bankruptcy and the value of
unemployment insurance and welfare benefits. These amounts accrue to the benefit of
the problem and pathological gamblers and to the detriment of either their creditors or the
government.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 52
.
NORC
lrml
UNlVERS!rr' OF CHICAGO
The costs of problem and pathological gambling minus transfers are $1,050 and $560 per
year, and $10,550 and $5,130 per lifetime, respectively. When these sums are multiplied
by the estimated prevalence of pathological and problem gamblers from the combined
RDD+patron data file (which was used for the cost calculations), they translate into
annual costs of about $4 billion per year, and $28 billion on a lifetime basis. If transfers
to the gambler from creditors and other taxpayers are included, the costs rise to about $5
billion per year and $40 billion per lifetime.
Annualizing the lifetime estimates is difficult without a fIrm estimate of the average
duration of problem and pathological gambling in the general population, which would
provide a denominator for the lifetime costs. If the average age of onset were in
adolescence or young adulthood and gambling persisted continuously or sporadically
throughout the adult life, the average duration could be as long as 50 years. In this case,
the lifetime costs would annualize to a present value in the neighborhood of $1 billion. A
lower estimate of the lifetime duration would increase the annual estimate.
Summary
.
This section demonstrates that problem and pathological gamblers experience a variety of
tangible consequences at rates that are significantly higher than would otherwise be
expected based upon their sociodemographic (and substance abuse) characteristics. Such
consequences include burdens to personal health, family, workplace, and the criminal
justice system. In other words, such gamblers impose costs on themselves, their families,
and on those around them, including employers, creditors, and taxpayers. It is possible to
estimate economic impacts experienced by, or at the level of, the individual problem or
pathological gambler. These estimates use standard and commonsense methods to attach
valuations on the consequences that could be measured. A verage annual costs per
pathological gambler are about $1,200 per year, and $715 per year per problem gambler.
"Lifetime" costs are estimated at $10,550 and $5,130. (Annual and lifetime costs should
not be added together, since they are measured over different time periods.)
It is instructive to compare economic cost estimates from this study with measurable
costs of other sources of morbidity, mortality, and productivity loss (see Table 20). The
annual cost estimate for pathological and problem gambling in 1998 of $5 billion
(somewhat more if we annualize the lifetime costs) compares with 1995 estimates for
drug abuse of $110 billion and alcohol abuse of $166.5 billion (Harwood et al. 1998).
Motor vehicle crashes in 1992 cost $71 billion (Blincoe and Faigin 1992). The most
recent estimates for other major health problems such as diabetes, stroke, and heart
disease have been compiled and compared by the National Institutes of Health (1997).
The current economic impact of problem and pathological gambling, in terms of
population or cost per prevalent case, appears smaller than the impacts of such lethal
competitors as alcohol abuse and heart disease. However, the costs measurable by
health-based estimation methods do not capture all of the consequences important to the
person, family, or society. The burden of family breakdown, for example, is outside of
these measures. And the value of further attention at the policy level may depend more
on the quality of efforts to respond as on the extent of costs we can presently measure.
.
Table 20. Economic Impacts of Major Health Problems
1---------------- ! Annua'j Cost-TPrevalenc-ej-.--A-nnu~iCost per-'j
~e of Problem ! (billions) I (millions) I Prevalent Case!
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Poge 53
.
NORC
......
UNIVERSin' OF CHICAGO
~i21~~~~~I~~~~I~~~~~~~~~
i Alcohol abuse .. [$166. i . 13.8 I. $7,000. i
t~~~f:--~~~~!-~~=:~~~=~=~~:~~E~~~:~l~I::-::~-:-~=r:~=:--~=:~~l==~:=~:::rrr:=~=~:~_::~ifi~~~~_:=::J
I Heart disease i $125 I 21 $6,000 I
lf~~t~~~~-I~=fiE:~[~~e~F~~~i~ ..
A major component of cost for most of these problems is the cost of treatment, which is
much more universally available, and administered much more often, to sufferers of
trauma, organic illnesses, and other mental and behavioral disorders than to pathological
or problem gamblers. Health care accounts for about one-half of the economic impact of
mental illness, stroke, heart disease, and diabetes. The measured economic impacts
therefore include the cost of society's determination to respond directly to such problems.
.
The findings of this part of the report directly raise the question of the extent to which
problem gambling behavior is the cause of the higher rates of consequences. This
analysis cannot rule out the possibility that the gambling problems are actually reflective
of certain underlying inclinations or values of these persons, such as a reduced
willingness to abide by social norms or an inclination to take extra risks (not simply in
gambling). To the extent that this is true, the gambling problems are as much
symptomatic of the other characteristics or issues as causes of difficulties in the life of
gamblers and their families. This is not to say that the gambling behavior is not in itself
damaging as documented in this study, but that the additional issues will probably need to
be addressed in order to ameliorate the tangible negative consequences of problem and
pathological gambling.
While the conclusions of this analysis are relatively robust, they must be tempered by
several factors. The small sample size was a limiting factor in the analysis. There were
too few problem and pathological gamblers in the survey, even after the random digit dial
and the patron surveys were combined and weighted to generate cost estimates for
consequences that were directly attributed by interviewees to "gambling problems." All
of the costs that have been estimated 'are associated with excess rates of consequences
that can be caused by factors in addition to problem and pathological gambling.
Analyses have been done to adjust for selected other factors such as alcohol and drug use,
age and educational attainment. Adjustment for these factors does result in smaller
estimates of costs than would otherwise result simply by comparing problem and
pathological gamblers to nongamblers and those with no problems.
.
Finally, the costs that we measured are tangible and relatively amenable to economic
analysis. However, many of the human burdens of pathological and problem gambling
are not so readily quantifiable into dollars, for conceptual and practical reasons. For
example, we calculated the cost of divorce in terms of the legal fees generated to
complete divorce actions through the court system. The cost in legal fees hardly begins
to capture all of the social and psychological meaning of divorce for the partners and
families directly involved, and for society as a whole. The economic costs that we
calculated are a lower bound. Without a substantially greater research base on the
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 54
.
NORC
-
UNNERSrn. OF CHICAGO
characteristics and consequences of pa,thological and problem gambling, it is impossible
to say with precision where the upper bound or midpoint of economic impact would lie.
Annex 1: Description of Outcome Variables
The economic and econometric analysis of the consequences of gambling problems
examined the following respondent-specific outcome variables:
. Not working, dummy variable set equal to one if the respondent is unemployed or not
in the labor force;
. Employed less than 12 months in past year, dummy variable set equal to one if the
respondent, if employed at all, was employed for less than 12 months in the past year;
. Lost job in past year, dummy variable set equal to one if the respondent lost or was
fired from his or her job in the past year;
. Employed in low-wage job, dummy variable set equal to one if the respondent's
hourly wage is $10 per hour or less;
.
. Received unemployment insurance in past year, dummy variable equal to one if the
respondent reported having received unemployment insurance/disability benefits in
the past year and reported being in fair, good, or excellent health;
. Received welfare benefits in past year, dummy variable equal to one if the respondent
reported having received any welfare benefits, including AFDC/TANF, general
assistance, and foster care payments, in the past year;
. Low-income household, dummy variable set equal to one if the respondent's annual
household income is less than $24,000;
. Bankruptcy, dummy variable set equal to one if the respondent ever filed for
bankruptcy;
. In poor or fair health, dummy variable equal to one if the respondent reported being
in poor or fair health;
. Mental health treatment in past year, dummy variable equal to one if the respondent
reported having gone to a clinic, doctor, or counselor, or outpatient treatment for
problems with his or her emotions, nerves, or mental health in the past year;
. Ever divorced, dummy variable equal to one if the respondent reported ever having
been divorced. This model only includes those respondents who have ever been or
are currently married;
. Ever arrested, dummy variable equal to one if the respondent reported ever having
been arrested; and
.
Gombling Impact ond Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adulis
Page 55
.
NORC
11IM
UNlVEHSfTY OF CHICAGO
.
Ever incarcerated variable equal to one if the respondent reported ever having been
incarcerated. The patron survey does not ask about incarceration; therefore, this
model only includes respondents from the RDD survey.
Annex 2: Description of Explanatory/Independent Variables
All of the logistic regression models include the following respondent-specific
explanatory variables:
· Dummy variables for lifetime gambling behavior, including never gambled in
lifetime, gambled with one or two problems in lifetime, gambled with three or four
problems in -lifetime, and gambled with five or more problems in lifetime, with
gambled with no problems in lifetime serving as the excluded base category;
· Age, included as a continuous variable in both linear and quadratic forms;
· Dummy variable for sex, set equal to one for men and equal to zero for women;
· Dummy variables for race, including black, Hispanic, and non-white other, with
white serving as the excluded base category;
.
· Dummy variables for current marital status, including married, separated, divorced,
and widowed, with never married serving as the excluded base category;
· Dummy variable for household with children, set equal to one if the respondent lives
in a household that includes at least one child under the age of 18;
· Dummy variables for educational attainment, including did not attend Ith grade,
attended Ith grade, attended technical school, attended one to three years of college,
and attended four years of college, with attendance at graduate/professional school
level serving as the excluded base category;
· Dummy variables for alcohol and drug use and abuse: Questions on alcohol and
drug abuse were not included in the patron survey; therefore, this set of variables
includes four variables specific to respondents to the RDD survey and two variables
specific to respondents to the patron survey. The RDD variables are: use of alcohol
at least 12 times in the past year; use of at least one drug on five or more days in the
past year; abuse of/dependence on alcohol; and abuse of/dependence on drugs. RDD
respondents coded as alcohol abusers/dependent and/or drug abusers/dependent are
not coded as being alcohol and/or drug users, respectively. The patron variables are
use of alcohol at least 12 times in the past year and use of at least one drug on five or
more days in the past year;
· Dummy variable for survey type, set equal to one if the respondent participated in the
patron survey and zero if the respondent participated in the RDD survey; and
.
· Dummy variables for region of residence, including dummy variables for the
Midwest, South, West and Missing, with the Northeast serving as the excluded base
category.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 56
.
NORC
......
UN/VERs/no OF CHICAGO
Annex 3: Methodological Notes for Costs
.
The following table presents certain values and calculations used to estimate the cost per
problem and pathological gambler. Specifically, the estimates of this study compare the
rate of costly consequences for these gamblers relative to "predicted" or expected rates
for inoividuals with similar characteristics, but who are low-risk gamblers (they have
gambled, but never experienced any symptoms of problem gambling).
Specifically, the analysis adjusts for a standard set of characteristics that are believed to
be predictive of the behaviors and outcomes of interest in this report. These factors were
identified and the variables were defined in the body of the report. They include age,
gender, ethnic identity, educational attainment, use/problems with alcohol and drugs,
respectively, and region of the country in addition to variables representing the gambling
Type of the individual. The purpose of these calculations is to adjust for basic and
systematic differences between different types of gamblers that might be related to the
outcomes of interest, rather than simply take the difference in outcomes for pathological
and problem gamblers and compare them to those with no history of problems.
The analysis has compared problem and pathological gamblers to low-risk gamblers
(never had any problems) rather than nongamblcrs for several reasons. First,
nongamblers have lower rates of problems than low-risk, thus we get more conservative
(smaller) cost estimates when we use low-risk gamblers for comparisons. Second, other
researchers have found persons that have never gambled to be relatively distinct in their
characteristics, e.g., less likely to work outside of the home, more likely to be female, and
otherwise more conservative in certain behaviors.
The costs are based on the "excess" or difference between the actual rate and the
predicted rate, where the predicted rate is calculated from the "odds ratio." This yields a
smaller or more conservative estimate than simple comparison of problem and
pathological gamblers to the unadjusted rates for low-risk and nongamblers.
For all of the costs examined that rate .of problems is materially greater for problem and
pathological gamblers than for low-risk gamblers (and nongamblers). Costs have only
been estimated where the rate for pathological and/or problem gamblers is statistically
significantly different (worse) than low-risk gamblers (or alternatively nongamblers).
For comparison purposes the rate of consequences/problems for low-risk gamblers is also
presented. Note that for all types of consequences except one the "predicted" rate of
problems for problem and pathological is greater than the unadjusted rate for low-risk
gamblers. This indicates that problem and pathological gamblers on average are more
likely to have characteristics that are associated with the consequences of concern, even if
they were not problem gamblers. For example, other tabulations have shown that
problem and pathological gamblers are more likely to have alcohol and drug problems
and lower educational attainment. If these factors are not adjusted for the cost estimates
will be somewhat inflated, as having these characteristics (alcohol and drug problems) is
generally significantly and negatively related to measures such as divorce, health, and
criminal justice involvement (as is the case in the analyses done for this study).
For example, the problem of "job loss" was reported by 13.8 percent of pathological
gamblers who had been employed during the prior year, compared to a rate of 4.0 for
low-risk gamblers. In the logistical regression the "odds ratio" is 2.62, which means that
the odds of pathological gamblers experiencing job loss is 2.62 times greater than for
low-risk after adjusting for other characteristics. These data imply that pathological
gamblers without their gambling problems would have a predicted rate of 5.8 percent.
.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 57
.
NORC
m>!I
UNlVEffSln'OF CHICAGO
This is greater than the value for low-risk gamblers of 4.0 percent, due to the other
characteristics which indicate that pathological gamblers are at higher risk of job loss
even without the gambling issues.
Predicted rates are estimated from the rates for pathological and problem gamblers,
respectively, and their "odds ratios" from multivariate logistical regressions comparing
each respective type of gamblers to low-risk gamblers. Odds ratios (and accordingly
costs) are only used where problem and pathological gamblers are significantly worse
than low-risk gamblers or those that have never gambled at the p<O.l 0 significance level
(one-tailed test) or better.
.
Table 21. Summary of Comparisons Between Pathological, Problem,
and Low-Risk Gamblers
j I Rate of !Odds Ratiol Predicted I Rate for ,
IType of Costly I Consequence [' Relative to I Rate withoutl Low-Rjs~ j
Iconsequence/pr~ble~___LPer .f!'.oble.!!!.. Low. Risk.___~ambling--L~amble'.:~
Pathological Gamblers
.-...--..--.-----....--..---.--..--.-.----..------r--.-.--...--.-""'"f---..---..-----..-.-r.-..-.--...--..----.
Uob loss i 13.8% ! 2.62 I 5.8% I 4.0%
jUnemployment Insurance _ 15.0% 2.81 I 5.9% I 4.0% I
IWelfar_~~~~i!~_____..__~-.---- 4.6%1___..-2:.?i.____~--2.4% --.l__1.3~_1
IBankruptcy I 19.2% 1.97 I 10.8% I 5.5% 1
; Divorced ever I 53.5% I 2.29 I 33.5% I 29.8%
t-;-;---;---------..-..--:--------L---____.______._.._.______
IHealthpoororfair i 31.1% I 2.43 I 15.7% I 13.9%
r---n-----......-----------......--...--..).-.---...---__..____..._..-.-.I.--.---.----..-----.r.-----.--....-----~.--...-..-----_______,
,Mental health utilization I 13.3% i 2.12 i 6.7% I 6.5% !
IArrested ever I 32.3% I 2.00 I 19.3% L-~~~
,'ncarceration ~ve!___.._______._L__~~!.~_J.____.i:~~._.L___~. 3% _.I____.~_P% ---l
i Problem Gamblers I
IJob loss I 10.8% I 2.07 I 5.5%, 4.0%
~~pl~~e~!i.~~~+_10~% -+----~.:3-~f-- 5.3~l___4.0~_
Welfare benefits I 7.3% 3.35 I 2.3% I 1.3%
Bankruptcy I 10.3% 1.71 I 6.3% I 5.5% I
t
~ ------. -.--------,---------______l______--;____.__
.~~v~~~~~;~~e~~air --.------+--i-~:~~-------~-~~:-..-~---~-?~-~~---t--- ~~~~-~-----.
Mental health utilization I 12.8% 2.47 I 5.6% I 6.5%
Arrested ever . I 36.3%' -- 3. 15 __L~:3o/". _I. -.!.]-: 1 % --1
Incarceration ever -----r--1O.5% - 2.34 I 6.2% 4.0%
.
References
Blincoe, L., and Faigin, B. 1992. The Economic Cost of Motor Vehicle Crashes.
Washington DC: Office of Plans and Policy, National Highway Traffic Safety
Administration, U.S. Dept. Of Transportation.
Everett, C., ed. 1991. The Consequences of Divorce: Economic and Custodial
Impact on Children and Adults. New York, The Haworth Press, Inc.
Frazis, Harley, et al. 1998. Results from the 1995 Survey of Employer-Provided
Training. Monthly Labor Review, June 1998, v121, n6, pp. 3-13.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 58
.
NORC
-
llNl1/EnS!rY OFCH!CAGO
.
Gropp, R., Scholz, J, and White, MJ. 1997. Personal Bankruptcy and Credit Supply
and Demand. The Quarterly Journal of Economics, 112(1), February.
Harwood, H., Fountain, D., And Livermore, G. 1998. Economic Costs of Alcohol
Abuse and Alcoholism. In Recent Developments in Alcoholism, Vol. 14. Marc
Galanter, Editor. Plenum Press, New York and London.
Lesieur, H.R. 1998. Costs and Treatment of Pathological Gambling, Annals of the
American Academy of Political and Social Science (Gambling: Socioeconomic
Impacts and Public Policy, lH. Frey, special editor). March 1998.
Mark T., McKusick, D., King, E., Harwood, H., Genuardi, R. 1998. National
Expenditures for Mental Health, Alcohol and Other Drug Treatment 1996. For
the Substance Abuse and Mental Health Service Administration, DHH_S
Publication No. SMA 98-3255, Rockville, MD.
