Since the 1970s, the availability of
gambling has grown ten-fold in the United States. Today, a person can make a
legal wager of some sort in every state except Utah and Hawaii; 38 states have
lotteries, 28 states have casinos and 22 states have off-track betting (National
Gambling Impact Study Commission, 1999; North American Association of State
& Provincial Lotteries, 2003). Just
as telling as the expansion of gambling into new jurisdictions is the growth of
the gambling industries. Between 1975
and 2001, revenues from legal wagering in the United States grew twenty-fold,
from $3 billion to $64 billion while gambling expenditures more than doubled as
a percentage of personal income (Christiansen, 2000; Christiansen &
Sinclair, 2002; Kallick et al, 1976).
The main purpose
of this survey, funded by the Responsible Gaming Association of New Mexico, was
to determine the scope of problem gambling in New Mexico and identify the
groups in the population most affected by the disorder. The results of this study are also intended
to provide information about the impacts of problem gambling in New Mexico and
will be useful to the Association, the State and other stakeholders in efforts
to help individuals and groups in New Mexico affected by this disorder.
This
report is organized into several sections for clarity of presentation. The Introduction includes a definition
of the terms used in the report, a brief review of methods for assessing
problem gambling and conducting prevalence surveys in the general population,
and background information on gambling and problem gambling in New Mexico. This is followed by a review of research on Risk
Factors for Problem Gambling.
The Methods section addresses the details of conducting the
survey. The next four sections present
findings from the survey in the following areas:
·
gambling in New
Mexico;
·
prevalence of
problem gambling in New Mexico;
·
comparing
non-problem and problem gamblers in New Mexico; and
·
comparing Native
Americans with other population groups in New Mexico.
The report
concludes with a summary of the findings of the study and consideration of the
number of problem gamblers likely to seek treatment in New Mexico on an annual
basis as well as suggestions for the future development of services for problem
gamblers and their families in New Mexico.
There are two appendices to the report including a technical section
comparing the performance of the two problem gambling screens used in the New
Mexico survey and a copy of the questionnaire.
Gambling is a broad concept that includes
diverse activities, undertaken in a wide variety of settings, appealing to
different sorts of people and perceived in various ways by participants and
observers. Failure to appreciate this
diversity can limit scientific understanding and investigation of gambling and
gambling problems. Another reason to
note the differences between various forms of gambling arises from accumulating
evidence that some types of gambling are more strongly associated with
gambling-related problems than others (Abbott & Volberg, 1999a). People take part in gambling activities
because they enjoy them and obtain benefits from their participation. For most people, gambling is generally a
positive experience. However, for a
minority, gambling is associated with difficulties of varying severity and
duration. Some regular gamblers develop
significant, debilitating problems that also typically result in harm to people
close to them and to the wider community (Abbott & Volberg, 1999a).
Gambling problems exist on a continuum and
there is mounting evidence that such problems may not necessarily be chronic
and progressive (Abbott et al, 2004c).
Gambling problems vary in duration and severity and a substantial
proportion of these problems occur in persons who are not pathological gamblers
but who engage in risky gambling. Risky
gambling includes a broad range of gambling behaviors (e.g., persistently
betting more than planned or spending more time gambling than intended, chasing
losses and borrowing money to gamble) as well as cognitions (e.g.,
superstitions, illusions of control and misunderstandings about the nature of
probability and randomness) that support the adoption and maintenance of risky
gambling behaviors. Although risky
gambling is not a clinically defined condition, it is generally viewed as
gambling in ways that may pose a risk of physical or emotional harm to the
gambler or others but has not produced effects that would result in a clinical
diagnosis. Figure 1 presents the continuum of gambling involvement and
gambling problems graphically. The
terms used in the present report are not identical to the terminology included
in this illustration; however, our view of the continuum of gambling problems
as highly dynamic and not inevitably progressive is very similar.
Figure 1: OPGRC Problem
Gambling Framework
Pathological gambling was first
included in the third edition of the Diagnostic and Statistical Manual
(DSM-III) of the American Psychiatric Association (1980). Each subsequent revision of this manual has
seen changes in the diagnostic criteria for pathological gambling. The essential features of pathological
gambling are presently defined as: (1) a continuous or periodic loss of control
over gambling; (2) a progression, in gambling frequency and amounts wagered, in
the preoccupation with gambling and in obtaining monies with which to gamble;
and, (3) a continuation of gambling involvement despite adverse consequences
(Rosenthal & Lesieur, 1992). A
formal diagnosis of pathological gambling is arrived at by an appropriately
qualified and experienced clinician following an extensive clinical interview. To make a diagnosis, a clinician must
determine that a patient has met five or more of the ten diagnostic indicators
associated with pathological gambling.
Table 1 on the following page presents the current diagnostic criteria
for pathological gambling (American Psychiatric Association, 1994: 618):
Table
1:
Diagnostic Criteria for Pathological Gambling
|
Persistent and recurrent maladaptive gambling behavior as indicated
by five (or more) of the following:
|
|
Preoccupation
|
Preoccupied with gambling (e.g. preoccupied with reliving past
gambling experiences, handicapping or planning the next venture, or thinking
of ways to get money with which to gamble)
|
|
Tolerance
|
Needs to gamble with increasing amounts of money in order to achieve
the desired excitement
|
|
Withdrawal
|
Restless or irritable when attempting to cut down or stop gambling
|
|
Loss of Control
|
Has repeated unsuccessful efforts to control, cut back or stop
gambling
|
|
Escape
|
Gambles as a way of escaping from problems or relieving dysphoric
mood (e.g. feelings of helplessness, guilt, anxiety or depression)
|
|
Chasing
|
After losing money gambling, often returns another day in order to
get even (“chasing” one’s losses)
|
|
Lying
|
Lies to family members, therapist or others to conceal the extent of
involvement with gambling
|
|
Illegal Acts
|
Committed illegal acts, such as forgery, fraud, theft or
embezzlement, to finance gambling
|
|
Risked Relationship
|
Has jeopardized or lost a significant relationship, job, or
educational or career opportunity because of gambling
|
|
Bailout
|
Relies on others to provide money to relieve a desperate financial
situation caused by gambling
|
|
The gambling behavior is not better accounted for by a Manic Episode.
|
The term problem gambling is used
in a variety of ways. In some
situations, its use is limited to those whose gambling-related difficulties are
less serious than those of pathological gamblers. In other situations, it is used to indicate all of the patterns of
gambling behavior that compromise, disrupt or damage personal, family or
vocational pursuits (Cox et al, 1997; Lesieur, 1998). From this perspective, pathological gambling can be regarded as
one end of a continuum of gambling-related problems. Problem gamblers, as well as
individuals who score even lower on problem gambling screens (at-risk
gamblers) are of concern because they represent much larger proportions
of the population than pathological gamblers.
These groups are also of interest because of the possibility that their
gambling-related difficulties may become more severe over time. Problem and at-risk gamblers are also
important because the prospects of changing their behavior through effective
public awareness and education campaigns are better than for more troubled gamblers
(Hodgins & el-Guebaly, 2000; Shaffer & Korn, 2002).
In considering the public health risks of
problem gambling, it is important to note that not all of the features of
pathological gambling need be present at one point in time (Abbott &
Volberg, 1999a; Gerstein et al, 1999).
Some of the impacts that at-risk, problem and pathological gamblers may
experience include psychological difficulties, such as anxiety, depression,
guilt, exacerbation of alcohol and drug problems and attempts at suicide as
well as stress-related physical illnesses such as hypertension and heart
disease. Interpersonal problems include
arguments with family, friends and co-workers and breakdown of relationships,
often culminating in separation or divorce.
Job and school problems include poor work performance, abuse of leave
time and loss of job. Financial effects
loom large and include reliance on family and friends, substantial credit card
debt, unpaid creditors and bankruptcy.
Finally, there may be legal problems as a result of criminal behavior
undertaken to obtain money to gamble or pay gambling debts (Lesieur, 1998;
Volberg, 2001a).
State governments began funding services for
individuals with gambling problems in the 1980s. As a first step toward establishing these services, policy makers
sought information about the number of people who might seek help for their
gambling problems and what they looked like.
In responding to these questions, researchers adopted methods from the
field of psychiatric epidemiology to investigate the prevalence of gambling
problems in the general population.
In the 1980s, few tools existed to measure
gambling problems, and only one—the South Oaks Gambling Screen (SOGS)—had been
rigorously tested for performance (Lesieur & Blume, 1987). Closely based on the original psychiatric
criteria for pathological gambling, the SOGS was developed to screen for
gambling problems in clinical populations.
Like other tools in psychiatric research, the SOGS was quickly adopted
for use in epidemiological research as well as in clinical settings. The SOGS was first used in a prevalence
survey in New York State (Volberg & Steadman, 1988). Since then, the SOGS—or one of several variants
of the original screen, most often the SOGS-R (Abbott & Volberg, 1996)—has
been used in population-based research in more than 50 jurisdictions in the
United States, Canada, Europe, Asia, and Oceania (Abbott & Volberg, 1996,
2000; Bondolfi, Osiek & Ferrero, 2000; Duvarci et al, 1997; Lund & Nordlund, 2003; Orford et
al, 2003; Productivity Commission, 1999; Shaffer, Hall & Vander Bilt, 1999;
Volberg, 2001a; Volberg, Abbott et al, 2001; Welte et al, 2001).
Beginning in the 1990s, dissatisfaction with
the SOGS grew, particularly among Australian and Canadian researchers. The main criticism of the SOGS was that this
screen was developed and tested in a clinical setting and the characteristics
of its performance in community samples were unknown (Walker & Dickerson,
1996; Wiebe, Single & Falkowski-Ham, 2001). However, this view ignores studies that did assess the SOGS and
SOGS-R in general population contexts (Abbott & Volberg, 1996; Stinchfield,
2002). There have been additional
criticisms of the SOGS, reflecting concerns that the screen does not reflect
the DSM conceptualization of pathological gambling; that some of the items
would be equally endorsed by non-problem gamblers; that the lifetime frame of
reference of the original screen overestimates the current prevalence of
gambling problems; and that the screen is insensitive to culturally diverse
contexts (Abbott et al, 2004c; Battersby et al, 2002; Thomas, Jackson &
Blaszczynski, 2003).
In 1994, the fourth edition of the Diagnostic and Statistical Manual (DSM-IV)
adopted a new set of criteria for the diagnosis of pathological gambling
(American Psychiatric Association, 1994).
The new criteria incorporated empirical research—primarily
epidemiological research—that more firmly linked pathological gambling
conceptually to other addictive disorders like alcohol and drug dependence
(Lesieur & Rosenthal, 1998).
One response to this and other changes in
the gambling studies field was the development of a large number of new screens
for problem and pathological gambling (Govoni, Frisch & Stinchfield,
2001). Some of these new screens are based on the
most recent revision of the DSM; others use a broader definition of problem
gambling. In addition to ongoing use of
the SOGS and SOGS-R, the screens and measures that have been most widely used
in prevalence surveys since the late 1990s include the DSM-IV-MR, the NODS and
the Canadian Problem Gambling Index (CPGI) (see Abbott & Volberg, in press
for a review). While performance on
these various measures generally shows moderate to high levels of agreement,
especially in the case of people with severe problems, they generate somewhat
different prevalence estimates.
On the face of it, finding out how many
people there are in a community with gambling problems appears to be
straightforward. A random sample of the
population is selected, assessed using a valid problem gambling instrument, and
a prevalence estimate is then generated from the results. In reality, for a variety of financial and
technical reasons, this process is an evolving and increasingly complex one.
For one thing, because problem gambling is a
relatively rare phenomenon, large sample sizes are necessary to conduct
meaningful analyses. Without a large
sample size, it becomes difficult to determine whether differences observed in
a study are in fact generalizable to the population from which it is
drawn. Most gambling researchers agree
that it is essential to interview large samples of respondents to establish
reliable prevalence estimates, particularly for subgroups of the
population.
Another issue that requires careful attention
is the sampling design, especially as it pertains to those who choose not to
particulate in surveys. For one thing,
increasing attention has been devoted to not only randomly sampling households,
but also randomly sampling withinhouseholds (using increasingly complex methods) in order to ensure
that those who answer the phone (often females) are not over-represented. Also, because of the fact that response
rates in general are declining, it is vital that researchers devote special
attention to achieving the highest possible response rates. In contrast to popular polls conducted by
major news organizations (generally done over the course of a few days), the
New Mexico problem gambling prevalence survey relied heavily on substantial
callbacks—re-contacting potential respondents several times to encourage their
participation. Completing substantial
callbacks requires significant resources and time and also means that only
interviewers with demonstrated success at completing lengthy interviews and
converting those whom researchers call “refusals” are employed. All of these developments mean that
prevalence research is getting more complex and more expensive.
Throughout the world, gambling participation
and attitudes toward gambling are linked to the communities in which these
behaviors occur and to the norms and values of members of those
communities. Differences have been
found in the types of gambling preferred by middle-class and blue-collar
gamblers, by white and black Americans and by men and women (Dixey, 1996; Drake
& Cayton, 1945; Henslin, 1967; Hraba & Lee, 1996; Light, 1977; Zola,
1964). It is equally important to note
that individual and community definitions of gambling can vary widely. For example, a recent Gallup poll found that
52% of respondents defined stock market investment as a form of gambling while
22% did not consider buying lottery tickets to be gambling (Gallup, 1999).
Gambling in New
Mexico
The citizens of New Mexico have access to a
wide range of legal gambling opportunities available throughout the state. The major forms of commercial gambling in
New Mexico include Indian casino gambling, pari-mutuel wagering on horse races,
the New Mexico Lottery and electronic gaming machines both at racetracks and at
social and fraternal clubs.