Mincer, l and Polachek, S. 1978. "Women's Earnings Re-Examined." Journal of
Human Resources 13(1): 113-34.
National Institutes of Health. 1997. Disease-Specific Estimates of Direct and
Indirect Costs of Illness and NIH Update. Department of Health and Human
Services.
Paringer, L., and Berk, A. 1977. Cost of Illness and Disease, Fiscal Year 1975.
Washington D.C.: Georgetown University, Public Services Laboratory.
Thompson, W., Gazel, R., and Rickman, D. 1996. The Social Costs of Gambling in
Wisconsin. A Report prepared for the Wisconsin Policy Research Institute.
U.S. Department of Commerce. 1998. Statistical Abstract of the United States,
1997. U.S. Government Printing Office, Washington, DC.
U.S. Department of Justice. 1993. Correctional Populations in the United States,
1991. Bureau of Justice Statistics Publication NCJ-142729, Washington, D.C.
U.S. Department of Justice. 1996. Sourcebook of Criminal Justice Statistics 1995.
Bureau of Justice Statistics Publication NCJ-158900, Washington, D.C.
Volberg, R.A., Moore, W.L., Christiansen, E.M., Cummings, W., and Banks, S.M.
1998. Unaffordable Losses: Estimating the Proportion of Gambling Revenues
Derived from Problem Gamblers, Gaming Law Review 2 (4): 349-360.
Volberg, R.A. 1998. Gambling and Problem Gaming in Oregon. A Report to the
Oregon Gambling Addiction Treatment Foundation. Gemini Research, Ltd.,
Northampton, MA. Also at http://www:gamblingaddiction.org/oregonreporU
frame.htm
WEFA Group. 1998. The Financial Costs of Personal Bankruptcy. Burlington,
MA, February. Available online at http://www.wefa.com/wefabkr.html.
WEFA Group. 1998. A Study Concerning the Effects of Legalized Gambling on the
Citizens of the State of Connecticut.. Burlington, MA.
Weitzman, L.l 1985.. The Divorce Revolution: The Unexpected Social and
Economic Consequences for Women and Children in America. New York: The
Free Press.
Westphal, l, Rush, J., and Stevens, L. 1998. Problem and Pathological Gambling
Behaviors Within Specific Populations in the State of Indiana. A Report to the
Indiana Department of Administration and Family Social Services
Administration.
.
Gambling Impact and Behavior Study
Chapter 3. Economic Anolysis of Consequences of Gambling Problems Among Adulis
Page 59
.
NORC
......
lINNEnS!n' OF CH!CAGO
Wilson, C.A., as reported in Marital Status: The Comprehensive On-Line Resource
for Divorce and Remarriage. Internet at: http://www.maritalstatus.coml
divorce/articles/dividing cover.htmI.
.
.
Gambling Impact and Behavior Study
Chapter 3. Economic Analysis of Consequences of Gambling Problems Among Adults
Page 60
.
NORC
mIlD
I)NIVEFlS1TY OF CHfCAGG
CHAPTER 4. GAMBLING AMONG 16- AND 17-YEAR-OLD YOUTHS.
As part of its data collection protocols for the Commission, the NaRC team interviewed
534 youths via a randomized telephone survey of U.S. households during the last 2
months of 1998. Surveys of small age groups, and particularly of minors, are more
complicated than general adult surveys, even when the same interview is being used.
(albeit some skip patterns were widened and others narrowed; see below). The
differences are attributable both to the screening requirements, under which the great
majority of households have no eligible respondent, as well as to the need to obtain two
contacts and consents for the interview-one from the parent and then one from the
youth.
.
No more than 7 percent of households have a 16- or 17-year-old in residence. Because
youths of this narrow age band are so rare, relatively speaking, they are time-consuming
to reach for interviewing. The cost of screening to obtain a sufficient number for typical
national estimation purposes is much higher than for an adult survey. Consequently,
researchers may take one of the following approaches: (1) the survey is conducted by
group administration in schools, (2) the survey protocol accepts a much wider age group
(such as 9- to 17-year-olds), or (3) residences are screened for a more extensive protocol,
so the relative cost of the screening is less significant. The cost of selecting youth for a
highly specialized one-interview sample survey is such that we recommend further
national-level research on adolescents instead be performed in the context of ongoing
longitudinal or cross-sectional research, in which gambling questions can be appended to
one or more rounds of questionnaires, rather than as a stand-alone survey.
In the present survey of youth gambling behavior, we began by obtaining two kinds of
randomized phone lists: a random-digit-dial sample like those used for the adult sample,
and a random selection of household telephone numbers from lists that were known to be
"enriched" with adolescents. The latter lists contain phone numbers that, due to their
neighborhood location or other known household characteristics (e.g., school enrollment
or consumer expenditure patterns), have a much higher than random likelihood of
providing access to adolescents. Numbers from this second type of list were in fact much
more accurate, in that about 19 of every 20 were found to be working residential
numbers, in contrast to about one-half of the phone numbers on the RDD lists. The
enriched lists also had a much higher percentage of 16- and 17-year-olds among the
successfully screened cases; at present, we have not yet finalized our calculations for
these separate sample sources.
The screening procedure for the youth sample was somewhat different from the adult
procedure. As with the adults, we first asked for the number of residents age 18 or older,
and then 17 or younger, in the household. If minors resided there, we then asked whether
one or more was of the requisite age; if there were more than one, we randomly selected
the youth with the most recent birth date. We then asked to speak with the child's parent
or guardian, from whom we would request consent to interview the child. Only in cases
where we obtained parental consent did we pursue further contact with the youth, at
which point we would request her or his personal consent to take part in the survey.
.
A small number (n=49) of the youth cases were obtained from the fully randomized lists;
all others were reached through the enriched lists. In examining the composition of the
Gambling Impact and Behavior Study
Chapter 4. Gambling Among 16- and 17'Year-Old Youths
Poge 61
.
NORC
11IM
UNiVERSff,' OF CHICAGO
total sample this procedure gave us, we found that the North Central region of the United
States was over-represented, and that black and Hispanic youth were under-represented.
We therefore weighted the sample by poststratification (described previously for the adult
surveys) on region, race, and ethnicity to yield proportionate weights equal to those
obtained in Current Population Estimates. The final sample represented 8.3 million 16-
and 17-year-olds, with a slight preponderance (less than 1 percent) of males; the sample
was 74-percent white, 13-percent black, 9-percent Hispanic, and 3-percent assorted other
backgrounds. Nearly 36 percent lived in the South, 19 percent in the Northeast, and 23
percent in the Midwest (North Central) and in the West. Further calculations revealed
that 82 percent lived in lottery states.
Our overall finding was that adolescents gamble appreciably less often than adults.
About one-third of the 16- and 17-year-olds have never gambled, versus less than one-
seventh of adults. However, the most striking finding in the youth sample was not the
lower overall participation rates, but the difference in their reported pattern of gambling
when compared with that of adults. The past-year data most clearly demonstrate this
difference. As indicated in Figure 9 below, adolescent gambling was predominantly
composed of private betting on games of skill, particularly card games (named by more
than 40 percent of those who mentioned a favorite game). Nearly 3 out of 10 youths,
versus just more than one-tenth of adults, bet on such games in the past year.
.
Figure 9. Past-Year Gambling
Participation by Type of Game
60
51.8
50
40
II>
Dl
o
C
II>
...
~
II>
IL
30 25.7
-11 91 102
:: - ,., ,., 3.' "'.' r-.
0-' 11 _~[]II._~__~_
Casino Track Store Binga Unlicensed Lottery
28.2
Private
D 18 years or older
. 16 or 17 years old
The other most prominent youth games, albeit much less popular than private games of
skill, were betting in sports pools and buying lottery tickets. Youths particularly favored
instant lotteries; about three-quarters of the young lottery players bought instant (scratch-
off type) lottery tickets in their most recent purchase period, and no more than 15 percent
bought either multi-state, daily, or big-jackpot tickets.7 For adults, the lottery was the
.
7 There is a reasonable prospect that a small payoff, such as that delivered by most wirming instant-lottery tickets,
could be collected by an underage gambler. The likelihood that a youth could collect on a winning Powerball
ticket, or for that matter a MegaBucks slot machine, is very small (Cummings, personal communication, 1999).
Gambling Impact and Behavior Study
Chapter 4. Gambling Among 16- and 17-Year-Old Youths
Page 62
.
NORC
......
UNIVEAS!TY OF CHICAGO
most frequent mode of gambling. More than one-half of adults bought lottery tickets in
the past year, versus about one-eighth of 16- to 17-year-olds. In addition, adults strongly
favored lotteries with big payoffs; during their most recent purchase period, more than
one-third bought Powerball-type (multistate) tickets, more than one-half bought big-
jackpot (state) tickets, and fewer than one-quarter bought instant lottery tickets.
Casino gambling (especially slot machines) was the second most common form of adult
gambling, with one-quarter. of all: adults participating in the past year. The adolescents
were notably absent from casino play, with barely 1 percent reporting any casino wagers.
This presumably reflects well on the enforcement efforts (particularly against fake IDs)
of casino operators, among other factors.
.
The data also show that 16- and 17-year-olds have wagered (and won or lost)
substantially smaller amounts of money when compared with adults. For example,
approximately 22 percent of adults ever lost more than $100 in a single day of gambling,
compared with only about 2 percent of the 16- and 17-year-olds. When asked about their
wagering in the past year, about one-quarter of all adult lottery players reported losing
more than $100, while less than 2 percent of youth who played the lottery reported losing
this much. Even in their preferred form of gambling-private games-only about 2
percent of all 16- and l7-year-olds lost more than $100 in the past year, compared with 6
percent of adults.s Finally, youths who gambled did so less frequently than adults who
gambled. For example, when looking at those who did play private betting games in the
past year, only one-fourth of the 16- and 17-year-olds, compared with one-third of adults,
made private bets at least once a month in the past year. .
If we use adult guidelines standards to gauge the sheer financial riskiness of youthful
wagers, we would have to conclude that adolescent gambling is not nearly as serious a
problem as adult gambling. But this would be a premature conclusion. For example,
many 12- and 17-year-olds hold part-time jobs and earn incomes. These incomes are
generally much lower than those of adults, and few adolescents are in a position to "bet
the rent." However, the amounts they wager may in fact comprise an appreciable
percentage of the income they do control. Unfortunately, the pattern of income questions
in the survey was poorly suited to adolescent economic circumstances, which require a
different approach. The relationship between the discretionary income of adolescents and
their pattern of gambling is an important subject for further research.
The NODS screen was a second area in which there was a slight difference between the
adult and adolescent survey, but in this case a more revealing one, in that all adolescents
who ever gambled were asked the NODS questions, without regard to reported their
reported levels of gambling losses. By using the answers to the daily and annual greatest
loss questions, it is still possible to apply to youth the same parameters when considering
their "problem gambling type" as we applied to adults-namely, that gamblers who have
never lost more than $100 in a single day, or as a net yearly loss, are automatically
considered to be low-risk gamblers. When we apply these parameters to our youth
sample, about 2 percent (roughly 150,000 youths) would then be classified as at-risk
gamblers, which is about one-fourth the proportion seen among adults. About 1.5 percent
.
S
We should note that, of the population of youth who have lost more than $100 in a single day, or netted a
$100 loss in any given year, about two-thirds are male.
Gambling Impact and Behavior Study
Chapter 4. Gambling Among 16- and 17-Year-Old Youths
Page 63
.
NORC
mm.
UNJ~'Er.SITY OF CHICAGO
(roughly 100,000 youths) would be classified as problem or pathological gamblers,9
which is less than the figure for adults.
Most prior research on adolescents has not applied the same yardstick to youthful
gambling behavior as to adult gambling, but instead designated behavior as problematic
with a lower required number of behavioral criteria, or by admitting less severe levels as
diagnostically important than the same researchers would accept for adults. If, for
example, the adult financial criterion that we used is not considered, so that all 16- and
17 -year-olds are screened with the NODS regardless of the largest amount lost in a day or
year, the percentage of problem and pathological and problem gambling youth doubles to
about 3 percent, which is similar to the percentage for adults. Furthermore, the
percentage of at-risk youth increases even more dramatically, tQ about 15 percent-which
is more than double the incidence among adults. .
.
There is no single "right" decision on what kind of yardstick to use at this stage of
developing research on youthful gambling. It is plausible to argue that the limited
discretionary funds available to adolescents are largely spent to purchase entertainment,
and that private games of skill and luck, around which most youthful gambling occurs,
provide a relatively protected environment in which to learn how to discipline one's
gaming expenditures. That is, one can say that it is potentially a good thing for youths
interested in gambling to be able to discover, with relatively limited financial exposure,
that excessive wagers can be costly and disruptive, both in dollars and in the absorption
of time that might be devoted to other pursuits.
But one can also argue that these "protected" environments, in which there is no
continuing "house cut" from the money being wagered, present a misleading and
seductive picture of gambling as it exists for most adults. Even the notion of protection
may be mistaken, insofar as adolescents may be prepared to enforce the collection of
debts (or to attempt to evade collection) with less discipline and more ill temper than in
the more businesslike world of regulated adult credit and loan arrangements. Moreover,
one might argue that many forms of learning compete for the limited attention of youths
in the chrysalis of adolescence, and the real cost of time spent learning how to gamble-
that is, how to manage and live with the operations of chance-is the lost time from
learning about other foundations of life, including responsibility, love, trust, charity, and
work. Finally, an. early introduction to gambling, even if it proved salutary for many
adolescents, may also encourage earlier onset and a longer and more difficult course for
those who are vulnerable to development of gambling problems.
At any rate, depending on which measurement approach one prefers, adolescents can be
seen as less at risk of gambling problems than adults, about equally at risk, or at greater
risk. A deeper understanding of the economic, social, and psychological dimensions of
adolescent gambling is very much needed.
9
The sample sizes of problem and pathological gamblers are too small to permit comparisons between them.
.
Gambling Impact and Behavior Study
Chopier 4. Gambling Among 16. and 17-Year-Old Youths
Page 64
.
NORC
~
UNlVc.I1SJry OF CHICAGO
CHAPTER 5. IMPACTS OF CASINO PROXIMITY ON SOCIAL AND
ECONOMIC OUTCOMES, 1980-1997: A MULTILEVEL TIME-SERIES
ANALYSIS
Casinos have opened in and near many U.S. population centers during the past two
decades, especially in the 1990s. To determine the effects of these establishments, we
analyzed social and economic changes between 1980 and 1997 in 100 non-tribal sample
communities. The outcome (or response) variables are the published social and
economic indicators for the counties of the sample communities by year, from 1980 to
1997. These _variables include annual county-level measures of criminal activity,
employment, income and earnings, bankruptcy filings, and demographic behavior.
The key explanatory variable is "casino proximity," in other words, whether one or more
casinos were in operation within a 50-mile radius of a given community between 1980
and 1997.10 In the sample of 100 communities, casino proximity was an irreversible
community attribute during these years; in other words, all communities that were near at
least one casino had this attribute steadily from their first year of casino proximity up
until 1997. Only 5 of the 100 sample communities were near casinos in 1980, while 45
were near casinos in 1997. Thus, any effects of casino proximity in this sample are
largely attributable to casino openings in 40 communities between 1980 and 1997.
.
A statistical model is needed to analytically separate the effects of casino openings from
the effects of other factors operating during this timespan. From a statistical perspective,
the laO-community sample is close to being ideal for this purpose. Both the sample of
communities, as well as the total populations residing in the counties of the sample
communities, are roughly balanced between communities that were near casinos and
communities that were not. Based on county-level data, the 45 communities near at least
one casino in 1997 accounted for about 46 percent of the approximately 42 million
individuals who resided in the 100 sample communities. Moreover, in the subsample of
45 communities that were near at least one casino in 1997, the majority of years from
1980 to 1997 antedated the opening of any nearby casino. The average year of first
casino opening equals approximately 1990.6, and about 90 percent of the 45 communities
had their first casino opening in 1989 or later. The statistical model seeks to isolate the
effects of casino proximity by (1) comparing communities with and without a nearby
casino between 1980 and 1997 and (2) comparing years before and after first casino
openings in communities that were near a casino sometime during 1980-1997.11
.
10
Altemately, we use the phrasing "near [one or more casinos]" in this section to mean "casino proximity."
IIIn contrast to the favorable balance of the IOO-community sample, the worst sample composition for
statistical purposes would occur if casinos opened near all sample communities in the same year. No model
could hope t9 isolate casino effects given such a data set, because casino effects could not be distinguished
from effects of other factors, such as changes in the U.S. economy, that tend to affect all communities in the
same way at the same time.
Gambling Impact and Behavior Study Page 65
Chapter 5. Impacts of Casino P(Oximity on Social and Economic Outcomes, 1980-1997
.
NORC
-
UNIVEF?SHY OF C!i,'CAGO
Data
A set of social and economic indicators was selected for analysis based on hypotheses
about the effects of casino proximity presented in Reuter (1997). For example, casinos
may create conditions conducive to crime, including larcenies and burglaries that target
gamblers carrying large amounts of cash; the risks associated with gambling may result
in non-business bankruptcies; casinos may provide employment; visitors who come to an
area to gamble at casinos may boost earnings in other sectors, especially hotels and
construction; and visitors who come to the casinos may drain business away from other
sites. These are all empirical questions.
.