As in many other states, pari-mutuel wagering
on horse races is the oldest major form of legal gambling in New Mexico. In 1997, after decades of declining
attendance, the five commercial horse tracks in New Mexico were permitted to
begin operating slot machines. In 2004,
attendance reached over 1 million at the five tracks and total handle reached $166
million of which $130 million was returned to the public in winnings (New
Mexico Racing Commission, 2005).
Separately from racing handle, slot machines were expected to generate
approximately $176 million in gross revenues in FY 2005. One-fifth of these gross revenues goes to
purses and another 25% goes to the State’s general fund. In 2001, the New Mexico racetracks were
permitted to increase the number of machines from 300 to 750 per location
(Cole, 2005).
The New Mexico Lottery was authorized in 1995
and launched in 1996. There are
approximately 1,100 lottery ticket sales outlets throughout the state. State lottery proceeds in New Mexico are
earmarked for education and, since the lottery’s inception, approximately
32,000 college scholarships have been funded with lottery proceeds (Heild,
2005). In FY 2004, lottery ticket sales
reached $147 million of which 24%--or $36 million—went to the scholarship
program (New Mexico Lottery, 2005).
In the wake of the Indian Gaming Regulatory
Act, Governor Gary Johnson negotiated compacts with a number of New Mexico
tribes. These compacts were approved by
the 1997 New Mexico State Legislature, paving the way for the 1999 Compact
Negotiation Act, which established the process for negotiations between the
Tribes and the State. Thirteen tribes
presently operate 18 casinos throughout New Mexico. Under the compacts, the tribes pay the State a percentage of the
“net win” from slot machines and, in calendar year 2004, this net win reached $484
million (New Mexico Gaming Control Board, 2005).
Finally, the state’s veterans and fraternal
clubs are permitted to offer charitable gaming through electronic gaming
machines and bingo. Sixty-one clubs are
permitted to operate a maximum of 15 slot machines per location although most
have fewer machines. Of the more than
$10 million that was won on club gaming machines in fiscal year 2004, the state
received just over $1 million, with an additional $1.9 million going to
charitable causes. Bingo offerings have
declined substantially since the introduction of Native American casinos but
bingo still earned $30 million in FY 2005 prior to paying out prizes and
expenses (Gallagher, 2005).
Altogether,
there are presently about 14,750 gaming devices operating at New Mexico’s five
racetracks, 18 casinos and 61 veteran/fraternal clubs (or approximately 1
machine for every 1,000 New Mexico adults).
Overall, gambling in New Mexico generates approximately $1 billion
in gross annual revenues. Figure 2
shows the relative proportion of the New Mexico gaming industry represented by
the different sectors of the industry.
Even without considering revenues from table games—which typically
represent between 20% and 30% of casino revenues—tribal casinos represent the
largest sector of the gaming industry in New Mexico. The racing industry, with revenues from both pari-mutuel wagering
and gaming machines, is the second largest sector of the industry followed by
the New Mexico Lottery. Bingo and
gaming machines at veterans and fraternal clubs are the smallest sectors of the
gaming industry in New Mexico, representing only about 2% of gross gaming
revenues.
Figure 2: New Mexico Gaming Industry
Problem
Gambling Services in New Mexico
Although a growing number of states fund
services for problem gamblers, the major sources of help for problem gamblers
and their families remain the self-help groups, Gamblers Anonymous and
Gam-Anon, and not-for-profit state councils on problem gambling.
In New Mexico, both the tribes and the
racetracks are required to contribute Ľ of 1% of gross slot machine revenues to
problem gambling programs. In FY 2004,
with additional contributions from the New Mexico Lottery, funding for problem
gambling services in New Mexico reached $2 million (or about 2/10ths of 1% of
gross gaming revenues). New Mexico has
a well-advertised toll-free helpline, a fund that finances treatment for those
who cannot afford help and training programs for healthcare workers and casino
workers. However, there is no state
coordination of spending on problem gambling services and critics have argued
that there is a need for a more balanced approach to targeting these resources
(Heild, 2005).
The New Mexico Council on Problem Gambling
was established in 1998 and operates the state’s bilingual, 24-hour
helpline. Callers are referred to local
professional treatment services, debt counseling programs, Gamblers Anonymous
meetings and/or Gam-Anon (New Mexico Council on Problem Gambling, 2006). In 2001, the New Mexico Council established
an Indigent Care Treatment Fund that has provided over $250,000 to New Mexico
residents in need of problem gambling treatment who could not afford to pay for
these services.
In addition to the New Mexico Council,
Gamblers Anonymous provides assistance through self-help (12-step) programs in
Albuquerque, Los Lunas, Rio Rancho, Santa Fe and Tularosa. The group meetings characteristic of
Gamblers Anonymous, wherever these are held, offer a fellowship with other
individuals who are themselves working to overcome their problems with
gambling. Gam-Anon provides similar
services to those who have family members with gambling problems (New Mexico
Gaming Control Board, 2005).
There are different ways to characterize or
classify risk factors for problem and pathological gambling. The National Research Council (1999)
identified risk factors at three levels – those which initiated gambling, those
which caused progression from social to problem or pathological gambling, and
those associated with chronicity and maintenance of problematic gambling. A recent trend in the behavioral sciences
has been a convergence of biological, psychological, and social theories into a
biopsychosocial
perspective that attempts to explain psychiatric conditions (Engel, 1980). From this perspective, behavioral illnesses
are caused by a combination of risk factors from three separate domains,
including disturbances in brain function, altered psychological processes and
social factors.
In this section of the report, we summarize
the most current scientific evidence on the biological, psychological and
social risk factors that contribute to the development of pathological
gambling. As is the case with many
other psychiatric disorders, the current evidence suggests that there is a combination
of risk factors that contribute to pathological gambling. Understanding of these risk factors serves
to focus on areas in prevention, treatment and early intervention where efforts
may be most effectively and efficiently concentrated. A clearer understanding of the risk factors associated with
pathological gambling can also help direct public policies in relation to legal
gambling.
Investigating the biological causes of
pathological gambling is uniquely challenging because there are no consistent
animal models and because there are likely subtypes of pathological gambling
that may or may not share certain biological characteristics. Nevertheless, research into the biological
causes of pathological gambling is important, not least because, with no
neurotoxic substances involved, this disorder serves as a natural model of
addictive behaviors.
Genetic
Contribution. Genetic studies
are important in understanding psychiatric illnesses because they help prove
that these disorders are biological diseases and not simply a matter of
excessive appetites or immoral behavior.
There are several approaches to identifying the impact of genetics on
pathological gambling, including family studies to determine the heritability
of the disorder; twin studies to tease out genetic versus environmental
influences, and studies that focus on differences in genetic factors of
pathological and non-pathological gamblers.
Family studies have found high rates of
pathological gambling among family members of pathological gamblers as well as
among substance dependent patients (Gambino et al, 1993; Lesieur, 1985). A recent meta-analysis of 28 family studies
examining pathological gambling found a relatively weak effect overall although
a stronger familial effect appears to hold for those with more severe gambling
problems (Walters, 2001). This is
similar to findings related to alcohol dependence, suggesting a parallel
process and supporting the notion that a small genetic effect can have a
powerful impact on behavior when exposed to an environment that allows genetic
vulnerabilities to be expressed in a clinically significant manner.