Social and economic indicators for counties of the 100 sample communities were taken
from a number of published sources, including crime data from the Federal Bureau of
Investigation (FBI); employment, earnings, and income data from the Bureau of Labor
Statistics (BLS); health indicators from the National Center for Health Statistics (NCHS);
and numbers of business and non-business bankruptcies from the Administrative Office
of the U.S. Courts. Estimates of per capita casino spending were provided by
Christensen/Cummings Associates. An important caveat is that the estimates of per
capita casino spending are not completely independent of the key explanatory variable-
. casino proximity-because the Christensen/Cummings estimation algorithm used
proximity along with other variables, including reported revenues of local gambling
facilities, to approximate casino spending (Christensen/Cummings Associates 1999).
Prior to analysis, all dollar measures-including casino spending, income measures, and
earnings measures-were adjusted to 1980 price levels using the consurner price index
(CPI-U) for all items (Bureau of the Census 1997, p. 487).
To standardize outcomes across communities, we calculated per capita rates and means
by dividing the estimates of the numbers of crimes, bankruptcies, and dollars of earnings
or income of counties in specific years by corresponding estimates of the population sizes
of the same counties in the same years.12 For example, the larceny rate is calculated per
100,000 people (i.e., 100,000 x (number of larcenies)/( county population size)). Per
capita income is calculated as (number of dollars of income )/( county population size).
Details are provided in Table 22, at the end of this section.
Years of availability and communities for which data are available also vary depending
on the data source. For example, crime rates are available for 92 counties and years 1980
to 1995, while most earnings and income measures are available for all 100 communities
from 1980 to 1996 (see Table 22 for details). Sporadic missing values are also present in
the time series of specific communities. We chose a method of analysis designed to yield
robust results in the presence of missing values.
.
12The appropriate county population bases for rates in particular years vary slightly depending upon the data
source. The population bases of the rates and means analyzed in this section, and the population bases used
in constructing analytical weights (next section), are from the agencies that published the numerators of the
corresponding rates and means. Thus, the bases of crime rates are county population estimates of the FBI,
and the bases of average income and earnings estimates are county population estimates of BLS and the
Bureau of the Census. The base of the unemployment rate is the population size of the civilian labor force
(BLS), and the base of the infant mortality rate is number oflive births (NCHS).
Gambling Impact and Behavior Study Page 66
Chapter 5. Impacts of Casino Proximity on Social and Economic Outcomes, 1980-1997
NORC
.
-
UNI\/ER5!ff OF CHICAGO
Methods
A statistical model used to analyze survey data is a precisely stated concept about how
different measured (and unmeasured) features of the subject under study relate to one
another, such as how casino proximity relates to community characteristics over time.
By computing how well the data fit the model, we are able to infer the accuracy of its
conceptual structure and the strength of postulated relationships.
Multilevel models are models that represent the multilevel (hierarchical) structure of a
dataset and allow the random error component of a model to depend on units at different
levels. In the time series of social andeconornic indicators, there are two "levels"-years
(repeated measures of a specific indicator for a particular community) are "nested" with
communities (counties). That is, for each community, we have as many as 18 annual
measurements for years between 1980 and 1997 . Years within communities are "level-l
units," and communities are "level-2 units."
.
Multilevel models are especially appropriate for estimating casino prOXImIty effects,
given this data structure, for two reasons. First, community-level effects are probably
important in these data, because there are many factors-including factors that are
unmeasured or poorly understood-that vary across communities and affect the levels of
outcomes differently in different communities. Multilevel models allow inc'orporation of
a community-specific intercept that varies across communities and serves as a proxy for
diverse unmeasured factors that differentiate the communities in the sample. The results
corroborate the importance of community-level effects: For most outcome measures,
more than one-half of the total variance in the outcome measure is attributable to the
community-specific intercept, meaning that most of the variance in the outcome measure
is among, rather than within, communities (see Table 22). Second, unlike traditional
multivariate models for multiple time series, multilevel models can accommodate
missing values because the models do not require a balanced data structure (Goldstein
1995). The outcome measure can be available for different communities in different
years.
Another important modeling decision is the choice of scale for the outcome variable.
Prior to the multilevel analysis, we transformed each outcome measure by taking the
natural logarithm, for two reasons. First, most of the outcome variables are highly
positively skewed, and without exception, the mean level of the outcome measure across
communities exceeds the median. Based on histograms and residual plots, it appears that
the log transformation worked well in normalizing the distributions of the outcome
variables. Second, plots of the community-level means by community-level standard
deviations showed that the standard deviations tended to increase linearly with the means.
This is the case of "constant coefficient of variation," common in social datasets. The log
transformation operates to make the variance of the outcome measure constant across
communities, thereby simplifying the models that can be used to analyze the data (e.g.,
McCullagh and Neider 1989, Chap. 4).13
.
13The log transformation applied in our analyses was a slight modification of the usual transformation,
namely log(Y + I), where Y denotes the outcome measure-a rate, percentage, or mean. The addition of unity
obviates the problem of occasional zero rates and means.
Gambling Impact and Behavior Study Page 67
Chapter 5. Impacts of Casino Proximity on Social ond Economic Outcomes, 1980-1997
.
NORC
...".
UNIVEf?Slrr' OF CHICAGO
Given the log transformation applied to rates, means, and percentages, statistical
efficiency is enhanced by weighting the observations in proportion to the population
bases of these statistics (McCullagh and NeIder op.cit.). We used the weighting
technique available in the multilevel program MlwiN (Goldstein et al. 1998) to weight
communities in each analysis in proportion to their mean populations across years during
1980-1997 when the outcome measure was available. This procedure may tend to
understate the "good" effects of casinos, if any, with respect to employment, increased
tourism, and so forth, because in larger communities these effects would tend to be less
salient than in smaller ones (e.g., 100,000 "tourists" coming to gamble to Minneapolis/St.
Paul will have a lower relative impact than the same number coming to Tunica, Mississippi).
Adverse impacts (in terms of greater gambling-related pathology among local residents)
would appear to be more independent of the size of the community.
The results presented in this section are based on four "nested" multilevel models, called
Models 0, 1, 2, and 3. Each of the four models incorporates one or more of three classes
of explanatory variables:
· Community, represented by a community-specific intercept;
· Year, represented by a set of 0-1 variables indexing years when data were available;
and
· Casino, represented by a single 0-1 variable measuring whether a casino was within
50 miles of a specific community in a specific year.
.
Each model assumes that the logarithm of the outcome variable equals a linear function
of (1) one or more of the three classes of variables and (2) a "level-I" random error term,
i.e., a random error that varies across years within communities. Each model assumes
that both the community-specific intercept and the Ievel-1 error are normally distributed
in the population and mutually independent. Each model was estimated using likelihood-
based methods discussed in Goldstein (1995) and implemented in the program MlwiN
(Goldstein et al. 1998).14
14 The most general model IS Model 3, which can be written In the following manner:
97
~i = U + U)" + L Y, I;i' + P CASINOi' + e;i '
, ,.81' ~,
whereYij denotes the outcome measure for community j in year i; Iij, denotes a 0-1 dummy variables that
takes on the value I if i = t and takes on the value 0 otherwise; and CASINOij is a 0-1 dummy variable that
takes on the value I if community j has a proximate casino in year i and takes on the value 0 otherwise.
The parameters a, P, and the y,'s are fixed constants. aj and eij are random variables assumed to be normally
and independently distributed with zero mean and constant variance. The variance of aj is the level-Z
(cammunity-Ievel) variance, and the variance of eij is the level-I variance. The community-specific intercept
of the j-th community equals (a + aj)'
Models 0, I, and 2 result from setting specific parameters equal to zero in Model 3. Model 0 assumes that P
and the y,'s equal zero. Modell assumes that the y,'s equal zero. ModelZ assumes that P equals zero.
.
Directions for improving Model 3 include allowing for serial correlation of successive values of ejj.
Estimation of such models is currently possible but nontrivial in the multilevel framework (Goldstein 1995;
Gambling Impact and Behavior Study Page 68
Chapter 5. Impacts of Casino Proximity on Social and Economic Outcomes, 1980-1997
.
NORC
""""
UN/t/ERS!TY OF CH!CIIGO
The models differ in which of the three classes of explanatory variables are included, as
follows:
. Model 0: community only
. Modell: community + casino
. Model 2: community + year
. Model 3: community + year + casino
The presentation of results in the next section focuses on the comparison of Models 2 and
3. For each outcome measure arid model, community effects were highly statistically
significant. For each outcome measure and each of Models 2 and 3, year effects were
higWy statistically significant. For each outcome measure, the casino effect is significant
in Modell, but it is significant for only selected outcome measures in Model 3 (see the
next section). Estimates of year effects are highly consistent between Model 2 and 3, but
estimates of casino effects vary dramatically depending upon the choice of Model 1 or
Model 3.15
.
The presentation of results focuses on the comparison of Models 2 and 3 because year
effects represent an important competing hypothesis for the apparent effects of casinos.
The opening of casinos in recent years has been correlated with important changes in
outcome measures in many communities (Reuter 1997). The inference that casinos are
causing such changes in outcomes can be discounted to the extent that the changes are
fully accounted for by changes that occurred in specific years both in communities with
proximate casinos and in those without proximate casinos. The question is whether the
addition of the casino effect in Model 3 results in a statistically significant increment to
the goodness of fit of Model 2. In effect, the sequence of the model development serves
to control for changes that occur in communities independently of whether casinos are
becoming more accessible to them.
The comparisons between Model 2 and Model 3 in the next section use a likelihood-ratio
chi-square statistic, gauging the difference in goodness of fit between the two models.
We infer a casino effect if and only if the chi-square (one degree of freedom) is
statistically significant at the standard a = 0.05 level. It is also possible to use the Wald
criterion, comparing the casino parameter estimate to its estimated standard error, but the
likelihood-ratio chi-square criterion is generally more accurate in small samples
(Goldstein 1995).16
Goldstein et al. 1998). NORC has obtained preliminary estimates, extending Model 3 to allow serial
correlation, for several of the outcome variables presented in the previous section. This is an important area
for future research.
15For example, the efFect of casino proximity is to increase total per capita bankruptcy filings by 51 percent
according to Modell, but, as discussed in the next section, the same effect of casino proximity is small (+3
percent) and statistically insignificant in Model 3.
.
16A current limitation of multilevel modeling techniques is the absence of an overall measure of the
goodness-oF-fit of a model, analogous to the R2 measure used in linear regression analysis. We evaluated
Model 3 by examining plots of residuals, which did appear to corroborate the assumptions of the model.
Gambling Impact and Behavior Study Page 69
Chapter 5. Impacts of Casino Proximity on Social and Economic Outcomes, 1980-1997
.
NORC
-
UNlvERs!n" OF CHiCAGO
The log transformation of the outcome variable implies that the casino proximity effects
in Model 3 are multiplicative on the original scale of the outcome measure. This means
that casino effects can be expressed as the percentage change in the outcome measure that
is associated with the opening of a nearby casino. For example, based on Model 3, the
opening of a casino is associated with an increase of 237 percent in annual per capita
casino spending (this is discussed further in the next section). Thus, if annual per capita
casino spending equals $10.00 in a particular community before the opening of any
proximate casino, we would expect annual per capita casino spending to equal $33.70 in
the same community after the opening of a proximate casino.
Results
Table 22 presents comparisons of Models 2 and 3 and casino proXimity parameter
estimates based on Model 3. Table 23 presents detailed Model 3 parameter estimates for
four of the outcome measures with significant results: per capita casino spending,
percentage unemployed, per capita income maintenance, and per capita earnings from
hotels and lodging. To facilitate combining estimates, the estimates in Table 23 are
presented on the log scale rather than on the untransformed scale. For example, from the
fIrst column, the estimated mean per capita casino spending (that is, gambling
expenditures) in communities without proximate casinos during 1990 equals exp(2.72 +
1.24) = $52. The same estimate in communities with proximate casinos during 1990
equals exp(2.72 + 1.24 + 1.22) = $178.
.
We draw the following conclusions. First, the casino effect is not statistically significant
for any of the bankruptcy or crime outcome measures or for the infant mortality measure
(which is as close to a common measure of child welfare as can be obtained). This is not
to say that there is no casino-related crime or the like; rather, these effects are either small
enough as not to be noticeable in the general wash of the statistics, or whatever problems
that are created along these lines when a casino is built may be countered by other
effects. Also, the index crimes measured here are generally those which reflect the public
safety and security of persons and real property-that is, "Part 1" crimes reported to the
Uniform Crime Report system of the Federal Bureau of Investigation. The "Part 2"
offenses, which include "white collar" crimes such as embezzlement and fraud, are not
yet available for analysis.
.
Second, there is a statistically significant casino effect for (1) per capita casino spending
(that is, local gambling expenditures in casinos), (2) 4 of 5 employment measures, and (3)
7 of 16 income and earnings measures. The largest significant effect is for per capita
casino spending, which rises 237 percent. There is also a marked decrease in the
percentage of the labor force that is unemployed, or -12 percent from an average
unemployment rate of 6.5 percent; in other words, about one percentage point is taken off
the unemployment rate. A marked decrease is also seen in the receipt, on a per capita
basis, of income-maintenance (welfare) dollars (-13 percent), unemployment insurance
( - 17 percent), and other transfer payments (- 3 percent), which may be associated with
the drop in local government employment. There is a slight increase in the employment
rate in construction trades, matched by a larger increase (+18 percent) in actual per capita
construction earnings, an important component of the local labor market. Even more
substantial percentage increases are seen in earnings in hotel and lodgings (+43 percent)
and recreation and amusement industries (+22 percent); however, these start from quite
small bases in the communities under study. No change is seen in overall per capita
Gambling Impact and Behavior Study Page 70
Chapter 5. Impacts of Casino Proximity on Social and Economic Outcomes, 1980-1997
.
NORC
......
UN/VERS!n: OF CHICAGO
income, as the increases noted above are offset by the reductions in welfare and transfer.
payments as well as a drop-off in income from restaurants and bars, which may be losing
business to the food and beverage services in casinos and hotels. The net picture in the
economic and crime data is on the positive side, but not in an overwhelming way. There
appears to be more of a shift in the types and locations of work, and perhaps the overall
number of workers, than a rise in per capita earnings.
Table 22, Casino Proximity Effects in Model 3. p=casino proximity effect as %
change. p=% variance between communities. Baseline level=avg. in yrs. w/o proximate casinos.
---.------.----...--.---- ------~.-.--. .------.-------l
' Significance of~: I .
Baseline I Model 3 v. Model 2 Model 3 Estimates I
1.-- .
level I X2 (1 df) p ~ p
Outcome measure
Casino spending I
j Per capita casino spending
I Bankruptcy filings per 100,000 pop.'
0.00'
+237%' I
75%
.
j Total filings . --- -wa-- 0.5! 0.48 --r+3%I- 98% -I
1-~~~fJi;;g;----.---------.------- --'--'85-- -"---'2:1I---o-:T5-t----:;:9o/,,--t----92'%--J
I Non-business filings 1 ,054 0.5 I 0.48 I + 1 % 1 98% i
Crime'
FBI crime index 4,400 0.1 I 0.75 -0% I 96%
FBI modified crime index 4,430 0.1 I 0.75 -0% I 96%
larcenies per 100,000 pop 3,863 0.3 I 0.58 +2% I 69%1
I i
I 8urglaries per 100,000 pop 1,326 2.5 ! 0.11 -7% I 74% 1
I Motor vehicle thefts per 100,000 _._~ -~~l 0.75 +0% I 81% I
I. ~~~~I::e~~~7~~g~~~~~;p---'_-'-- ---fH-=~_- ~:~ ~- ~:~~ --'I--'-'~~- ~;~ -.-j
Employment-civilian labor force composition"
,----.---.--------.---.----.---,..-------- -.--27.2.-To.oo~r_---=-12%-;-r-- 53%-1
I % unemployed 6.5%
I % employed-local govt. 8.9% 5.4 0.02' -2%' 45%
% employed-construction 6.1% 15.6 I 0.00' +1%' I 58%
r- 0.75 _ i
~.Ioyed-services ____ 26.7% 0.1 +0% I 87%
I
% employed-retail trade 18.8% 7.9 I 0.01' i -3%' ! 82%
Demographic behavior
Infant deaths per 1,000 live births 9.2 2.0 I 0.16 I +7% ! 29% I
i
Income and earnings per capita.
Total incon1e $16,153 --~-- ~_~O% 94!~
Lincome maintenance .--'.._-
$157 __66~~ 0.00' -13%' 95%
-retirement $1,866 0.6 I--O~r--':I%- -~%-
-transfer payments $2,094 14.5 0.00' I -3%' 94%
-unemployment insurance $70 20.5 0.00' -17%' 58%
I __::,~!....:.arnings _ $10,976 1.2 I 0.27 +1% I 99%
fPrivate earnings: construction -- !
$679 5.6 0.02' +18%' 51%
-restaurants, bars $241 11.9 ! 0.00' -19%' I 25%
L -gen. merchandise _ $124 3.0 f....- 0.08 J -15%! 29%
._-- ----
I -local govt, govt enterprise ____ $ 83-4.._ 2.2 -0:14 I -1 % r~5%
r-::'h;;~I~dg;;;g----------' _M'__'M__'~_~_ 1---.0.01 ' -~I"---+43o/;'--r--55%--
$64 7.1
1 -recreation & amusement $64 4.3 0.04' +22%' I 51%
-reta il trade $1,104 0.1 I 0.75 -0% I 88%
-services $2,354 ~..9.:U_~~_ +1% I 83%
-=;;cial se~ic;s -- --
$69 0.1 ! 0.75 -2% I 42%
I
I -transportation $769 3.9 0.06 +11% I 71%
'Significantly different from 0 at level a = 0.05. 1. Data for 100 counties, 1980-97; source: Christensen/Cummings. 2.
Data for 100 counties, 1987-97; source: Administrative Office of U.S. Courts. 3. Data for 92 counties, 1980-95; source:
FBI Unifonn Crime Reports. 4. Data for 100 counties, 1980-94 except 1987; source: BLS. 5. Data for 100 counties,
1982-93 except 1983; source: NCHS. 6. Data for 100 counties, 1980-96 for per capita income, 1980-95 otherwise;
source: BLS.
.
Gambling Impact and Behavior Study Page 71
Chapter 5. Impacts of Casino Proximity on Social ond Economic Outcomes, 1980-1997
.
NORC
~
UNlVErlSln' OF CHICAGO
Table 23. Model 3 Parameter Estimates: Selected Outcome Measures
.
Estimate on log scale (standard error)
Casino Unemploy- Income Earnings,
Parameter --_._~!:~~~~~-- ment maintenance hotels
~--------'-'---~--'--- ------.-..---. ---.-.-.--.---..---..-. -.-..-.--......-..-------
Av!L.~~~~ve~._1_~~2_" f-_3.23J.~.:2~L_ __~:?.:IJQ:_2~L__ . __.'!2!J.2:9_9L.__ ..__]~~9..:..1jL
Year 1981 0.22 (0.04) 0.05 (0.02) 0.19 (0.02) 0.15 (0.15)
Year 1982 0.34 (0.04) 0.27 (0.02) 0.24 (0.02) 0.18 (0.15)
f----------
Year 1983 0.43 (0.04) 0.27 (0.02) 0.34 (0.02) 0.21 (0.15)
Year 1984 0.55 (0.04) 0.05 (0.02) 0.38 (0.02) 0.46 (0.15)
Year 1985 0.64 (0.04) 0.03 (0.02) 0.43 (0.02) 0.61 (0.15)
Year 1986 0.70 (0.04) 0.02 (0.02) 0.49 (0.02) 0.59 (0.15)
Year 1987 0.84 (0.04) 2.53 (0.02) 0.59 (0.15)
Year 1988 0.95 (0.04) -0.18 (0.02) __ ~_.65 i9._:Q31__. 0.75 (0.15)
Year 1989 ____~_~g~j'2:. 04 L__ -0.21 (0.02) 0.79 (0.02) -_...Q~~-~~~---
--~-_._._--_.._--_._-_. f----------- 1--.-----------
Year 1990 _..~.~ 2 4j'2:. 04 L_ ~-Q:~J2:..Q.3.L- __~!jS!..:02) ___ _~.:09 .i9..2?1___
-----.------..-.----.----
Year 1991 1 .28 (O~~_ _-=-0.03_ (0.02) 1.16 (0.02) f-_l.30 (0.15)
1-----------
Year 1992 1.37 (0.04) -0.04 (0.03) 1.36 (0.02) 1.29 (0.15)
Year 1993 1.61 (0.04) -0.03 (0.03) 1.46 (0.02) 1.45 (0.16)
Year 1994 1.85 (0.04) -0.10 (0.03) 1.54 (0.02) 1.40 (0.16)
Year 1995 2.08 (0.04) 1.19 (0.02) 1.16 (0.16)
Year 1996 2.19 (0.04)
Year 1997 -- 2.31 (0.04)
-- 1.22 (0.03) -0.13 (0.02) -0.14 (0.02) _-.2.:36 (0.131_
Casino proximity
_. ..--
Level-2 variance 0.20 (0.03) 0.03 (0.01) 0.35 (0.05) 1.43 (0.21)
(communities) ------
Level-2 variance (years 0.07 (002) 0.02 (0.001) 0.02 (0.001) 1.15 (0.04)
within communities)
References
Bureau of the Census, 1998. Statistical Abstract of the United States 1997.
Washington: U.S. Dept. of Commerce. .
Christensen/Cummings Associates, 1999. Personal communication with NaRC.
Goldstein, H. Multilevel Statistical Models. London: Edward Arnold, 1995.
Goldstein, H., M. Healy, et aI., 1994. Multilevel time series models with applications
to repeated measures data. Statistics in Medicine 13: 1643-55. -
Goldstein, H., 1. Rasbash et al., 1998. A User's Guide to MLwiN. London: Institute
of Education, U. of London.
McCullagh, P and J. NeIder, 1989. Generalized Linear Models, 2nd ed., London:
Chapman and Hall.
Reuter, P. The impact of casinos on crime and other social problems: an analysis of
recent experiences. Report for the Greater Baltimore Committee. College Park,
MD. January 1997.
.
Gambling Impact and Behavior Study Page 72
Chapter 5. Impads of Casino Proximity on Social Clnd Economic Outcomes, 1980-1997
.
NORC
".".
UN/VERSlrr- OF C,'iICAGG
CHAPTER 6. CASE STUDIES OF THE EFFECT ON COMMUNITIES OF
INCREASING ACCESS TO MAJOR GAMBLING FACILITIES
As a part of the NaRC team's work for the Commission, we conducted a 10-community
case study on the impact of increased access to legalized casino gambling. We randomly
selected our sites from eligible towns and cities across the United States; the one
requirement was that the community have a population of at least 10,000 and lie within a
50-mile radius of at least one major casino, as defined by the criterion of atleast 500
gaming devices, and measuring the distance with commercial place-mapping software
with longitude/latitude coordinates. We stratified the selection of cases so that all four
census regions of the United States are represented by at least-two communities.
.
Once the communities were selected, we began the process of deciding who the
appropriate contacts would be for a telephone interview process. The set of possibilities
included local planners, addiction specialists, social service personnel including
employment specialists, law enforcement officers, chamber of commerce members, and
other persons representing a spectrum of community experience and viewpoints. We
developed a series of open-ended questionnaires that consisted of a core set of questions
plus items that were tailored to each particular type of community member. We then
compiled lists of potential contacts from recommendations of NaRC field staff residing
in or near these communities, as well as phone directories of government numbers and
the local yellow pages. Finally, in the closing section of each interview, we included an
item asking the respondent whether ~here was anyone else in the area they would
recommend we speak with; this method yielded a number of additional interviews.
Once this process was complete, three experienced interviewers were selected to contact
and interview respondents, speaking with seven or eight people in every community.
These respondents were often in prominent positions, such as heads of local planning
boards and chambers of commerce, chiefs of police, executive directors of mental health
agencies. and addiction centers, community development directors, newspaper editors,
consumer credit counselors, attorneys, and program directors in local and county social
. .
servIce agencIes.
We selected our respondents based not only on their positions, but also on the length of
time they have lived in the area and how long they have been in their respective lines of
work. Many of our respondents have lived in their communities all their lives, and of
those who have not, many have lived in the area for most of their careers. One interview
had to be excluded from use due to respondent noncompliance. Our interviews averaged
approximately one-half hour each.
.
In the sections below, we begin with a brief introduction of the various types of gaming
available in the case study communities at the current time. We then provide an
aggregate overview of the case study sites, discussing common threads we have noted
among the diversity of cities and towns we examined. Finally, we present two of the case
studies in detail. Each case study is an independent, indepth look at how a community
has responded to legalized casino gaming; taken together, they are a study in contrasts of
the very different ways that people's lives can be affected (or not) by the introduction of
caSInOs.
Gambling Impact and Behavior Study Page 73
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
......
UN/VEF?Sln' OF CHICAGO
The Communities
It is important to keep in mind when reading these case studies that people are reporting,
more often than not, their own subjective impressions of numbers of people with
problems, of crimes and bankruptcies, and so forth. In many cases, the population of an
area has increased at least modestly (and in some cases dramatically) due to the
introduction of one or more casinos, and in at least two of our case studies, the population
has increased due to factors that were unrelated to the casinos. What cannot be
concluded based on these vignettes alone is whether the proportion of cases concerning a
particular outcome, negative or positive, has changed.
Where possible, we attempted to verify the descriptions of the types of gambling
available in and near the case study communities. With this exception, we should
emphasize that the information we report here, including the direct and indirect quotes
from respondents, is based on our interviews alone and have not been validated or
investigated through external means.
.
Types of gaming
The sites we chose to examine for our case studies all had a variety of gambling
facilities 11 within the city limits and nearby; we asked respondents to tell us which
facilities were available within the town proper, as well as any facilities within a 50-mile
radius that drew residents. All of these communities were within driving distance of a
casino per our selection criteria. Since these casinos are recent additions to the local
landscape, all being introduced within the last 9 years, the issues surrounding their
introduction were fresh in our respondents' memories.
Casinos
Until quite recently, casinos were confined to Nevada and Atlantic City. However, in the
last decade, casino gambling has spread to towns and cities across the country, partly in
response to the need for additional revenues for local and state governments, and partly
as a result of the Indian Gaming Regulatory Act of 1988. The earliest efforts to legalize
casinos outside of Nevada and Atlantic City came in South Dakota and Colorado, where
small-stakes casino gambling for the purpose of historic preservation was approved by
referendum ill several old mining towns. Published research has reported that the impact
on these towns has been enormous, with skyrocketing property values, conversion of
many businesses to casinos, increased traffic, and increased crime (Long, Clark & Liston
1994, in Cox et al. 1997).
11
.
According to the Commission's directives, we defme the term "gambling" as "an activity that involves
staking or risking something of value upon the outcome of a contest, sporting event, or game of chance, with
the understanding that one may receive something of value in the event of a certain outcome" (RFP, p. 7).
Hence, we interpret a "gambling facility" to be any structure where such activity takes place. For the
purposes of this study, we do not include bona fide business transactions governed by the securities laws for
the purchase or sale of stocks, bonds, or securities.
Gambling Impact and Behavior Study Page 74
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
......
UNNEf?srrr OF CHICAGO
Riverboat casinos in particular have also spread rapidly in the 1990s, and at least one of
our communities is in proximity to one or more riverboats.18 The first riverboats,
legalized in Iowa in 1991, placed strict limits on both wagers and losses. As riverboat
casinos were legalized in other states, including Illinois, Indiana, Louisiana, Mississippi,
and Missouri, these limits were lifted. While these casinos must be located on facilities
that look like boats, few of the riverboats actually leave shore. In Mississippi, as well as
in Iowa where the earlier restrictions have been eliminated, the term "dockside gambiing"
is a more accurate description than "riverboat gambling" (Cox et al. 1997), since the
minimum requirement is that the gambling occur above a body of water, which may pre-
exist or be dredged for the purpose.
The Indian Gaming Regulatory Act created a regulatory structure for gambling on Native
American lands throughout the United States. By establishing a framework for.
negotiation between the sovereign tribes and state governments, Congress opened the
door for Native American tribes to establish casino-style gambling in any state where
charitable or social gambling is permitted (Eadington 1991). In 1996, more than 700
privately owned or Native American casinos were operating in 28 states (International
Gaming & Wagering Business 1997). At least 3 of our 10 case study sites are based near
one or more tribally owned casinos.
.
Lotteries
States across the nation have been implementing lotteries since the mid-1960s to fund
government services. Today, 37 states have legalized lotteries, and we found state and/or
multi-state lottery games to be present in each of our communities with the exception of
one. With cutbacks in federal spending, pressures on state lotteries to provide revenues
for government programs have increased. State lotteries now offer a multitude of games
that blur the boundaries between their traditional products and other types of gambling,
including instant or scratch tickets, daily numbers games, and electronic gaming devices
offering keno, poker, and line games similar to slot machines at casinos (Cox et al. 1997).
In only one of our case study communities did the lottery seem to be a problem for a
significant proportion ofresidents.
Pari-mutuel gaming
In contrast to lotteries and casinos, the pari-mutuel industry has undergone a dramatic
struggle to stay competitive in the last 10 years. We found a strong representation of
pari-mutuel gaming in our case study sites, with two communities reporting jai alai
frontons in the area, and seven reporting one or more dog and/or horsetracks. However, a
full six of the sites reported that one or more tracks have moved or gone bankrupt due to
the increased competition from the casinos. In response, racetracks have sought relief
from taxation from state legislatures and have also sought to expand their activities.
Initially racetracks worked to increase access to their traditional product by establishing
off-track betting systems and broadcasting races from other tracks at their own facilities.
More recently, racetracks have sought to compete by offering other types of gambling. In
.
18
To ensure the confidentiality of our informants, we have tried wherever possible to eliminate information
that could readily identify an individual community. We found in the course of analyzing results from these
sites that the type of casino, particularly whether it was water- or land-based, did not appear to have any
differential impact on the communities. Therefore, we do not identify riverboats as such in our profiles.
Gambling Impact and Behavior Study Page 75
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
......
UNNEASI rf OF CHICAGO
California, several racetracks now have cardrooms where patrons may wager on poker
and other games. In Delaware, Iowa, Rhode Island, and West Virginia, racetracks have.
been permitted to add electronic gaming machines to their traditional products, often with
excellent results for their bottom line (Cox, Lesieur, Rosenthal, & Volberg 1997).
Respondents in two of our case study communities reported local referenda on this issue.
Video gamillg devices
The innovation widely thought to have the greatest impact in recent years has been the
electronic gaming device, specifically video lottery terminals (VLTs) and video poker.
Widely publicized incendiary remarks by one clinician who called video gaming the
"crack cocaine" of gambling (Bulkeley 1992), and observations by helpline counselors
who report increased calls from video poker players, indicate the potential value of
undertaking systematic studies of this form of gambling. There are now at least 21 states
where electronic gaming devices such as slot machines or video lottery terminals are
available (McQueen 1996). Of our 10 case study communities, respondents in 5 reported
video gaming devices operating (legally and illegally) in local small businesses. No
respondent mentioned this type of game to be particularly addictive or harmful; however,
respondents in five communities specifically opined that casinos-in which video gaming
devices are the predominant form of gaming-generate more problems for gamblers than
other types such as the lottery or racetracks.
.
Other types of games
In addition to these games, all but one community reported bingo. All of these sites had
charitable bingo, and some also had commercial bingo. In addition, three communities
were reported to have bookmaking, and one community each was said to have a sports
bar, a bar with a pull-tab dispenser, and keno.
One of the interesting if unexpected outcomes of our interviews was that when we asked
respondents about the effects legalized gaming has had on their communities,
respondents seemed to forget all the other kinds of gaming in town and just speak to the
casinos. This is understandable if a community has no other gaming except charitable,
which many people do not perceive as really "gambling" but, rather, giving money to a
good cause. However, communities reported to have, on average, five major game types.
One possible explanation for this focus on casinos may be simple timing. In other words,
since, according to our respondents, bingo parlors, lotteries, and tracks have typically
been operating in their communities for at least 15 years, it seems reasonable that these
types of games have become more a part of day-to-day reality, and that fewer people visit
these types of facilities today, as in the case of racetracks and bingo.
.
Economic outcomes
A variety of economic themes came through in our case studies; some reflected positively
on casinos and gaming as a whole, while others did not. The silver linings and dark
clouds seemed inextricable from each other. For example, a recurrent positive theme was
the new employment opportunities that casinos create for local people. In five
communities this was cited as a very positive advantage (all but one of these communities
was located within 10 miles of a major casino). Respondents in the other four
communities indicated that unemployment remained a problem, despite former hopes to
the contrary. In two of these sites, as well as in four that cited more jobs, residents
Gambling Impact and Behavior Study Page 76
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
NORC
.
11IM
lINiVEnS!rr' OF Ch'!C1GO
complained that the casinos for the most part provide low-paying and/or part-time
positions with no benefits (four of these communities were located within 10 miles of a
major casino, and two, between 20 and 40 miles of a major casino). One respondent
indicated that because of the part-time nature of most of the jobs available in his town,
the welfare rolls had not decreased, even though some families had found work in the
casinos. He stated that the reason for this is that people are afraid to take jobs without
benefits; while on welfare, they have some assurance that if something happens to
themselves or one of their family members, their medical bills would still be paid.
Similarly, in another community, a respondent reported that the welfare rolls have
decreased, but not due to the casinos, because they only brought part-time jobs. On the
other hand, one respondent did indicate very clearly that welfare rolls have decreased due
in part to the opening of casinos. 19
.
In addition to casino operations as such, there was growth in the hotel and motel industry
(three communities), as well as more funds flowing either directly or indirectly from the
casinos into local government (six communities) and into charitable organizations (three
communities). Respondents in five communities indicated an increase in construction,
and four of these mentioned housing construction in particular. Two communities were
reported to have had property values increase. In communities where casinos were built
within the city limits, one of the most common concerns was the burden on public
infrastructure. One town was able to use increased revenues from casinos to upgrade its
water and sewage systems; four reported that either road improvements were needed or
that such improvements had taken place but more work needed to be done. In six
communities, at least one resident complained about the increased traffic brought by the
casinos. Three communities reported a growth in the number of retail establishments, but
two reported a decline since the casinos were built, with a variety of places going out of
business.
All but one of the commumtIes reported an increase in debt problems and/or
bankruptcies. One possible explanation that is likely at least in part for this, as told to us
in most of these communities, is that people are gambling on their credit cards and taking
out cash advances at A TMs in or near the casinos. However, it is also true according to a
few of our respondents that because new casinos promise jobs and financial stability for
disadvantaged persons or families, many people move to these towns with their existing
debt problems to try for a better life. - If the only work they find available is part-time
service positions, this may lead to further financial distress and eventually, in severe
cases, bankruptcy. Respondents in five communities, all but one of which had casinos
within the town limits, noted that one of the primary problems in the community was the
large number of working poor. Three communities also mentioned problems with
homeless ness, and about four mentioned problems with low-income housing or housing
stock.
.
19
The vast majority of U.S. communities located within 50 miles of a major casino are not located near a
casino with unionized workers, as most unionized casino employees work in the relatively larger, destination-
style/resort casinos located in Nevada and Atlantic City. However, due to the concentration and size of these
casinos, their workers comprise a fair percentage of the total number of casino workers. The sample of case
studies would have to be larger and selected to reflect union presence (for example, 10 communities with and
10 without unionized casinos) to permit comparison of the differential effects (if any) that unionization would
have on the type of jobs available and related economic effects.