Twin studies are considered more powerful
than family studies because both genetic and environmental impacts on the
heritability of disorders are incorporated.
If a disorder has a true genetic component, monozygotic (identical)
twins will have a higher frequency of the disorder compared to dizygotic
(fraternal) twins who will, in turn, have a higher frequency of the disorder
than other first-degree relatives or the general population. The largest twin study of pathological
gambling, based on the Vietnam Era Twin Registry, found that this disorder was
as heritable as alcohol dependence and that genetic factors were the
predominant contributor to familial transmission of pathological gambling (Slutske
et al, 2000, 2001). In a smaller study,
heritability explained “high action” gambling in male twins but not “low action”
gambling (Winters, 1999).
Overall, genetic studies of pathological
gambling support the notion that there are clinically significant, inheritable
risk factors for pathological gambling. These factors may determine one’s
initial emotional response to gambling or code for a predisposition to
impulsivity/addictive behaviors. They
may also be responsible for an inability to control behavior or an inability to
adapt and learn from losing.
Neurotransmitter
Functioning. Neurobiological
research has identified genetic differences between pathological gamblers and
controls in the dopamine, serotonin and norepinephrine systems. Several studies have found differences
between pathological gamblers and controls in dopamine receptor genes and in
serotonin transporter genes, suggesting that the disorder may be associated
with deficiencies in the brain’s reward systems (Comings et al, 2001; Ibanez et
al, 2000, 2003; Perez et al, 1999).
Recent advances in neuroimaging techniques
have allowed researchers to identify abnormalities in areas of the brain that
control decision-making, reward processing and information processing in
pathological gamblers similar to those among persons with substance use
disorders (Goudriaan et al, 2004; Potenza et al, 2003; Potenza & Winters,
2003). Pathological gamblers have been
shown to have alterations in levels of the dopamine, serotonin and
norepinephrine systems, all implicated in the neurobiological roots of
impulsivity (Chambers & Potenza, 2003; Potenza, 2001).
Serotonin has been implicated in the regulation
of impulsivity and compulsivity, norepinephrine in the mediation of arousal and
novelty seeking, and dopamine in reward and reward dependency. Some researchers believe that all three
neurotransmitters are involved in pathological gambling, but at different
stages of the gambling cycle.
Anticipatory arousal may be linked to the noradrenergic system, the
‘high’ of the actual gambling episode may be associated with the serotonergic
system, and difficulties extinguishing the behavior may be under the aegis of
the dopaminergic system (Rosenthal & Fong, 2004).
While these results are important, a great
deal more work is needed to investigate the role of neurotransmitters in the
development and maintenance of pathological gambling. As such work proceeds, it will be important to include larger samples as well as paying greater
attention to racial composition and subtypes of problem gamblers, in order to
clarify the relationship between these genetic risk factors and the precise
behaviors they may encode.
Psychological factors determine how people
interact with the environment and with others and how they view themselves and
the world. Personality traits, ways in
which people manage stressful events, and comorbid psychiatric disorders are
all important psychological factors related to the development of pathological
gambling.
Comorbidity. Like
other addictive disorders, pathological gamblers have much higher rates of
co-occurring psychiatric conditions and substance use disorders than are found
in the general population. Rates of these
disorders are particularly high among pathological gamblers, both clinically
and in the general population. For
example, two recent national surveys found rates of alcohol and substance
dependence among problem and pathological gamblers in the general population
that are approximately ten times higher than among low risk gamblers and
nongamblers (Gerstein et al, 1999; Welte et al, 2001). There is also evidence that mood disorders—primarily
major depression—frequently co-occur with problem and pathological gambling
(Gerstein et al, 1999; Specker et al, 1995).
There are several theories as to why comorbid
disorders are so common in pathological gamblers. There is disagreement about whether these disorders are caused by
the same biological and psychological risk factors or whether one disorder
causes the other (i.e., depression causes pathological gambling or vice
versa). In developing effective
interventions for pathological gambling, it is important to understand not only
why comorbid conditions are so common but how they may cause pathological
gambling. It is also important to
improve our understanding of how gambling may be used to self-medicate for
other disorders, whether psychological or physical.
Personality
Traits. There is research
suggesting that certain aspects of personality development, including
impulsivity and competitiveness, can predispose toward pathological
gambling. However, simply having these
personality traits is not enough to “cause” pathological gambling nor does an
absence of these traits protect from the development of gambling problems.
Pathological gambling is classified as an
impulse control disorder and it is important to understand precisely how
impulsivity, which contains elements of risk-taking, sensation seeking and
arousal, contributes to loss of control over gambling. Research in clinical settings shows that
pathological gamblers tend to be highly impulsive compared to healthy controls
and suggests that pathological gamblers are less likely to think about future
consequences and are more likely to act in the moment (Blaszczynski et al,
1997; Petry, 2001; Vitaro et al, 1999).
Sensation seeking tends to be high among
casino and racetrack gamblers and low among electronic gaming machine
players. The difference seems to
conform to a distinction that is made in the gambling studies field between
those who play competitive, skill-based games (“action seekers”) and those who
play non-competitive games primarily based on luck (“escape gamblers”)
(Lesieur, 1988; Lesieur & Blume, 1991).
Pathological gamblers who are sensation seekers are more apt to be early
onset male gamblers who wager primarily on competitive skill-based games and
are likely to have other addictions involving risk or danger, including
alcohol, drugs and sex.
Stress
and Coping. Addictions research has
made major strides in recent years in demonstrating the contributions of
internal and external stressors in the initiation and maintenance of substance
use disorders. However, research on the
relationship between pathological gambling and stress is in its infancy. Nevertheless, it appears that early
interventions for problem gambling that focus on stress reduction may be
helpful in preventing full blown development of the disorder.
Research into mood disorders has linked early
adverse experiences as a contributing factor to the development of depression
as well as a mediator of treatment response (Heim et al, 2004). Recent research by Petry et al (in press)
found high rates of childhood maltreatment, including emotional abuse and
neglect, physical abuse and neglect, and sexual abuse among male and female
treatment-seeking pathological gamblers with severity of maltreatment strongly
associated with earlier age of onset of gambling and increased severity of
gambling problems. These results
suggest the importance of further investigation into the role of childhood
maltreatment in the development of pathological gambling as well as the need
for research on resiliency factors shown by some who experience childhood
maltreatment but do not develop addictive disorders including pathological
gambling. This area of research is
critical in order to begin to identify protective factors that can be utilized
for prevention.
Coping (or defense) mechanisms are dynamic
processes that are used to resolve psychological conflicts. Such mechanisms are learned responses to
stress that people use to minimize uncertainty or emotional pain. Pathological gamblers are more likely than
non-problem gamblers to make use of a range of coping mechanisms that are
considered immature and counterproductive, including avoidance, procrastination
and dissociation (Brown, 1986; Diskin & Hodgins, 1999; Jacobs, 1988;
Rosenthal, 1996, 2004). Pathological
gamblers appear to be more boredom-prone although the relationship between
boredom susceptibility, depression and problematic gambling requires further
exploration. Finally, studies have
demonstrated that gambling in general is highly arousing and there is research
suggesting that some pathological gamblers are motivated by the excitement of
gambling rather than by the desire to win money (Anderson & Brown,
1984).
Learning
Theories. Some researchers believe
that addictive behaviors occur as a direct result of learned experiences. While learning theories are likely to be
useful in understanding pathological gambling, much more research is needed in
this area. Gambling activities operate
directly on the principles of intermittent reinforcement, one of the most
effective approaches to reinforcing and perpetuating behavior. Gambling also promotes cognitive distortions
and irrational thinking, an area that has received far more research attention
(Gilovich, 1983; Ladouceur & Walker, 1996; Langer, 1975; Toneatto et al,
1997).
What remains unclear is exactly how cognitive
distortions are acquired and maintained although we can speculate that these
distortions probably arise in response to a combination of personality traits,
adaptation strategies and biological mechanisms that are responsible for learning. Further research is needed on the
relationship between specific forms of gambling and the acquisition of
cognitive distortions as well as the identification of modifiers of cognitive
distortion.
There are a number of social factors that influence
gambling behavior and may contribute to the development of pathological
gambling. From a policy perspective, one
of many important questions is whether increasing access to gambling increases
rates of pathological gambling in the population and, if so, whether putting
prevention programs in place prior to increasing access will limit the number
of people who develop problems.
Age. Internationally, research has identified
high rates of problem gambling among adolescents. This, along with reports of especially early ages of onset among
treatment-seeking pathological gamblers, has formed the basis for the widespread
belief that early initiation into gambling is a risk factor for later
pathological gambling (Gupta & Derevensky, 1998; National Research Council,
1999). However, Rosenthal and Fong
(2004) point out that early experiences with gambling occur as part of normal
social development and that early exposure to family card games or other
socially managed gambling activities could serve as a protective factor in the
development of problem gambling. The
question is whether adolescent experimentation with gambling can be managed in
ways that promote “maturing out” and transition to non-problematic involvement
in gambling.
Any consideration of age as a risk factor for
problem gambling must consider the other end of the life span and the impact of
legal gambling on older adults.
Prevalence surveys do not support the notion that older adults are at
greater risk than younger adults for the development of problem gambling
(National Research Council, 1999; Volberg & McNeilly, 2003). However, research does show that older
adults are more likely to gamble now than in the past (Gerstein et al, 1999)
and it is possible that developmental issues such as impaired physical status,
loss, isolation and limited recreational alternatives may contribute to growing
numbers of older adults experiencing gambling-related problems. There is evidence that older adults
represent a growing proportion of callers to problem gambling helplines in the
U.S. (Volberg & McNeilly, 2003).
Gender. In most of the United States and other
Western countries, rates of problem and pathological gambling are about two
times higher among men than among women (Abbott & Volberg, 1996; American
Psychiatric Association, 1994; Gerstein et al, 1999; Volberg, 2001a,
2003b). In some jurisdictions, notably
Australia and some U.S. states where electronic gaming machines are widely
distributed, rates of problem and pathological gambling are about equal for men
and women (Productivity Commission, 1999; Volberg, 2003b).
Compared to female pathological gamblers,
male pathological gamblers are younger, have higher incomes, began gambling at
an earlier age, have a longer duration of gambling problems, have more severe
legal problems, are more likely to have alcohol or drug related problems, to be
diagnosed with antisocial personality disorder, and to gamble on cards, sports
or the racetrack (Grant & Kim, 2002; Ladd & Petry, 2002; Potenza et al,
2001). Women are more apt to describe
loneliness and relationship problems as precipitants of their gambling; they
are also more likely to be diagnosed with depression. Women also report starting to gamble later in life than men.
These studies seem to support a longstanding
characterization of men as early onset gamblers who play competitive, skill
based games, and women as late onset gamblers who play non-competitive, luck
based games. According to this
description, men gamble for excitement or action while women gamble to numb
themselves or escape. However, an
analysis of “early onset” and “late onset” gamblers in the general population
in Arizona found that the majority of “action gamblers” in that sample actually
identified slot machines as their favorite gambling activity (Volberg,
2003a). Clearly, more research is
needed to understand the relationships between gambling careers, gambling
preferences and the development of gambling problems.
Another consistent finding is that women’s
gambling progresses more rapidly to problematic gambling (Ladd & Petry,
2002; Paton-Simpson, Gruys & Hannifin, 2004; Potenza, 2001; Tavares et al,
2001). Various explanations have been
offered for this phenomenon, including the greater stigma attached to women’s
gambling problems, the limited financial resources available to women compared
with men, experiences of loss and the stresses of caring for children and aging
parents, and the greater difficulty of hiding gambling excursions and debts
from family and friends. Breen and
Zimmerman (2002) present data on gambling problems related to electronic
gambling machines to suggest a radically different explanation: that it is not
gender which accounts for the telescoping phenomenon, but rather involvement in
machine gambling.
Ethnicity
and Culture. Most research on
problem and pathological gambling has focused on white male gamblers. However, there is growing evidence to
support the notion that disproportionate numbers of African Americans,
Hispanics, Asians, and Native Americans are problem and pathological gamblers
(Abbott et al, 2004c; Volberg, 2001a, 2003b; Volberg & Abbott, 1997; Welte
et al, 2001; Zitzow, 1996). While there
is research suggesting that a strong ethnic identity can act as a protective
factor against drug use in some ethnic groups, there is no research examining
this relationship with regard to gambling.
Another
cultural factor that appears to contribute to pathological gambling is the
immigration process. Gambling may
appeal to immigrants as an enticing way to make money but also as a
recreational activity that does not require English language ability, provides
opportunities for socialization and relieves the stresses of
acculturation. In one small study, Petry
et al (2003) surveyed Southeast Asian refugees in the community and identified
59% of their sample as pathological gamblers.
Societal
Attitudes Toward Gambling. Attitudes
toward gambling in the U.S. have always been highly ambivalent. On the one hand, gamblers have been
stigmatized as greedy and immoral. On
the other hand, gambling has often been identified with American ideals of
independence, risk-taking and entrepreneurship. Prior to the involvement of governments in legislation and
regulation, gambling was viewed as a morally suspect industry with close
associations to organized crime. Over
the last 30 years, as state legislatures have turned to gambling as a way to
raise revenues without increasing taxes, attitudes have shifted and gambling is
now generally viewed as an acceptable form of recreation and
entertainment.
This change in attitude has been accompanied
by two other significant developments.
The first development is the “normalization” of gambling as these
activities spread far beyond the confines of gambling-specific venues and out
into the community. The second
development is the “democratization” of gambling as groups that would not have
gambled previously—particularly women and older adults but also youth and
ethnic and cultural minorities—now do.