Gambling Impact and Behavior Study Page 77
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
~
U!'JI',,'Ei1$JTY OF CHICAGO
Other social benefits and costs
Two communities reported an overall reduction in crime in their communities, and three
reported an overall increase. Interestingly, the youth in particular appear to be having
problems; five communities reported increases in youth crime, one specific to illegal
gambling, and one including illegal gambling. Two of these communities were among
the three reporting an overall increase in crime, and one was a community reporting an
overall decrease in crime. Illegal gambling did not seem to be much of an issue in the
communities we looked at; two reported a decrease and one reported an increase, with the
remainder either not knowledgeable or not aware of any change having taken place. So-
called white collar crimes such as forgery and credit card theft were reported to be on the
rise in seven of nine communities, including the two towns that reported an overall
decrease in crime; some respondents attributed this change to increased gambling in the
casinos, citing the timing of the increase or the circumstances given in particular news
reports they had seen.
Domestic violence also appeared to be a theme, with respondents in six commumhes
telling us they have seen increases in this behavior. Respondents reported, on the whole,
no increase in child abuse in their communities. However, six communities had one or
more respondents who said they had seen increases in child neglect, and attributed this
increase at least in part to parents leaving their children alone at home or in casino
lobbies and parking lots while they went to gamble.
.
A number of social service staff across several communities mentioned they have seen an
overall increase in "family stress" due to gambling. In one community, a mental health
specialist attributed a recent increase in divorces to casino gambling. Seven communities
reported either an increase in suicide since the casinos opened, or having seen cases
where people ended their lives due to problems stemming from their gambling. In
addition to the need for more gambling treatment, residents of four communities told of
an increased need for general mental health services for gamblers and their families.
Respondents in four communities noted concern over the increased numbers of older and
retired persons gambling in the casinos, and respondents in two communities noted an
increase in youth gambling. (As noted in Section 1, data from the national survey do not
indicate that particularly large numbers of young adults or elderly are gambling; in fact,
past-year gambling statistics for persons between the ages of 18 and 24 and 65 and older
show that these two groups gamble disproportionately less than one would expect, given
their proportion of the adult population. See Section 4 for a discussion of gambling
among 16- and 17-year-olds.)
.
Prob'em gambling
In seven of the nine communities we investigated, respondents told us they have seen an
increase in the number of people who are problem and pathological gamblers. Aside
from the fact that simply more people are gambling, one interviewee pointed out that
once a community builds a casino, the area may become more attractive to persons who
have the potential to develop into problem gamblers. Furthermore, as the casinos are still
a relative novelty in the communities we investigated, it may be that in time, the numbers
of people who fmd themselves having difficulty controlling their gambling behavior will
learn ways to keep their behavior in check, or will simply lose interest when another
novel and exciting activity comes along.
Gambling Impact and Behavior Study Page 78
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
......
UNI'/EASJr,. OF CHICAGO
While the national prevalence data from the 1998 survey versus its 1975 counterpart
indicate an increase in lifetime gambling in the United States, the data do not indicate an
increase in the proportion of people who have gambled in the past year. Therefore, most
of the increase in gambling appears to be from increased activity among those in the
population who gamble regularly. From the perspective of our interviewees in at least
four communities, casino gambling is more habitual than previously available gaming
opportunities, so those who gamble do so more frequently and intensively. In fact, it is
clear that casinos are "more popular" than other forms of legal gambling in that people, as a
group, spend much more money at them than on other games, other things being equal.
People who live within 10 miles of a casino typically spend $400-600 per adult per year on
casino gambling, versus $50-100 per adult per year on state lotteries (in states that have
them) and $10-30 per year on horse and/or greyhound racing (if they live within 10 miles of
a track).
Finally, a major theme that came through in every single case study is that substance
abuse is a major problem in these communities. Respondents complained about the high
proportion of drug-related crime, arrests for public intoxication and DWI, youth drug use,
and addiction in general. Many interviewees also drew a connection between substance
abuse and gambling, noting that a high proportion of people with one problem will also
have the other. It is plausible that casinos are affecting substance-using behavior and/or
the reverse. It should also be kept in mind that some of the changes reported in the
communities may potentially be due to substance abuse.
.
Public opinion regarding gambling
One community was strongly in favor of legalized gambling, six indicated a slight bias in
favor or a general acceptance of the casinos, two communities were very mixed, and one
was clearly negative. Despite a sample size of only 10 communities, these results are
reasonably consistent with how the general population characterizes gambling.
. w
. Case Study One: Florissant
The case study interviews were organized in sections, and for the most part, we found it
sensible to retain this organization when writing up our findings. We structure each case
study in the following way: First, we discuss the backgrounds of the interviewees who
participated in our survey. Second, we present a summary of the types of gaming that are
available to residents both within the community itself, as well as nearby, meaning within
a 50-mile radius of the city or town. Third, we present our findings on the types of
changes that have occurred in the community in the past 10 years. Although we stressed
to respondents that this question referred to general changes and not just those due to
gaming, most often respondents told us about the latter; we make this difference explicit
where needed in the text. Next interviewees were asked what they currently saw as the
most important socioeconomic issue in their community. Finally, we asked respondents
what their views were on how the community generally feels about having legalized
gaming readily available.
.
20 This name and all other proper names in these case studies are fictitious.
Gambling Impac1 and Behavior Study Page 79
Chapter 6. Case Studies of the Effect on Communities of InCl"easing Access to Gambling
.
NORC
11IM
UNiVEnS!TY OF CHfCAGO
Below we present two of our case studies, chosen in part for the very different impacts
gambling has had on these two communities. In Appendix C, we repeat this section, and
include full discussion of the remaining eight case studies.
Our respondents
for our case study of the Florissant community, we interviewed a detective in the
florissant police department, the director of a nonprofit organization specializing in child
abuse prevention, an addiction counselor, a city planner, the manager of a consumer
credit counseling agency, a tourism office commissioner, and the editor of a local
newspaper. Our respondents have been in their line of work for an average of 15 years
(ranging from about 7 years to more than 25). All have lived in the Florissant area or
nearby for at least 10 years.
Gaming in Florissant
When people talk about gaming in Florissant, they are uSlially talking about casinos.
Bingo games are run by nonprofits such as the War Veterans Association, so it appears
that people view bingo more as making a charitable contribution than as actual gambling.
We asked four respondents if they had noticed any trends that appeared to be inlluenced
by the opening or closing of gaming facilities, or by the availability of certain kinds of
games, and three of them told us that comparisons could not be made, since "We really
only have one type."
.
Around 1990, state residents voted to allow casino gambling, and within florissant, the
first casino opened its doors within a year of the vote. Residents also travel more to a
nearby state to play in the casinos there. Residents reported some turnover of casinos in
the area, with some of the initial, smaller operations leaving or being bought out by larger
firms. In recent years, several casinos near but not within florissant have closed or gone
bankrupt and changed owners. The two original casinos within the city limits are the
only casinos still operating there and no others have opened and closed in town in the
meantime; however, a new casino/hotel complex is expected to open soon.
Community changes
This city is one of the largest in the state; the opening of the casinos has turned what was
once a "ghost town" into a popular tourist destination. "[Gambling] is part of the
community. Gambling is an active part, it has been involved with community activities.
The casinos in [town] are corporate citizens." In this vein, another stated, "The reason
we brought [gaming] in was to use it as a leverage point to bring in other business.
Gaming is certainly different than most expect it, but the casinos are run by Fortune 500
companies, not by gangster types." In the midst of these changes, the city has been
striving to maintain a "clean" image. One interviewee reported that "[a] strip joint
opened up in downtown [florissant], but the city took them to court and they left town."
.
We asked individuals what community changes they have seen over the last 10 years and
whether they thought any of these changes was related to gaming. Everyone mentioned
the population boom or some change that was related to the increased number of people
moving to the area, such as the lack of housing or the traffic and demand for new roads.
The thriving economy appears to be the main reason for this influx. People without jobs
or many resources come to the Florissant area to start a new life.
Gambling Impact and Behavior Study Page 80
Chapter 6. Case Studies of the Effeci on Communities of Increasing Access to Gambling
.
NORC
......
l)NlVEA-S!ri OF CHIC,\GO
Due to the casinos, according to one interviewee, "Investment into schools and public
services has been greater. Also, there has been more investment in highways and
sanitation services.... More people are working; there are more two-income families than
ever before. The MSA now has more businesses than any other region. . ."
Gaming, according to the newspaper editor we interviewed, comprises one-quarter to
one-half of the local economy, and all respondents except one indicated that they were
pleased with the overall direction of the local economy (the exception abstained). The
thread that connects most of these comments is the dramatic improvement in employment
opportunities. Since so many people mentioned this, we break out their comments below.
.
· According to one respondent who is involved in transportation planning, "We've had
to build roads faster, increase housing stock, service businesses.... About 60 percent
of these changes are due to gaming."
· A cn;dit counselor stated, "We see mostly people who have been out of work in other
areas moving here, then paying off creditors...."
· A staff member on the tourist commission stated, "We've had significant increases in
employment. We have more two-income households, with more spending power.
There's been a [substantial] rise in per capita income since 1990. It's generated a
positive image about the area, and has attracted other sectors. The timing [of the
casinos being built] was very good.... It has impacted a large portion of the economy,
... but not all of it. [New kinds of businesses] have appeared, [and] the
manufacturing sector... has strengthened due to independent world conditions. [The
casinos have] increased exposure for tourism-related business, which we've seen in
data from [two local associations).
· According to another, "We've had a huge increase in employment and tax revenues.
It's a tight labor market, so salaries are increasing. People are moving to Florissant
from all over the country."
· A newspaper editor stated that the lower end jobs in the area now have benefits,
where before they did not.
· Another respondent stated, "The region's population has grown (by more than half]
in the past 10 years. We have more service-oriented jobs riow."
Respondents also mentioned what could be considered less desirable changes attributable
to the casinos, including the rise in problem gambling. According to a respondent in law
enforcement, "there are now Gamblers' Anonymous meetings in [Florissant] every single
night of the week, when there used to be none." An addiytion counselor stated that
problem gamblers were not known before the casinos. The detective stated, "We already
had alcohol and drug addiction services. The casinos opened up . . . and we saw our first
problem gamblers [within 18 months]. These people were often both alcoholics and
problem gamblers. They have free booze" in the casinos. Two respondents mentioned
an increase in bankruptcies. But help is advertised-"They have help-lines now, and
more people are aware that help is available." Another told us, "Casinos are proactive in
combating gambling problems and are bringing money into the community."
.
An interviewee in social services mentioned seeing an increase in neglect, though not
abuse, over the past several years; "We've seen children left unattended, people losing
Gambling Impact and Behavior Study Page 81
Chapter 6. Case Studies of the Effeci on Communities of Increasing Access to Gambling
.
NORC
11IM
UNiVEF?srr, OF CHICAGO
their money who can't afford to pay for their food and rent...." Another reported that
"[h]ousing prices have gone up. We have traffic congestion and crowded schools.
There's been a strain on infrastructure, construction is up. Not a huge rise in crime.
Some traditional neighborhoods have been stressed by growth-high-rise condos,
shopping centers, hotels appearing. All of [this change was due to gambling], though
some of it was indirect." One person indicated that chain restaurants were moving in and
"chasing away the local restaurants." A respondent summed up the two sides this way:
It's been both good and bad. Constmction is booming, but there are complaints
of illegal immigrants. The schools have more money, but there's also the
increase in Gamblers' Anonymous groups and bankruptcies. Property values
have gone up, but renters have been hurt, especially seniors. Average rent has
gone from $200 a month to $700a month.
We asked our informants whether they thought the nature or number of crimes in
florissant had changed. Everyone mentioned an increase in robberies. According to an
officer in law enforcement, crime in general has increased. He added that burglaries
account for the greatest proportion of crime in the city. Furthermore,
.
Robberies are up, and there has been a rash of bank robberies... that appear to
have been gambling related. Attempted suicides have also gone up. We have
had some illegal gambling activity in the area. There has also been a rise in
prostitution.... The casinos give away alcohol for free, and we have seen a big
rise in alcohol-related crimes. The casinos try to self-regulate, but they are
ineffective. In Vegas, people catch cabs, but in [Florissant], drunk driving is an
Issue.
On the other hand, according to a newspaper reporter, prostitution is now less visible than
it was before the casinos, and it should be noted that we were unable to find any escort
agencies in the Florissant Yellow Pages. The reporter also stated that "there have been
some isolated, weird crimes, from people freaking out after losing their money. One
woman faked her own kidnapping, then disappeared for a month."
Current community issues
We asked respondents what they thought was the primary social or economic problem
their community was facing right now, and whether they thought gambling had helped
the community to control this problem. One respondent was not sure, and did not
elaborate. Another interviewee, in social services, had only good things to say:
"Economically, we're doing much better, unemployment is extremely low. Socially, I
don't know what the primary problem would be." A respondent on the tourist board
complained about the transportation congestion and workforce availability, but added,
"Gaming is the cause but also employs [thousands] directly. Tax rates have actually gone
down as we have... extra capital there now." Similarly, the planner stated, "The cost of
housing is increasingly too rapidly, but gaming has increased salaries, too."
The law enforcement interviewee was less positive and pointed to an array of problems
he felt had all been created by the opening of casinos:
.
Pawnshops are now all over the place operating 24 hours a day, and they don't
worry about dealing in stolen merchandise. There are even places that will let
you sign over the title to your car. We've seen some organized crime. We had a
Gambling Impact and Behavior Study Page 82
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
~ -
UNfi/ERSJn' OF CHICAGO
.., [situation] where the FBI came in and arrested people.... Casinos are very-
image conscious. If there are rapes in the parking lot, the casinos ask us not to
report them. Casinos don't go after problem gamblers unless they exhibit signs
of substance abuse. Also, the increases in bankruptcy.
According to the addiction counselor, the primary problem is Florissant is "[d] rug abuse.
Crack has been around 1 0 or 15 years, and hit a high in the last 4 years." He added that
gambling does not appear to be related to the drug problem in Florissant. Finally, the
newspaper editor stated that the primary problem Florissant faces is "uncontrolled
growth. And gambling's causing it!"
Public views on gaming
Next we tried to get a feel for the climate in the community regarding gaming. People
did not seem to differ much on their perception that the community felt positively toward
the casinos, although some emphasized the fact that a minority exist who do not approve.
Since the community voted to allow casino gaming, it is not surprising that most
interviewees told us that the majority supports it. However, a few either expounded on
this or disagreed; we include their responses below.
.
· "There's no middle of the road. The community is divided. The town voted
gambling down before it passed. [Florissant] is a casino town now."
. "For the most part, it's understood to be an industry, albeit a recreation/leisure
industry. It's seen to be positive for our economy."
. "We had a vote, and it passed. It would depend on whom you talk to. It offends the
religious beliefs of some, but the economic benefits have been significant."
We also asked people what their expectations had been for when the casinos opened-
whether they thought that the community would change in any way. Only one person
mentioned having no expectations beforehand. Three people mentioned they believed
that the casinos would bring more tourism to the general area, and two stated that the
result far exceeded their expectations. The flip side to this was expressed by another
respondent, who stated, "My expectations were that gaming would cannibalize the area.
And it's happened." Another stated he had thought that the casinos "would contribute
more to treatment programs for compulsive gambling and drug addiction on the local and
state level, though [one casino] did contribute to an awareness program." Finally, a
respondent stated, "I expected a lot more crime and a lot more traffic. Traffic has
increased, crime has not really increased as much as I expected."
When it came to their own views on the impact the casinos have had, respondents either
reiterated their feelings that they have been positive for a majority of the community or
were even more glowing with their praise. In the words of one, "It's been very good.
Very positive. It's like having Disney World move to your community."
.
Gambling Impact and Behavior Study Page 83
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
/
NORC
......
UN;1.,'ERS/n' OF CHICAGO
Case Study Two: Hansen
Our respondents
For this case study, we interviewed the local chief of police, a mental health program
administrator in the local department of social services, a psychotherapist, a director of a
council on problem gambling, the local director of community development, a program
director at a consumer credit counseling service, an administrator in the local chamber of
commerce, and a reporter for a local newspaper. On the whole, respondents have lived in
the general area for a significant amount of time; with the exception of one respondent
(the administrator in the chamber of commerce, who has lived there for just over 5 years);
interviewees have lived in the vicinity for a minimum of20 years.
Gaming in Hansen
We asked respondents about the kinds of gambling available in their area. We found it
significant that we received such an inconsistent set of responses, and we repeat them
below:
· "We have lottery games at the gas stations. None, otherwise."
.
· "There are three popular bingo halls."
· "Every kind you can imagine, from legal to illegal.... Charity gaming, illegal sports
books, illegal video poker machines in bars."
· "Zero."
. "The state lottery."
· "Absolutely none. [Hansen] was started to be a religious community. It is still dry,
and originally no doctors were even allowed."
· "None that I am aware of, outside of the lottery."
Respondents also told us about locations outside the immediate community, but within a
50-mile radius, that attract local residents. These include a handful of casinos all close
by, which opened around 1990; "numerous bingo halls," including a tribally owned
parlor; and a dog track. According to one respondent, the dog track has been operating
for more than a generation. One interviewee stated that the municipal government
sponsors casino junkets for local senior citizens to a casino about an hour or so from
town, but he did not know how popular these excursions were.
When asked if any facilities have closed, the majority of the respondents mentioned a
horse track that was a short drive from town; one interviewee added that after the track
closed, a more local OTB parlor closed. Another respondent stated, "None [have closed],
because there haven't been any."