Access
to Legal Gambling. The relationship
between increased access to legal gambling and the prevalence of problem and
pathological gambling is an important issue in light of the remarkable
expansion of gambling throughout the United States and internationally. Commissions and official government reviews
in a number of countries including the United States, Great Britain, Australia
and New Zealand have all concluded that increased gambling availability has led
to an increase in problem gambling and that future increases will generate
additional problems (Abbott, 2001; Gambling Review Body, 2001; National
Research Council, 1999; Productivity Commission, 1999). Historically, the introduction and expansion
of new forms of gambling, especially continuous forms such as electronic gaming
machines, track betting and casino table games, have resulted in substantially
increased rates of problem gambling.
This has been documented across whole populations as well as within
sub-populations that previously had low levels of gambling participation.
Expansion of gambling has been largely due to
legislation permitting increases in gambling opportunities, demonstrating how
public policies can intersect with clinical conditions. Increased gambling opportunities create more
problem gamblers by increasing the risk of exposure. As more people gamble, the risks are greater that individuals
with specific vulnerabilities will gamble and develop problems related to their
gambling. Results from a number of studies
demonstrate that the location of a major gambling venue in a community is
associated with rates of problem and pathological gambling that are
approximately double the rates in communities without such venues (Gerstein et
al, 1999; Welte et al, 2004).
There is research to suggest that the
prevalence of problem gambling will eventually level out, even when
accessibility continues to increase.
However, rates are likely to rise dramatically before stabilization
occurs and active measures, including raising public awareness, expanding services
and strengthening regulatory measures are probably required to achieve
stabilization sooner rather than later (Abbott et al, 2004c).
Role
of Technology. The
gambling industry has taken advantage of recent technological advances to
increase the efficiency, reliability and accessibility of gambling
options. The most dramatic changes have
been the introduction of computer technologies in electronic gaming machine
design, changes in the accessibility of credit and financial services for
gamblers, and the creation of new, online forms of gambling.
There is a strong belief among gambling
counselors and researchers that electronic gaming machines are more addictive
than other forms of gambling (Turner & Horbay, 2004). Electronic gaming machines (EGMs) are the
most profitable form of gambling; they account for 80% of casino profits in the
U.S. and Canada and are found in a growing number of non-traditional gambling
locations. Internationally, a growing
proportion of problem and pathological gamblers contacting helplines or
accessing treatment are identifying EGMs as their primary form of gambling
(Abbott et al, 2004c; Doiron & Mazer, 2001; Productivity Commission, 1999;
Smith & Wynne, 2004). In addition
to high intensity play and intermittent reinforcement, EGMs possess additional
highly addictive features including near misses, frequent small wins, the
possibility of large jackpots, non-availability of payout probabilities and
illusions of skill (Turner & Horbay, 2004).
Natural Recovery
Natural recovery refers to the process by
which individuals with maladaptive behaviors attain a state of recovery without
the help of a formal treatment program or self-help. In the case of problem gambling, the exact number of individuals
who recover on their own is unknown but is likely to be much higher than the
number of problem gamblers who access professional treatment (Abbott &
Volberg, 1996; Abbott, Williams & Volberg, 2004b; Smith, Volberg &
Wynne, 1994). Research has begun to
shed some light on natural recovery from pathological gambling.
Prospective studies of adolescents, college
students, casino employees and problem gamblers in the community have all found
high rates of “problem resolution” over periods ranging from one to seven years
(Abbott et al, 2004b; Hodgins & el-Guebaly, 2000; Shaffer & Hall, 2002;
Slutske, Jackson & Sher, 2003).
These studies challenge the notion enshrined in the DSM of pathological
gambling as a chronic and inevitably progressive disorder. The data further suggest that natural recovery
may be the rule rather than the exception, particularly among subclinical
problem gamblers.
The likelihood that natural recovery is
common among problem gamblers provides hope for effectively preventing gambling
disorders in the community (Abbott et al, 2004c). If problem gamblers’ behavior is as susceptible to change as these
few studies indicate, prevention messages could be targeted to specific groups
in the population most at-risk for progression to pathological gambling. It would also be possible to target specific
behaviors associated with progression towards more problematic gambling. Finally, given the relationship between
problem gambling and hazardous drinking, treatment initiatives are needed to
screen for gambling problems in alcohol treatment programs and either refer for
specialty gambling treatment or train providers in effective approaches to
treating gambling problems among substance abusers.
The survey of gambling and problem gambling in New
Mexico was completed in three stages.
In the first stage of the project, staff from Gemini Research consulted
with the Responsible Gaming Association of New Mexico as well as O’Neil
Associates, the organization responsible for data collection, regarding the
final design of the questionnaire and the sample. In the second stage of the project, staff from O’Neil Associates
translated and programmed the questionnaire and completed telephone interviews
with a sample of 3,596 residents of New Mexico aged 18 years and older. Data collection was carried out
between September 20, 2005 and January 12, 2006. O’Neil Associates then
provided Gemini Research with the data for the third stage of the project,
which included analysis of the data and preparation of this report.
The questionnaire included sections on
gambling participation, problem gambling, alcohol and drug use, experience of
depression and manic episodes, help-seeking, other impacts of gambling
including bankruptcy and involvement with the legal system, and demographics
(see Appendix B for a copy of the questionnaire).
Researchers in the field of gambling studies
recommend using more than one measure of problem gambling in surveys of the
general population (Abbott & Volberg, 1999b; Gambino, 1999; Shaffer, Hall
& Vander Bilt, 1997). Indeed,
Shaffer and his colleagues argue that the use of multiple problem gambling
screens should be one measure of the quality of problem gambling prevalence
studies. As noted above (see Measuring Gambling Problems on Page 4), several problem gambling screens based on the most
recent psychiatric criteria for pathological gambling have recently been
developed. The NORC DSM-IV Screen for
Gambling Problems (NODS) was used in the present survey to provide a measure of
problem gambling based on the most recent psychiatric criteria for pathological
gambling as well as comparability with recent national and statewide
surveys. The Problem Gambling Severity
Index (PGSI), developed in Canada for use in population studies of gambling
problems and impacts, was also used in the New Mexico a U.S. state-level survey
for the first time (see Appendix A for a comparison of the performance of these
two problem gambling screens).
Translation
of the Questionnaire
Census data show that 42% of the adult population
of New Mexico is Hispanic or Latino. To
enable interviews to be completed with Hispanic and Latino individuals who did
not speak English, it was necessary to translate the questionnaire. The questionnaire was translated into
Spanish by specialists at O’Neil Associates.
The translation process entailed one translator translating the
questionnaire from English into Spanish and a second translator translating the
questionnaire back from Spanish into English.
The original English version and the Spanish-to-English translation were
then compared. The two
translators discussed discrepancies between the two versions, including the
nuanced meaning of discrepant words and phrases, before reaching a consensus on
the Spanish translation’s final wording.