.
The respondents in this case study were unique in that they interpreted our use of the
phrase "legalized gambling" with the understanding that this category includes a variety
of types of games. In every other community in which we conducted interviews,
respondents appeared to equate "legalized gambling" with casino gaming. In this
particular town, one interviewee told us that people tend to spend more money at bingo
Gombling Impact and Behavior Study Page 84
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
""""
UN/l,'ERs!rY OFO;:C.~GO
halls than at the casinos, but added that the share taken by casinos is growing. Another
respondent commented that' she has seen people in town spending "a disproportionate
amount of their meager incomes" on lottery tickets, partly because it is "just so
convenient to play." A credit counselor told us that about one-half of the people she
works with play the lottery regularly, and guessed that they spend about ten dollars per
week on tickets. Another resident commented on the popularity of the lottery as well,
adding "We occasionally see thefts of lottery tickets from convenience stores, or a
domestic dispute over a winning lottery ticket."
Community changes
Hansen is small town that has recently experienced some economic troubles. Several
respondents told us that a particularly large plant recently closed in the area, which
affected the tax base significantly. According to an administrator in the local chamber of
commerce, this shutdown "had a major impact on our community. [Hundreds of] jobs
were lost." However, the newspaper reporter indicated that most of the plant's
employees were from out of town. The real impact seems to have been in revenue loss;
three respondents told us that this shutdown has impacted as much as half of the tax base.
The chief of police added that the schools have been particularly hard hit.
.
The director of community development defined the town as a "low-to-moderate income
community, primarily residential in nature. Like many older, inner-suburban
communities, we are struggling to provide services, meet demands." Another issue the
town is trying to handle is that people are moving out to newer suburbs. "We face a lack
of local jobs, and our taxes are based on residential property." Another respondent told
us that the town has a high percentage of rental housing per capita, and that they are
trying to take advantage of a local housing boom in the county to improve the housing
situation.
We asked respondents if they personally were pleased with the overall direction of their
economy at this point. Two respondents indicated they were, while four respondents
indicated dissatisfaction, including
. One interviewee who cited the recent closing of a large plant;
. A program administrator in mental health, who told us that "medical services are
outrageous in our community. If they're going to legalize gambling, they should put
money into treatment, across all areas";
. The director of community development, who stated the community needs "a broader
economic base, more industrial and business development in the community"; and
. A newspaper reporter, who commented that "[Hansen] really needs to wean itself off
of the tax money [from the power plant]."
One respondent declined to answer.
.
The director of community development told us, "We are part of [a] metro region, and
are influenced by [the city].... We constantly face movement to the green fields, new
suburbs. Urban sprawl contributes to our problems and is part of our planning process.
We face a lack of local jobs, and our taxes are based on residential property." Aside from
the plant closing, however, "The community has not changed much since 1990.... We
Gambling Impact and Behavior Study Page 85
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
-
UNNERS~rr" OF CHICAGO
have an ongoing economic development program, though. We have no specific sectors
targeted, but would like to attract light manufacturing jobs. We're too far from the
interstate to attract warehouse and distribution business."
According to another interviewee, a couple of new restaurants and banks have opened
over the past several years, but overall "There's not a huge commercial base.... Basically,
service industries are all that's moving in.... It tends to be a strictly residential
community." The credit counselor commented on the job situation, saying, "We have
low unemployment, but they're all low-paying, service-industry jobs."
None of our informants indicated that the local gaming establishments have had any
affect on the local economy, not even on tourist-related businesses. People from out of
town visit the community to go to the park or to walk along the marina, and the toWn has
an excellent medical center. In the past 2 years, the chamber of commerce representative
said, only one person has asked her about gambling in the area.
.
On the other hand, some controversy exists surrounding whether the opening of casinos
nearby has had an affect on individuals in the town. The program director for a consumer
credit counselor service said that the town has "seen higher and higher credit card debt....
The number of individuals filing for both Chapter 7 and Chapter 13 bankruptcies [i.e.,
personal bankruptcy] has increased." She told us she did not see any connection between
these trends and gambling, saying, "gambling losses play a minor role in the debt
problems that we are now seeing." But she added, "Casino gambling has increased in the
area. A few families have come in with gambling problems." According to a mental
health administrator, the town's relatively high poverty rate has made residents more
vulnerable to potentially addictive behaviors. "People think that they can get rich [if they
gamble], and the people that we treat tend to be on the lower end of the income scale in
[the] county.... The economic impact creates stress, which creates substance abuse and
other negative impacts."
With regard to changes in social services in the community, the mental health program
administrator told us she has seen "extreme changes" in her agency, and the demand for
addiction services far exceeds their resources. Not-for-profit companies in particular
have been suffering. "It's been a steady increase, maybe there was a surge since the early
1990s. The courts have adopted an interventionist approach as opposed to a punitive one.
More people are getting court-ordered treatment instead of jail time. [However,] we're
limited in the number of sessions we can provide and the number of people that we can
treat because of fmancial concerns."
Since the agency is not set up to deal with gambling addiction, and none of the
counselors are certified gambling counselors, staff only deal with gambling indirectly.
The administrator added that no services exist in the immediate area specifically to help
people with gambling problems, but she said that gambling treatment centers have been
opening in the state. She told us that her agency now refers more people to gambling
treatment than they used to. "I'm sure there's an indirect effect with drinking and other
kinds of drug abuse. I think it's a cumulative effect. Gambling, drinking and drugs
increase domestic violence and poverty. Can I track that as a statistic? No. That's more
my opinion than anything I could prove."
.
The director of the council on problem gambling also spoke to this issue for the state as a
whole:
Gambling Impact and Behavior Study Page 86
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
!!1m
lJf'Jii/ERs!n. OF CHICAGO
There's a tremendous need for treatment of gambling problems. There is better
awareness now, people show up stating that they and their families have
gambling problems. . Taxpayers also spend much more for prosecution and
incarceration of compulsive gamblers. This is often not factored into the
equation.... We have seen a doubling, at least, in the number of Gamblers
Anonymous groups in the state ... over the past 5 to 6 years. The need for
services has expanded dramatically.... We've trained [hundreds of)
professionals in diagnosis, assessment, and treatment over the last 3 to 4 years
statewide.
We also asked respondents about ways in which crime and law enforcement may have
changed in their communities over the last decade or so. According to the chief of police,
all types of crime are prevalent in this community, and most of it is related to drugs and
gangs. In particular, sales of illegal drugs "are way up." The newspaper reporter told us,
"Substance abuse is still a big problem.... Crack-cocaine [for instance], but it's
declining. Drug raids have tapered off, but are very severe when they occur. The
improved economy also helps, though it's really not better in [Hansen]." TIW local police
department has responded to this problem by becoming more involved in community
policing: "The districts haven't changed, but officers are permanently assigned to areas
now, and we used to rotate them."
.
The chief of police faxed us a brief table displaying major offenses (e.g., homicide,
burglary, arson) in the town from 1994 to 1997. These data show a downward trend in
violent offenses, and particularly in robberies and property offenses, although it we have
not calculated whether these changes are statistically significant. The chief of police told
us that although specific types of crime have increased, these trends are not connected to
gambling. In addition, he reported that illegal gambling has not increased since the
casinos have opened, and a program administrator in mental health told us she is not
aware of any trends in underage gambling.
When we asked the director of the council on problem gambling whether he thought that
crime has increased, he did not speak to the Hansen community in particular, but did
speak to the state as a whole:
Certainly. We have noticed a statewide increase in paper crime-forgery, fraud,
theft and embezzlement-none of which are classified by law enforcement
officials as gambling-related crime. Criminal activities of compulsive gamblers
are often not recorded. The crime question is being asked incorrectly, and the
results could be completely misleading.
The newspaper reporter told us that, overall; c~ime in the area has gone down. However,
he added that "transient residents" are susceptible to gang activity, and .that the town
mostly has drug-related crime like robberies and vandalism. "I don't think the police
force would want gambling in the area. Crime is already somewhat high for a
community our size."
Adolescent suicide rates may also be on an upward trend. The mental health
administrator told us, "I think teenage suicide is up slightly in this community, though I
don't know what the statistics are county-wide. I'm not aware of any connection
between suicide trends and gambling, but I'm not saying that they don't exist."
.
Gambling Impact and Behavior Study Page 87
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
""""
UNii:EAS!fr OFC.'i:CAGO
Current community issues
We asked respondents to tell us what they thought was the primary social or economic
problem their community faces at this time, and whether they believe legalized gambling
has had any affect on the community's efforts to control this problem. One interviewee
told us he did not know. Two respondents mentioned drugs, one also mentioning gangs,
and the other eXplicitly including alcohol. The latter respondent stated, "Crack cocaine is
far too accessible, far too cheap. Our treatment of crack-cocaine users is at an all-time
high. [Legalized gambling] certainly has a negative impact." Two respondents
mentioned the tax base, one calling it "unstable," and the other saying "We have a serious
crisis." Neither could say that gambling has helped or worsened this problem. One
interviewee mentioned the lack of high-paying jobs, adding that she was not aware of any
change in this area due to the opening or closing of gambling facilities. Finally, one
respondent told us he saw the primary problem as "pathological gambling.... Legalized
gambling has brought the problem more to the surface, but at the same time, has not
made it easier to deal with. In [this state], there's absolutely no state funding to address
compulsive gambling problems. The casinos also kick in no money for treatment
programs. "
.
Public views on gaming
Next we asked respondents how they thought the community generally feels about having
legalized gaming readily available. We found the nature of our responses to depart rather
radically from the kinds of responses we were given in other case study communities.
Basically, the respondents in this town were split into two camps: Interviewees who felt
that overall, the community was pleased with having gambling available due to their
being uninformed about the negative impacts, and interviewees who felt that the majority
of the community was opposed to gambling for religious reasons. A sampling of
responses follow:
· "I think they're ecstatic. They think that more money's going into education, that
they won't have to pay higher taxes. They don't grasp the negative impact. They are
used to segregating off crime and substance abuse in a societal kind of way."
· "I think most of the community likes it, and it's based on ignorance. It's not even the
gambling that the community likes, it's the fantasy of winning."
· "[Hansen] was originally formed as a religious community. Most residents are not
interested in having gambling." .
· "I'm sure it would fail miserably in this town. They tried to get alcohol on the ballot
in the November election, and it failed to pass."
Next we asked interviewees what their personal expectations were for the effects of a
nearby gaming establishment on'the community before the casinos opened, and whether
their expectations have been borne out. Most said they did not have expectations, and so
had no surprises when the casinos opened. The two respondents who did have
expectations expressed them in the following way:
.
. "My expectations were that it would be a net loss to the community, and they have
been borne out. I see things that others don't see. For every winner, there's a loser.
It's easy to spot fire plugs painted, streets paved, buildings built... [Also, I have been
Gambling Impact and Behavior Study Page 88
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
.
NORC
-
UNfVERsrrr' OF CHICAGO
surprised] that the public would continue to be as blind as they are to the loss side of
the equation."
· "I would expect it to be detrimental. In [other communities in the state that have
opened casinos], jobs were created, but the economic runoff that they expected has
not been realized."
Finally, we asked the respondents what their overall feelings were about the effects
legalized gambling has had on their communities. We repeat their comments below (two
respondents were not asked this question):
· "It has social and economic negative impacts, and contributes to addiction. I don't
support it, personally."
. "Negative."
· "I can't say that there's any [effects] that I know of."
· "I'm not aware of any community effects, but gambling can be devastating for an
individual family."
· "I've never even bought a lottery ticket, but don't see a big problem with gambling
personally.... I'm sure there's an impact, probably slight."
.
References
Bulkeley, W. July 14, 1992. "Video Betting, Called 'Crack of Gambling,' is
Spreading," The Wall Street Journal B1.
Cox, S., Lesieur, H., Rosenthal, R., and Volberg, R. 1997. Problem and
Pathological Gambling in America: The National Picture. New York: National
Council on Problem Gambling.
Eadington, W.R. 1991. "Public Policy Considerations and Challenges and the
Spread of Commercial Gambling." In Gambling and Public Policy:
International Perspectives. W.R. Eadington and J.A. Cornelius (eds). Reno,
NY: University of Nevada Press.
Harrah's Entertainment, Inc. 1996. Harrah's Survey of Casino Entertainment.
Memphis, TN: Harrah's Marketing Communications.
Heubusch, K. May 1997. "Taking Chances on Casinos." American Demographics.
h tip:/ /www.demographics.com/publications/ad/97ad/9705ad/ad970530.htm
IGWB (International Gaming & Wagering Business) Online. August 15, 1997.
"1996 Gross Annual Wager: Growth in Gaming Industry Slowed Last Year in
Both Handle and Revenue." http://www.igwb.com/news/gaw.html. accessed
February 4, 1998.
McQueen, P.A. 1996. "North American Gaming at a Glance." International
Gaming & Wagering Business 17(9): 52-56.
Scripps Howard News Service and Ohio University. "Recent Bush Center Research
Projects: Gambling." http://www.scripps.ohiou.eduJresearch.htm#gamb Ie,
accessed February 6, 1998.
.
Gambling Impact and Behavior Study Page 89
Chapter 6. Case Studies of the Effect on Communities of Increasing Access to Gambling
TAB
"6"
.(
.
.
.
CUTTING THE CARDS AND CRAPS:
Right Thinking About Gambling Economics
Earl 1. Grinols*
21 December 2001
For most of the twentieth century the American gambling industry could be described as
no more than a footnote to the na.tional scene. Attention to the industry, when it occurred
at all, was regional or episodic. Only after the advent of the 1990s' spread of casinos from
14 counties, all of them in Nevada except one, to hundreds of counties in dozens of states
did the industry become nationally significant and the question of whether casinos were
desirable a serious policy matter. This paper deals with the notional issues surrounding the
social value of gambling. The emphasis is on correcting common misconceptions about the
meaning of economic development and applying logic to the valid cost-benefit evaluation of
casino gambling. .
I. Historical Context
During its rise to national attention in the early 1990s, the gambling industry and its spokes-
men espoused the view that opposition to the spread of modern gambling should be puzzling
to any sensible observer and perhaps could be understood only as the emotional outpourings
of moralists or religious zealots. It was not to their advantage to acknowledge the possibility
that opposition was the reasonable resporrse to social costs that exceeded social benefits.
Those in the news media at first took their lead frorn the industry. To provide balance
to their news coverage, they unenthusiastically but dutifully included in their reports the
misgivings of gambling opponents. Questioning why gambling might be opposed gradually
changed, however, as newsworthy and often spectacular examples of suicides (such as losers
throwing themselves from the tops of tall casinos), bankruptcies, embezzlements, murders,
robberies, government corruption, and other crimes began to surface. Reporters needed
less education about the harmful effects of gambling when they knew a family member or
acquaintance whose life was seriously injured by garnbling. Interest in the media and in
.University of Illinois, former Senior Economist, Council of Economic Advisers.
1
.
.
.
government grew about understanding the reasons why gambling might differ from other
forms of entertainment. Three questions of the time were:
1. Did the gambling industry deliver on its often extravagant promises of economic de-
velopment?
2. What were the social costs that opponents seemed to assert so vehemently?
3. How did problem and pathological gamblers enter into the evaluation of gambling?
The answers to these questions required serious research, of which there was little. Mean-
ingful academic interest in gambling did not exist in most parts of the country because
commercial casino gambling had been criminalized for most of the century, often specifically
prohibited in state constitutions. A handful of respected scholars outside Nevada and New
Jersey had paid some attention to lotteries which were more widespread by the 1980s, but
answers to most questions were unavailable.
Research from states with the most experience with casinos-New Jersey and Nevada-
provided little that could be used in the social evaluation of gambling. Even though casinos
had been operating in Nevada since 1931, research from Nevada had not identified, let
alone quantified, the social costs of casino gambling. No measures of the benefits of casinos
had been estimated and the effect of casinos on causing problem and pathological gamblers
was unknown. In fact, the gambling industry initially denied that the spread of casinos
affected the number of problem and pathological gamblers until the June 1999 publication
of the research from the National Gambling Impact Study Commission made that position
untenable. The industry's public position was that people would gamble regardless of the
access to casinos and so access did not matter. Nevada experts did not know what the
prevalence rate of pathological gambling was in Nevada because they did not study it.
Pioneering research by Politzer, Morrow and Leavey (1981) at Johns Hopkins University
did quantify large costs associated with pathological gamblers. For a number of years this
research was left undisturbed and unchallenged by the gambling industry. The industry only
began to understand its significance when its implications about the damage that gambling
did to society were pointed out by researchers such as Kindt (1994a,b, 1995, 1998). There-
after, the American Gaming Association endeavored to harm the reputation of this work
by questioning its methodology. 1 The industry did not conduct research of its own on this
topic.
As data about problem and pathological (P&P) gamblers became more widely available,
often from the studies of treatment specialists, it suggested that P&Ps lost far more in
casinos than the average adult. Grinols and Omorov (1996) were among the first to raise
this issue, but other research soon followed. Combining loss figures with prevalence numbers
for problem and pathological gamblers implied that this tiny group of individuals (problem
lSee Farhenkopf press conference, 1998.
2
.
.
.
and pathological gamblers generally constitute fewer than 4.5 percent of the population2)
was responsible for a large share of casino revenues, as much as one third, or even in some
studies one half.3 Not surprisingly, research in Nevada had not investigated what share of
casino revenues came from problem and pathological gamblers, nor questioned what share
of revenues came from the small portion of the population that gambled most. One Nevada
academic testified that he had been involved in gambling research for 25 years, but that it
was "surprising how little we definitely know about the social and economic impacts of this
$40 billion a year industry." 4
By the late 1990s a growing number of scholars (Kindt, Henrikkson, Gazel, Goodman,
Grinols, Lesieur, Mustard, Omorov, Quinn, Rickman, Thompson, and others) had begun to
ask tougher questions about problem and pathological gambling, its social costs, how much of
casino industry revenues depended on the losses of problem and pathological gamblers, and
whether the economic development claims of the gambling industry were credible. Whereas
in the early 1990s there were almost no studies with firm numbers, by the end of the decade
there were, and research continues.
II. The Issues
The public debate over gambling has matured. We now understand, for example, that
gambling creates certain social costs that other entertainment industries do not. These
costs, have direct impact on other businesses and households, and do not operate through
prices and markets. They therefore fall into the classification that economists call negative
externalities. The social costs of gambling can be identified, in principle, and compared to
its social benefits.
In spite of the improvement of our quantitative knowledge about the effects of gambling,
however, misconceptions persist. Gambling opponents emphasize social costs that outweigh
social benefits; proponents "increased" number of jobs and "economic development." It
is not uncommon for industy-financed studies to report the effect of a casino on jobs, yet
never compute the value to residents of these jobs or incorporate them into a cost-benefit
assessment of casinos.
The situation is often not much more enlightened in government circles. Legislators still
believe against logic that taxes raised from casinos are "free" because they are "voluntary."
They commission reports to estimate how many tax dollars a casino project may provide, but
never incorporate these numbers in an applicable cost-benefit study. They ask for information
about jobs and taxes, but not about social welfare. What they hear of costs often must come
2Schaffer, Hall, Vanderbilt (1997)
3Lesieur (1998b), Productivity Commission (1999), Volberg, Gerstein, Christiansen (2001).
4Testimony before the Senate Committee on Governmental Affairs on the bill to create the National
Gambling Impact Study Commission (Committee on Governmental Affairs), 1995, pp.156-7, 162.
3
.
.
.
from the citizen groups that appear before their committees.
An outside observer of the way in which casino expansion has been done can rightfully
be excused for being appalled at the lack of rigor. There is not even agreement on what
constitutes a review. Industry apologists sell casino projects by counting jobs, government
evaluates projects by looking at tax revenues, and the citizens split according to their status:
Those who personally benefit-the casino owners and those with a direct stake in selling
casino games to the public or who think their taxes will be paid for them by the casino-are
in favor. Those who are indifferent to gambling and ignorant of its social costs are neutral
to mildly in favor since they believe they have no reason to prevent others their freedom of
choice. Few are knowledgeable enough to understand what constitutes a theoretically valid
cost-benefit review and sufficiently unprejudiced to pursue it.
III.
Jobs are Not A Benefit
The most common misconception about the social evaluation of gambling is the belief that
counting jobs created by casinos is the way to measure benefits. Jobs are not a benefit,
and more jobs in an area may even be harmful to existing residents. In other cases more
jobs may be the means to benefits. In this section we discuss the connections between job
creation and correctly counted costs and benefits. The concepts discussed are general; they
apply to the evaluation of other businesses as well as casinos.
A. Economic Development and The Creation of Wealth
A faulty emphasis on jobs derives from an incomplete understanding of economic develop-
ment. To answer the question, "What is economic development?" recall what the creation
of wealth is. When individuals undertake productive activity, they engage in the production
of goods and services that provide greater welfare or satisfaction than the inputs used. Paint
and canvas, for example, are re-arranged in the hands of a master painter into a work of
art that has more value than the components had previously. A barber provides grooming
that is more valuable to the customer than the value to the barber of the time given up.
Consumer utility rises when activities are pursued and productive assets used in better ways.
If society ca~re-organize so that someone is made better off without harming anyone in the
process, the welfare of society is enhanced. Money and prices provide a convenient measure
of the value of productive activity. When paint and canvas worth $100 dollars become a
painting worth $10,000 a month later, the painter has generated income of $9,900 over that
month.
Income is a flow per unit time, while wealth is a stock. Wealth is the claim to something
of value. Income that has not been used for consumption can be used to acquire wealth.
Claims to some assets have value to buyers because of expectations about how the assets will
4
.
.
.
produce a flow of income in the future. When these expectations change, the value of the
assets may change up or down, leading to capital gains or losses. With proper accounting for
capital gains and losses, wealth is the accumulation of past unused (unconsumed) income.
Economic development is the creation of greater value by society. Greater value can
result from a greater quantity of activity, activity that is more valuable, or both. Economic
development means greater income and wealth, which lead to greater utility for members of
society. The following constructed examples show that job creation is neither necessary nor
sufficient for economic development.
In the first example, let casinos be introduced into a community of 10,000 employed
individuals. The casino causes the demise of a nearby racet~ack and .takes business from
neighboring restaurants. The casino hires 100 employees, but these are matched by the loss
of 100 jobs at other businesses, and casino revenues are matched by reduced revenues at
these other businesses. The net effect is the enlargement of the casino sector matched by an
equal shrinkage of the rest of the economy. In this example, no economic development takes
place because greater value is not created. Casinos in this example act much like another
restaurant might in a town with many existing restaurants. The new entrant takes business
from others, shifting the location of activity but not increasing it.
From the first example one might conjecture that hiring by a new business need not
indicate an increase in jobs and that the failure to create jobs means there is no economic
development. It is true that hiring by a firm may not represent job creation. However, the
next two examples show that economic development is not linked to job creation: Develop-
ment may occur with or without a net increase in jobs.
In the second example, assume that casinos attract clientele from the surrounding areas.
Local residents do not gamble. Casinos hire 100 new employees who move in from sur-
rounding regions, paying them out of casino revenues earned from outside clientele. In this
example, the noncasino local economy continues to employ 10,000 people, whose earnings
are unchanged and who buy and sell as before at unchanged prices.
In this case regional net new jobs are created, but there is no economic development. The
local economy is enlarged and employment rises 10 percent, but this enlargement provides no
benefits to residents. Casinos operate like a toll house that uses the town as a platform for
conducting its business. Money enters and money leaves. While total economic activity in
the vicinity of the town rises, shifting the location of jobs without an increase in well being is
not economic development, even though the local economy experiences enlargement. Viewed
from the national perspective, this example could be consistent with casinos operating like
restaurants of the first example: Increased employment in the casino sector is matched by
reduced employment in other sectors. It is generally accepted, for example, that employment
in Nevada is larger at the expense of lower employment in California, from where many Las
Vegas clients come.
The third example shows that economic development can occur without job creation.
Assume that casinos begin operation, hire 100 employees, but cause the number of jobs in
5
.
other sectors to shrink to 9,900. Because gambling is so desirable to residents and outsiders
alike, the casinos earn higher profits than other businesses in town and bid up the prevailing
wage rate. Housing prices also rise due to the willingness of casino owners to pay higher
prices. Residents are better off because they can gamble nearer to home than before and this
provides them greater welfare, they receive higher wages for their labor which also makes
them better off, and their housing experiences capital gains. Because the work of the 100
casino sector employees generates greater profit than the profit lost from the businesses whose
employment shrank, the total of wages and profits of all area business is higher. Casinos
have brought economic development without new jobs.
In the first example, casinos did not increase jobs or provide economic development. In
the second, casinos provided jobs but no economic development, while in the third, casinos
provided economic development but no increase in jobs. What principle explains the different
outcomes? Anything that increases the value of social product leading to greater welfare
of residents is economic development and anything that does not is not. Job creation is
immaterial unless it happens to be a vehicle for increase in social value.
The increase in social value gets distributed to society in different forms. It could appear
as direct consumer benefits to households, such as closer access to an activity that was already
available. It could appear in the form of better prices for consumers. We will call the former
benefit distance consumer surplus, the latter, price consumer surplus. For example, a new
firm that is located closer to the consumer provides distance consumer surplus. A firm
whose entry into the market causes the price for its product to fall provides consumers with
greater price consumer surplus. A new firm might cause wage rates to rise, also providing
households with greater price consumer surplus, even though the consumer acts in the role of
labor supplier. Economic development could appear in other forms as well, including higher
profits to firm owners, greater taxes collected by government, and even improved functioning
of markets. For example, the introduction of internet auctions such as eBay links markets,
giving buyers lower prices and sellers higher prices for their goods. This also creates welfare.
Nowhere in the description of economic development was the creation of jobs part of the
discussion. Jobs matter only in the sense that they proxy for one or more direct effects that
benefit residents. Additional jobs in an area might not be assm::iated with any benefits to
residents or could even be associated with negative benefits. The point is. that the benefit of
an additional job to residents depends on specifics of the region and the case at hand. Using
jobs as a proxy requires that the value of a job to residents be identified.
There is a close analogy between the construction of major league sports stadiums and
social costs of gambling. In the case of major league teams, a community is usually told that
a sports team will generate jobs, but the community must pay to build a sports stadium. In
the case of gambling, a community is told that a casino will generate jobs, but the community
must pay the social costs of casino gambling. Commenting on major league sports teams,
Rappaport and Wilkerson (2001) write,
.
.
6
.
.
.
The large public spending on sports facilities has been controversial. Usually
these costly projects are justified by claims that hosting a sports franchise spurs
local economic development by creating numerous new jobs and boosting local
tax revenue. However, independent econornic studies suggest that taxpayers may
not be getting such a good deal. 5
They go on,
the impact studies almost always fail to measure benefits in a form that can be
compared with public outlays....Correctly measuring the benefit from job creation
requires both accurately accounting for the net number of new jobs associated
with a team's presence along with valuing the benefit oj these jobs to the host
metro area.6 [Emphasis added.]
Based on their review of the available research literature, Rappaport and Wilkerson conclude
that the value of a job to the host area could be nothing at all! Or, it could be as much as
$1,500 per job created, depending on circumstances. Or, jobs could provide negative value to
the host area because housing is made less affordable, or traffic and other forms of congestion
increase. "For some existing residents, the net result may be that they are hurt rather than
helped by net job creation."7... "The ambiguity of whether a metro area benefits from net
job creation sharply contrasts with the common perception that local net job creation is a
benefit in and of itself."s
IV. What Counts and How to Count It Right
The previous section tackled the prevalent confusion about job creation and economic de-
velopment. In this section we provide a rigorously grounded formulation of the costs and
benefits associated with the introduction of an industry such as casino gambling that imposes
costs on society in the form of real-resource-using harmful externalities.
We base our approach to cost and benefits on the change in individual consumer utility,
u1-UO. Superscripts 0 and 1 identify pre- and post-change variables. Thus, uO refers to utility
"before" (e.g. before casinos) and u1 refers to utility "after." We assume that the individual's
utility is a continuous function of the consumption of goods and services. With respect to
casino gambling, utility also depends on the distance that the consumer has to travel to the
nearest casino, the number of visits made, and the amount gambled per visit. The number of
5 "What are the Benefits of Hosting a Major League Sports Franchise?" Economic Review, Federal Reserve
Bank of Kansas City, 86, 1, 55-86, p. 55.
6Ibid., p. 60-61
7Ibid., p. 64.
8lbid., footnote 5.
7
.
.
.
visits and the amount gambled are under the individual's control, but distance is taken as a
given feature of the consuming environment, as are prices. Consumption is quantified in the
vector uRn. The quantity of goods and services consumed appear as positive components of
x, while the provision of a good or service such as labor supplied by the consumer, is shown
by a negative component. Consumption is assumed to be locally nonsatiable. That is, for
any x there exists a another vector near to x that is strictly preferred by the consumer.
Define e(d,p, u) as the minimum expenditure needed to achieve utility u when distance
to the nearest casino is d and prices are p. For fixed d and p, e(d,p, u) rises and falls with
u. In other words, e(d,p, u1) > e(d, p, uO) if and only if u1 > uO. The welfare increase of
individual i is therefore, ei(di,p, un - ei(di,p, u?), measured in dollars,where the subscript i
reminds us which variables are specific to individual i. For all consumers welfare change is
~w = L [ei(dLpl,uf) - ei(d;,pl,u?)].
(1)
Equation (1) treats a dollar of utility gain to one household as equal to a dollar of utility
gain to another. This is a natural assumption and has a number of important implications9
including that firm profits are equally important regardless of which firm generates them,
and to whom they ultimately go. We assume that household i owns share ()ij of firm j, where
L.i ()ij = 1 .
The economy's available consumption goods come from one of three sources: endowments
D, current production Y, or international trade z. Endowments are goods inherited from
nature or the past. Without loss of generality, assume that households own the endowments,
Wi E R~, where L.i Wi = D, the economy endowment vector. The current production of firms
Yj adds to the economy total L.j Yj = y. An input, say labor time or the services of physical
capital, appear as elements of Yj with a negative sign, while outputs are positive elements
of Yj. International trade is a form of production where exports, entering z as negative
numbers, play the role of inputs, and imports, entering z as positive numbers, play the role
of outputs.
Government is also an agent in the economy. Government collects tax revenues and
spends them for the benefit of consumers. Although the accounting framework can be
made to accomodate as detailed treatment of government as needed, we make the standard
simplifying assumption that government returns to the household in lump-sum fashion tax
dollars used for the purchase of goods and services that enter household utility. Households
spend these transfers. The remaining government tax revenues pay for resources to deal with
costly externalities. For example, if casinos increase crime, government might need to hire
more police. If casinos create pathological gamblers, government might need to provide more
treatment counselors. Goods and services, g, used to counter externalities do not directly
affect household utility, but because they divert available resources from productive activity,
9See Grinols and Mustard (2001) for a discussion.
8
.
.
.
their removal represents a drain on the economy. Tangible resources used is the social cost
of externalities. Accounting in real terms requires
x+g=y+D+z
where x == 2:i Xi is aggregate consumption.
We assume that households are unconstrained in the long run in their purchase of goods
and services. This implies that pI . xI = ei (di, pi, uI) because an unconstrained consumer
selects the least costly bundle for attaining a given utility uI. In the initial situation, however,
we allow for the possibility that the consumer may be constrained. If constraints are present,
p? x? > ei(d?,p?,u?). The most prominent example of a constraint is unemployment. If
the consumer wants to work at the going wage, but is prevented, he is unemployed. We will
return to unemployment as it relates to the effect of casinos below.
The following identity rewrites (1) as a telescoping sum where each term cancels part of
the preceding term. It decomposes the change in welfare into interpretable components.
L [ei ( d~ , p~ , u~) - ei ( d~ , pi, u?) ]
L [p~ . x~ - p? . x?]
(2.1)
Income Effects
+ L [p? . x? - ei (rii, p?, u?) ]
(2.2)
Consumption Constraint Gains
+ L [ei( rii, p?, u?) - ei( dI, p?, u?)] (2.3)
Distance Consumer Surplus
+ L [ei(dI,p~,u~) -ei(dLpI,U~)]. (2.4)
Price Consumer Surplus
To explain (2.1), make use of the budget identity
Pi' Xi = L BijIIj + PO' Di + 7i - Ei
j
(3)
where IIj is the profit of firm j, pn . Di is earnings from the consumer's endowment, Ti is
household taxes, and Ei is the household's share of the cost of gambling-induced externality
expenditures. Summing (3) over consumers and differencing between the initial and final
9
.
situations breaks (2.1) into the following terms,
2:= [pi . xi - p? . x?]
2:= ~rrj
j
~ Profits
+ ~pn. [2
+
6.T
6.E.
(4)
.
6. Social
Costs
The change in firm profits, taxes, and social costs are self-explanatory. t:.pn. [2 is the value
to the consurner of induced price changes-increases or decreases in the price of something
the consumer owns such as housing-that affects consumer welfare. We characterize these
as capital gains because they represent pure price effects.
The next component, (2.2), is the amount of money the individual would be willing
to pay to remove the consumption constraints that are present, if any. If the constraint
is unemployment, this is the value of reducing unemployment, for example. Expression
(2.3) gives the value of having the nearest casino d} miles away compared to d? miles.
Assuming that the consumer values gambling, the income needed to maintain original utility,
ei(dLp?, u?), is smaller when d} < cf!. (2.3) represents the surplus accruing to the consumer
when the nearest casino is closer. It is proper to interpret (2.3) as the amount of money the
consumer would be willing to pay each year to have the nearest casino closer. Expression
(2.4) is a conventional consumer surplus term. It is the amount of money that the consumer
would be willing to pay each year for the opportunity to trade at prices p; instead of p?
For example, if casinos cause the wage rate to rise, the consumer is better off by this price
change and would be willing to pay amount (2.4) to have the higher wage prevail.
Using equation (4) with equations (1) and (2) gives:
Capital Gains
t:. Taxes
t:. w = t:.Profits + t:. Taxes
+Distance Consumer Surplus + Price Consumer Surplus
+ Capital Gains + Consumption Constraint Gains
- t:.Social Costs.
(5)
Equation (5) is the form we want. It lists aH cost-benefit components, and, with equations
(1)-(4), specifies how each term should be constructed. Nothing is left out and there is no
double counting.
EconomiC development is synonymous with greater creation of social value. The decom-
position in (5) identifies the ways in which this increased value is captured by agents in the
economy. An increase in profits is the share of economic development that goes to the owners
of firms. Since profits are summed over all firms, it is the net profit increase that matters
to (5). The increase in taxes captures the portion of economic development that citizens
collectively capture through government. Taxes are used for the public good, so higher tax
collection summed over all payers, not just casinos, represent a public benefit. Better ac-
cess to casinos is captured in the distance consumer surplus term, while the possibility that
.
10
.
.
.
casinos cause prices to change in ways that favor residents is captured in the price consumer
surplus and capital gains terms.