Interviewers were instructed to arrange to conduct
the interview in Spanish if the person answering the telephone spoke Spanish or
indicated that they wanted to complete the interview in that language. Four percent (N=114) of the interviews were
conducted in Spanish.
Pretest
The questionnaire was pre-tested with 15
randomly selected residents of New Mexico.
The main goal of the pretest was to test respondent comprehension and
the programming of the questionnaire. Respondents
had no difficulties comprehending the content of the questionnaire and
responding to items. The programming of
the questionnaire worked well and only a few minor changes were necessary prior
to fielding the full survey.
The main sample
for this survey included 3,007 residents of New Mexico aged 18 and over. Participants in the main sample were
selected by means of random-digit dialing (RDD), a method that ensures that
each telephone-owning household in New Mexico had an equal probability of
selection into the sample. This
sampling approach ensures that the overall sample is representative of New
Mexico residents within a known margin of error. The study also included an oversample of 589 Native American New
Mexico residents aged 18 and over.
These respondents were selected from a random sample of telephone
numbers likely to belong to a Native American household. Native American ethnicity was confirmed for
all respondents in the Native American oversample before conducting the interview.
All interviews were conducted at the O’Neil
Associates facility in Phoenix, Arizona by
trained interviewers under close supervision and with random monitoring for
technique and adherence to procedures. In
addition to general training in telephone interviewing techniques, interviewers
received training in the specific requirements for this study. Interviews were conducted using a computer-aided
telephone interviewing (CATI) system which minimizes the potential for
interviewer errors by controlling progression through the questionnaire and
preventing out-of-range responses.
Interviews
were conducted afternoons and evenings on weekdays and weekends. A minimum of eight attempts to establish
contact with each piece of sample was made, unless the interviewer received a
definitive refusal. If contact was made
with a household but an interview was not completed in the course of eight calls,
interviewers continued to make attempts to complete the interview during the
fieldwork period.
Sample Disposition and Response Rate
Table 2 on the following page presents
information about the disposition of the main sample and the Native American
oversample for the New Mexico prevalence survey. Table 2 shows that a total of 18,621 numbers were called over the
course of the data collection period.
At the end of this period, interviewers were able to determine that 4,565
of these numbers were not valid for the study, leaving 14,056 potentially
eligible numbers. Of these, 6,483 numbers
were persistently unavailable (i.e. numerous attempts were made without
reaching anyone or it was only possible to leave messages on an answering
machine or voice mail) and 416 were determined to be ineligible, leaving a
total of 7,752 households with which contact was made and eligibility was
determined. Of 7,752 screened households,
3,596 completed the interview.
Table
2: Disposition of New Mexico Sample
|
|
Main
Sample
|
Native American Sample
|
|
Total Numbers
|
14960
|
100.0
|
3661
|
100.0
|
|
|
|
|
|
|
|
Invalid Sample
|
3938
|
26.3
|
627
|
17.1
|
|
Not in Service (Disconnected)
|
3110
|
|
606
|
|
|
Non-Residential
|
825
|
|
18
|
|
|
Language Barrier - Non-Spanish
|
3
|
|
3
|
|
|
|
|
|
|
|
|
Total Non-Contacts or Ineligible
|
4646
|
31.1
|
2253
|
61.5
|
|
Answering Machine/Voice Mail
|
2770
|
|
595
|
|
|
No Answer
|
1537
|
|
560
|
|
|
Busy or Fast Busy Signal
|
339
|
|
87
|
|
|
Not Native American
|
N/A
|
|
416
|
|
|
|
|
|
|
|
|
Eligible Contacts
|
6376
|
42.6
|
1376
|
37.6
|
|
Completed Interview
|
3007
|
|
589
|
|
|
Callback Scheduled
|
244
|
|
201
|
|
|
Refused to Participate
|
3049
|
|
571
|
|
|
Partial Interview
|
9
|
|
6
|
|
|
Appointment
|
3
|
|
1
|
|
|
Other/Sick
|
64
|
|
8
|
|
There are a variety of ways to calculate
response rates. One definition is the
number of completed interviews divided by the number of units in the sample determined
to be eligible (i.e. the number of completes divided by the total of completes,
refusals, callbacks, partial interviews, and others). This approach is more properly termed the completion rate rather
than response rate. Based on this
approach, a completion rate of 47% was achieved in the main body of the New
Mexico prevalence survey. Another, more
conservative approach is to multiply the completion rate by the screening rate
(i.e. the proportion of numbers for which it was possible to determine
eligibility). Using this approach, the
main body of the New Mexico survey achieved a 37% cooperation rate.
Response rates for telephone surveys in the
general population have declined precipitously in recent years as individuals
in the general population become increasingly reluctant to participate in this
type of research and as technological barriers proliferate (e.g. answering
machines, caller id). Given these
declines, the completion and cooperation rates achieved in this survey are excellent
compared with similar surveys.
The
data from the main survey were weighted with regard to gender, ethnicity and
age. The sample weights were derived from 2000 U.S. census data (Census Table
DP-1). Since the demographic profile of
the respondents in the main survey differed from that of the 2000 New Mexico
census, the weights were designed to match the sample demographics to that of
the census.
The
sample weights were algebraically derived by solving equations for unknown
values. In some cases, such as gender,
these initial weights remained intact.
However, the initial weights for age and ethnicity were adjusted via
sensitivity analysis to minimize the variance between the achieved sample
demographics and the population parameters from the census. At the group level, twelve unique weights—two
for gender (male and female), three for ethnicity (Caucasian, Hispanic and
Other) and two for age (18-34 and 35+)—were ultimately produced. The twelve unique weights were used to
describe all of the cases in the main survey sample. For each case, the corresponding gender, age and ethnicity
weights were multiplied to produce the case weight so that the weight applied
to a 37-year old, Hispanic female was different from the weight applied to a
25-year old, Caucasian male.
Since
gender, age, and ethnicity variables were employed in the sample weighting
scheme, a response was required in each of these categories for a case to be
included in the data analysis. No
response in any of these categories would result in a case weight of zero,
effectively removing the case from any data analysis. In a preliminary data screening process, it was determined that
231 cases were missing a valid age response and 157 cases were missing a valid
response to the ethnic origin item.
Several
attempts were made to predict age by producing various multiple regression
equations from the existing data. However,
none of these attempts was deemed successful. Consequently, the mean and standard deviation of the age variable
were used to create imputed age values for the 231 missing cases. Once the missing age values were replaced,
157 cases with missing ethnicity responses remained. Ethnicity was a nominal rather than continuous variable and there
were fewer effective options for estimating missing values. The research team elected to omit these 157
cases from the analysis. This decision
reduced the number of valid cases in the main sample to 2,850.
The
weighted sample results were produced by multiplying the original sample cases
by derived case weights. Despite some
considerable differences between the demographic profile of the unweighted
sample and that of the 2000 census, the use of unique weights made it
relatively easy to achieve a sample demography that was nearly identical to
that of the 2000 New Mexico census. Table
3 compares the demographics of the achieved sample to those of the 2000 census
and the weighted sample.