If the economy exhibits consumption constraints, and casinos are the cause of their
elimination, then this also represents a benefit of casinos. It is theoretically possible that
introducing casinos into the economy might reduce unemployment. The difficulty in pro-
pounding such a claim in practice, however, is that unemployment is generally understood to
be a transitory phenomenon. The economy would eventually move to full employment with-
out casinos. Comparing the final position of the economy with casinos to the final position
of the economy without casinos would reveal no unemployment in either. Term (2.2) would
then be zero. If one believes that casinos caused the swifter return to full employment, then
during the period of transition this would be a temporary benefit of casinos. I am unaware
of this argument being made and documented in practice.
The social costs of casinos are the cost of resources removed from other productive uses
to deal with the social problems caused by casinos.
Nowhere in equation (5) do casino jobs or jobs created appear. Jobs matter only to
the extent that they might be inputs to the creation of some other benefit identified in (5).
Although the number of jobs sometimes is a good indicator of economic development, it is
not always: The value of an additional job to an area may be zero, as documented in the
literature on the job effects of major league sports teams, already discussed.
v. The Bottom Line
The primary purpose of this paper was to correct misconceptions surrounding the evaluation
of socially costly industries like gambling. A full scale treatment of the literature estimat-
ing the costs and benefits of gambling would require more space than can be devoted here.
Nevertheless, we would be remiss to close without summarizing the essential numbers in a
short precis. Details can be found in the growing body of research that addresses the costs,
benefits and social implications of gambling, including Politzer, Morrow, Leavey (1981),
Maryland Department of Health and Mental Hygiene (1990), Florida Executive Office of
the Governor (1994), Kindt (1994a, 1994b, 1995, 1998, 2001), Thompson, Gazel, Rickman
(1996), Lesieur (1998a, 1998b), Thompson, Gazel (1998), South Dakota Legislative Research
Council (1998-99), Australia Productivity Commission (1999), National Gambling Impact
Study Commission (1999), Ryan, Speyrer, Beal, Burckel, Cunningham, Scott, Wall, West-
phal (1999), Gerstein, Murphy, Toxe, Volberg, Harwood, Tucker, Christiansen, Cummings,
Sinclair (1999), Thompson, Quinn (1999), Grinols (1996, 1999), Grinols and Omorov (1996),
Grinols and Mustard (2001a, 2001b), and Gazel, Rickman, Thompson (2001) among others.
There are two main ways in which gambling social costs have been estimated. One is
by identifying the costs of problem and pathological gamblers and combining these numbers
with estimates of problem and pathological gambler prevalence in the general population
11
.
.
.
to infer the size of total costs. The second is through direct observations of the impact
of casinos on selected variables such as crime rates. The approach based on problem and
pathological gamblers is the more widely employed. Its advantage is that the required inputs
are more readily accessible through sampling and survey techniques. Its disadvantage is that
it understates social costs by failing to count costs that do not operate through problem and
pathological gamblers. .
A. Direct Crime Cost Estimates
Direct estimates of the crime costs due to casinos, for example, has led to numbers that
are similar to the crime cost estimates derived from the criminal activity of problem and
pathological gamblers but they tend to be a little higher. Many experts believe that the
primary cause of increased crime observed in counties with casinos is due to problem and
pathological gamblers. However, crimes could result in other ways, and tallying only crime
due to P&P gamblers would understate the true amount of crime due to casinos. To directly
estimate the amount of crime due to casinos, Grinols and Mustard (200la) compiled data
for every county in the Unites States for the twenty year period from 1977-96. Based on
statistical analsysis they reported that 8 percent of observed crime in casino counties was
due to the presence of the casino. They found that it generally takes 3 or 4 years after casino
introduction before crime rates begin to rise. The cost of this crime was $63 per adult per
year in casino counties. This is comparable to the results of a study by Thompson, Gazel,
Rickman (1996) conducted in the state of Wisconsin. They reported that crime costs per
additional problem gambler were $4,225, or 42 percent of the total costs found. On a per
capita basis crime cost $57 per adult. Both figures are higher than the $46 estimate used in
Table 1 below, derived from the crimes of problem and pathological gamblers alone.
B. Costs Estimated From Problem and Pathological Gamblers
Pathological gambling is a recognized impulse control disorder of the Diagnostic and Sta-
tistical Manual (DSM-IV) of the American Psychiatric Association. Pathological gamblers
experience repeated failures to resist the urge to gamble, lose control over their gambling,
personal lives and employment, rely on others to relieve a desperate financial situation caused
by gambling; commit illegal acts to finance gambling and engage in other characteristic be-
havior. Problem gamblers have similar problems, but to a lesser degree.
According to a 1990 Maryland Department of Health and Mental Hygiene survey 62
percent of problem gamblers in treatment had committed illegal acts as a result of their
gambling, 80 percent had committed civil offenses, and 23 percent were charged with criminal
offenses (Maryland, 1990). A similar study of nearly 400 members of Gamblers Anonymous
showed that 57 percent admitted stealing to finance their gambling. On average they stole
$135,000 each. Total theft was over $30 million (Lesieur, 1998b). The National Gambling
12
.
.
.
Impact Study Commission's final report issued in June 1999 reported that among those who
did not gamble (had not gambled in the past year) only 7 percent had ever been incarcerated.
In contrast, more than three times this number (21.4 percent) of individuals who had been
pathological gamblers at any point during their lifetime had been incarcerated.
By studying problem and pathological gamblers the resource burden to society of an
additional pathological or problem gambler (some studies lump the two groups together)
can be determined. There are at least nine distinct types of gambling social costs that have
been identified:
1. Crime (apprehension, adjudication, incarceration and police costs),
2. Business and employment costs (lost productivity on the job, lost employment
time, other costs to firms),
3. Bankruptcy (lawsuits and legal costs, bill collection costs),
4. Suicide,
5. Illness (costs associated with depression, stress-related illness, anxiety, cognitive dis-
tortions, cardio-vascular disorders, chronic or severe headaches among others),
6. Social service costs (treatment/therapy costs, welfare, food stamps, costs associated
with unemployment),
7. Government direct regulatory costs,
8. Family costs (costs associated with divorce, separation, spousal abuse, child neglect),
9. Abused dollars (resources acquired from family, friends, employers under false pre-
tenses) .
Averaging the results of eight studies conducted between 1994-99 that contain original
research on one or more of the nine social costs10 resulted in the following estimates of
the costs to society per additional pathological gamblerll: Crime-$3998, Business and em-
ployment costs-$3995, Bankruptcy-$316, Suicide-Not estimated, Illness-$700, Social Ser-
vice Costs-$631, Government regulatory costs-Not estimated, Family Costs-$111, Abused
dollars-$3834. The total was $13,586. The comparable totals for problem gamblers based
on two studies (Gerstein et al. (1999) and SD Legislative Research Council (1998-99)) are
lOFL Executive Office of the Governor (1994), Thompson et at. (1996), Thompson et at. (1998), 3D
Legislative Research Council (1998-99), Ryan et al. (1999), Gerstein et al. (1999), Thompson and Quinn
(1999)
llGrinols, Mustard (2001a)
13
.
.
Business and employment costs-$200, Social Service Costs-$712. No other categories of
costs are available. The total for those costs that were estimated was $912.
Based on a meta-study 'of others' field 'research, Shaffer et al. (1997), estimated that
pathological gamblers were 1.4 percent of the population and problem gamblers 2.8 percent.
95 percent confidence intervals for the percentage of pathological gamblers in the popula-
tion were 0.90-1.38 %, and for problem gamblers 1.95-3.65 %. For an average 100 adults,
these numbers imply social costs of $14,006-$22,077, depending whether one uses both lower
bounds or both higher bounds. On a per capita basis, the range is $140-221. For the nation
as a whole, social costs for 197.5 million adults would be $27.6-43.7 billion. This places gam-
bling among the handful of social problems such as drugs and alc?hol that impose substantial
costs on society, including on those who do not engage in the activity.
Table 1: Summary of Per Capita Casino Costs and Benefits
Fewer estimates of the benefits of casinos exist. Grinols (1999) reports distance consumer
surplus as between $34-50 per adult, depending on whether one adjusts for the demand of
problem and pathological gamblers or not. How the other benefits of casinos are distributed
depends on the market structure and the casino tax rate. In the case where entry is free to
any casino firm meeting the licensing standards, as in Nevada and some other places like the
14
.
.
.
Gulf Coast of Mississippi, casinos exert no monopoly power and profits, through competition,
are driven down to the ordinary business level. In the case where casinos are taxed at the
same rates as other business, the increase in casino taxes is matched by comparable drop-
in taxes collected from businesses competing for the same dollars. In this case, the benefits
of casinos in society accrue entirely to consumers in the form of distance surplus. The net
change in profits plus taxes measured over all business is zero. These gains are conceptually
easiest to identify and estimate. vVe report them in Table 1 using cost figures in the middle
of the estimated range. When casinos are granted regional monopoly licenses, as in the case
of states like Illinois, some of the benefits of casinos are captured by casinos in the form of
monopoly profits. The increase in profits of business reduces the benefits that end up in
the hands of consumers. The net increase in profits to all business are harder to estimate
for a number of reasons, including the fact that American Indian casinos are not required
to report their profits. Similarly, if government levies heavier taxes on casinos than other
businesses, some of the benefits of casinos are captured by government, at the expense of
benefits that would have accrued to the casinos and consumers directly.
Based on the costs and benefits in Table 1 the social costs outweigh benefits by a factor
of $190 to $34 or 5.6 to 1. An economy that includes casino gambling is worse off by $156 per
capita compared to the same economy where casino gambling is prohibited. These figures do
not yet include any capital gains or price-related benefits of casinos. Following the literature
on the benefits of major league sports teams, let us assume, therefore, that additional jobs
due to casinos are associated with $750 in capital gains and price-related benefits to residents.
$750 is the mid-point of the $0-1,500 range identified earlier. Would these benefits change
the unfavorable cost to benefit ratio?
The answer is probably not. To overcome the benefit-cost deficit in Table 1 through job
creation would require 21 new jobs for every 100 adults in the population. Since labor force
participation in the US is 67 percent, this means that casinos would have to cause the labor
force to increase by 31 percent. Such a large increase is extremely unlikely.
The conclusion is inescapable: Based on the numbers currently available, casino gambling
fails a theoretically grounded, theoretically valid cost-benefit analysis.
REFERENCES
Executive Office of the Governor. 1994. Casinos in Florida. Office of Planning and Budget-
ing: The Capitol, Tallahassee.
Frank Fahrenkopf, Press Conference, American Gaming Association, CEO and President,
at the National Gambling Impact Study Commission, Chicago, Illinois, May 21, 1998.
(Lobbying organization of the gambling industry. Although the national press was well
represented at the NGISC hearings, none of the press except the Las Vegas papers carried
15
.
.
.
the story or any of Frank Fahrenkopf's criticisms.)
Florida Department of Law Enforcement. 1994. The question of casinos in Florida: Increased
crime: Is it worth the gamble? State of Florida: Tallahassee, FL.
Florida Sheriffs Association. 1994. Casinos and crime: Is it worth the gamble? A summary
report and position paper. Florida Sheriffs Association: Tallahassee, FL.
Frank ML, Lester D, Wexler A. 1991. Suicidal behavior among members of gamblers anony-
mous. Journal of Gambling Studies. 7: 249-254.
Gerstein D, Murphy S, Toce M, Volberg R, Harwood H, Tucker A, Christiansen E, Cummings
W, Sinclair S. 1999. Gambling impact and behavior study: Report to the National
Gambling Impact Study Commission. National Gambling Impact Study Commission,
(1 April).
Gazel R, Rickman DS, Thompson WN. 2001. Casino Gambling and Crime: A Panel Study
of Wisconsin Counties. Managerial and Decision Economics. 22, 1-3, 65-76.
Goodman, R. 1994. Legalized Gambling as a Strategy for Economic Development. Northamp-
ton, MA: United States Gambling Study.
Grinols EL, Omorov J. 1996. Development or dreamfield delusions?: Assessing casino gam-
bling's costs and benefits. The Journal of Law and Commerce. 16, 1: 49-87.
Grinols EL. 1996. Incentives explain gambling's growth. Forum for Applied Research and
Public Policy 11, 2: 119-124.
Crinols EL. 1999. Distance effects in consumption. Review of Regional Studies. 29, 1: 63-76.
Grinols EL, Mustard DB. 2001a. Business Profitabity versus Social Profitability: Evaluating
Industries with Externalities-The Case of Casinos. Managerial and Decision Economics.
22, 1-3, 143-162.
Grinols EL, Mustard DB. 2001b. Measuring Industry Externalities: The Curious Case of
Casinos and Crime. University of Illinois, University of Georgia, 1-35.
Henriksson LE. 1996. Hardly a quick fix: Casino gambling in Canada. Canadian Public
Policy. XXII, 2: 116-128.
Kindt JW. 1994a. Increased crime and legalized gambling operations: The impact on the
socio-economics of business and government. Criminal Law Bulletin. 43: 538-539.
Kindt JW. 1994b. Increased crime and legalizing gambling operations: The impact on the
socio-economics of business and government. Criminal Law Bulletin. 30, 538-555.
Kindt JW. 1995. Legalized gambling activities as subsidized by taxpayers. Arkansas Law
Review. 48, 889-931.
16
.
.
.
Kindt JW. 1998. Statement to the National Gambling Impact Study Commission, US and
international concerns ove the socia-economic costs of legalized gambling: Greater than
the illegal drug problem? Chicago, Illinois, 21 May.
Kindt JW. 2001. The costs of addicted gamblers: Should the states initiate mega-lawsuits
similar to the tobacco cases? Managerial and Decision Economics. 22, 1-3, 17-64.
Legislative Research Council, S. Dakota. 1998. Economic and fiscal impacts of the South
Dakota gaming industry. December 8, 1-145.
Lesieur H. 1992. Compulsive gambling. Society. 29, 4: 43-50.
Lesieur H. 1995a. Testimony before the National Gambling Impact Study Commission,
Atlantic City, New Jersey, 22 January 1998, Institute of Problem Gambling.
Lesieur H. 1998b. Costs and treatment of pathological gambling. The Annals of the Amer-
ican Academy of Political and Social Science (Gambling: Socioeconomic Impacts and
Public Policy, Frey JH, special editor). 556: 153-171.
Louisiana Compulsive Gambling Study Committee. 1996. Report to the state legislature of
the state of Louisiana.
Maryland Department of Health and Mental Hygiene, Alcohol and Drug Abuse Adminis-
tration. 1990. Task Force on Gambling Addiction in Maryland, Author: Baltimore,
Maryland.
National Gambling Impact Study Commission. 1999. National' Gambling Impact Study
Commission Final Report, (June 18), www.ngisc.gov.
Phillips DP, Welty WR, Smith MM. 1997. Elevated suicide levels associated with legalized
gambling. Suicide and Life- Threatening Behavior. 27: 373-378.
Politzer RM, Morrow JS, Leavey SB. 1981. Report on the societal cost of pathological
gambling and the cost-benefit/effectiveness of treatment. The Johns Hopkins Compulsive
Gambling Counseling Center, presented at Fifth National Conference on Gambling and
Risk-Taking, 1-48.
, Productivity Commission. 1999. Australia's Gambling Industries: Final Report, Summary
(Nov. 26, Rep. No. 10).
Ryan TP, Speyrer JF, with Beal ST, Burckel DV, Cunningham BR, Kurth MM, Scott
LC, Wall JL, Westphal JR. 1999. Gambling in Louisiana: A Benefit/Cost Analysis.
Prepared for the Louisiana Gaming Control Board, Louisiana State University Medical
Center (April).
SMR Research Corporation. 1997. The Personal Bankruptcy Crisis, 1997: Demographics,
Causes, Implications fj Solutions. SMR: Hackettstown, NJ.
17
.
.
.
Shaffer HJ, Hall MN, Vander Bilt J. 1997. Estimating the prevalence of disordered gambling
behavior in the United States and Canada: A meta-analysis. Harvard Medical School:
Boston.
Smith F, Craig WJ. 1992. Who's in for how much? CURA Reporter, Center for Urban and
Regional Affairs. 22, 1: 11-14, 16. (Plus data supplied to the authors.)
South Dakota Legislative Research Council. 1998. Economic and fiscal impacts of the
South Dakota gaming industry. 1-146, + Attachment: Letter of January 7, 1999 to Mr.
Terry Anderson, Director, SD Leg. Res. Council, Table "South Dakota Total Estimated
Incremental Social Costs."
Rappaport J, Wilkerson C. 2001. "What are the Benefits of Hosting a Major League Sports
Franchise?" Economic Review, Federal Reserve Bank of Kansas City, 86, 1, 55-86.
Thompson WN, Gazel R, Rickman D. 1996. The social costs of gambling in Wisconsin.
Wisconsin Policy Research Institute Report. 9, 6: 1-44.
Thompson WN, Gazel R, Rickman D. 1998. Social costs of gambling: A comparative studY0f
nutmeg and cheese state gamblers. Twelfth National Conference on Problem Gambling,
NV, published in Gaming Research fj Review Journal. 1998. 5, 1: 1-15.
Thompson WN, Quinn FL. 1999. An economic analysis of machine gambling in South
Carolina. The Education Foundation of the South Carolina Policy Council, May.
Tice DL. 1995. Why the press hates gaming. Casino Executive, August, 57-58.
Volberg, RA, DA Gerstein, EM Christiansen, J Baldridge. 2001. Assessing Self-Reported
Expenditures on Gambling Managerial and Decision Economics, 22, 1-3, 77-96.
Westphal JR, Johnson LJ, Stevens L. 1999. Estimating the social costs of gambling disorders
in Louisiana for 1998, Study Performed Under Subcontract From University Of New
Orleans, UNO #313-20-4124 (23 March).
18