Table
3: Demographics of Achieved and Weighted Samples
|
|
|
Achieved Sample
%
|
2000
Census
%
|
Weighted Sample
%
|
|
|
|
|
|
|
|
Gender
|
|
|
|
|
|
|
Male
|
39.6
|
49.2
|
48.9
|
|
|
Female
|
60.4
|
50.8
|
51.1
|
|
|
|
|
|
|
|
Age
|
|
|
|
|
|
|
18 – 34
|
16.4
|
31.9
|
31.2
|
|
|
35 and over
|
83.6
|
68.1
|
68.8
|
|
|
|
|
|
|
|
Ethnicity
|
|
|
|
|
|
|
White
|
65.1
|
44.7
|
44.1
|
|
|
Hispanic
|
29.4
|
42.1
|
42.0
|
|
|
Other
|
5.5
|
13.2
|
13.9
|
Once
the data were delivered to Gemini Research, all of the variables were checked
carefully for correct skip procedures.
The data were analyzed using the Statistical Package for the Social
Sciences (SPSS 13.0). Numerous analytic
variables were constructed from the raw data, including generalized gambling
participation levels, scores on the problem gambling screens, levels of alcohol
and drug use, experience of depression and mania, and help-seeking. Chi-square analysis and other nonparametric
techniques were used to test for statistical significance.
In
the three sections of the report that follow, we present information on the
results of the main sample separately from the results of the Native American oversample. There are two reasons for this
approach. First, as noted above, the
sampling frames for the main sample and oversample were somewhat different,
with the main sample consisting of a random probability sample of New Mexico
households with telephones and the oversample comprising a list-assisted
sample. Second, given the prominence of
Native American gambling issues in New Mexico, it seemed appropriate to present
the results from our Native American respondents separately from those of the
main sample.
As
noted above, the majority of data analyses were carried out using SPSS 13.0. Minitab was used in our analyses comparing
data from the Native American respondents with data from the general population
(see Comparing Native American and
Non-Native Americans in New Mexico on Page 46). The general
population and Native American data were delivered in two separate SPSS files. However, there were 193 Native American
respondents in the weighted general population data set. The research team elected to remove these
respondents from the main data set and add them to the Native American data set. To protect the integrity of the data, the
research team further elected to maintain the files separately.
To test
for statistically significant differences in response frequencies across the two
samples, crosstabulations were first produced in the general population file. A categorical variable identifying the Native
Americans in this file isolated their responses. Once categorized and isolated, these responses were added to the
responses on the same items in the Native American data file. As the counts from the general population
file were simply added to the counts from the Native American data file, the
data were now at the summary level. Since
SPSS will not compute chi-square tests on summary level data, Minitab was
employed for these analyses.
A
separate issue relates to the case weights when comparing data across the two
samples. It was determined that the
general population data should remain weighted when comparing results against
the Native American sample. As a
result, weighted responses from the general population (i.e. non-Native
American respondents) were compared to the pooled responses from all of the
Native American respondents. These
pooled data consisted of weighted responses from the Native American cases in
the general population file and unweighted responses from the cases in the
Native American data file.
There
was also the issue of the case weights when comparing data across the two
samples. It was determined that the data from the general population file
should remain weighted when comparing results against the Native American
sample. As a result, weighted responses from the general population (i.e,
non-Native Americans) were compared to the pooled responses from the Native
Americans. These pooled data consisted of weighted responses from the Native
American cases from the general population file and unweighted responses from
the cases in the Native American data file.
This chapter examines gambling participation among adults in
New Mexico. To assess the full range of
gambling activities available to New Mexico residents, the instrument for the
survey included questions about ten different wagering activities. All respondents were asked if they had ever
gambled or bet money on the following activities:
·
casino games
·
gaming machines outside of a casino
·
lottery games
·
numbers games other than the New Mexico State Lottery
·
horse or dog races
·
bingo outside of a casino
·
private games
(cards, dice or dominoes in someone’s home or at a club or organization, or a
game of skill such as golf, pool or bowling)
·
the outcome of sports or other events with friends,
co-workers, a bookie or some other person
·
Internet or World Wide Web
·
any other kind
of gambling (e.g. raffles, sweepstakes, baby pools, pull-tabs, betting on a
dogfight or cockfight)
In every recent
survey of gambling and problem gambling, the majority of respondents
acknowledge participating in one or more gambling activities. Nationally, the proportion of the population
that has ever gambled ranges from 81% in the Southern states to 89% in the
Northeast (Gerstein et al, 1999). In
2005, 85% of the New Mexico respondents acknowledged ever participating in one
or more of the ten activities included in the questionnaire.
Table 4 on the
following page shows lifetime, past-year, monthly and weekly participation for
all of the types of gambling included in the New Mexico survey. Lifetime participation among New Mexico adults
was highest for casino gambling and lottery play. Just over six in ten New Mexico adults acknowledge having ever
been to a casino or played the lottery.
Two in five New Mexico adults has bet on horse or dog races and one in
four New Mexico adults has gambled privately or bet on sports. One in five New Mexico adults has ever
played non-casino bingo or non-casino gaming machines.
Past-year
participation rates among New Mexico adults were highest, again, for lottery
play and then casino gambling. About
one in six New Mexico adults acknowledge gambling in the past year on sports or
on a private game of chance or skill.
Past-year participation in all other activities is much lower. The majority of monthly and weekly gambling
participation among New Mexico adults is explained by lottery play and casino
gambling.
Table
4: Gambling Participation
in New Mexico
|
|
Lifetime
Participation
(2850)
%
|
Past Year
Participation
(2850)
%
|
Monthly
Participation
(2850)
%
|
Weekly
Participation
(2850)
%
|
|
|
|
|
|
|
|
Lottery
|
64.0
|
51.7
|
14.1
|
5.8
|
|
Casino
|
64.0
|
36.6
|
3.9
|
1.4
|
|
Pari-mutuel
|
29.9
|
7.0
|
0.4
|
0.1
|
|
Sports
|
25.7
|
14.7
|
2.9
|
1.0
|
|
Private
|
24.9
|
13.3
|
2.3
|
0.9
|
|
Non-casino bingo
|
17.6
|
5.5
|
0.8
|
0.2
|
|
Non-casino gaming machines
|
17.6
|
5.5
|
0.4
|
0.2
|
|
Other
|
16.0
|
9.4
|
1.0
|
0.3
|
|
Non-lottery numbers
|
3.1
|
1.3
|
0.1
|
0.1
|
|
Internet
|
1.8
|
1.4
|
0.8
|
0.5
|
|
|
|
|
|
|
|
Total
|
85.0
|
67.6
|
19.8
|
8.6
|
Nearly one-fifth (17.9%) of the respondents in the New
Mexico survey only acknowledge having gambled on one activity in their
lifetime. The majority of these
respondents (N=374) are casino and lottery players. Over half of these respondents (57%) have played the lottery and
42% have been to a casino. Much smaller
percentages of this group (between 3% and 7%) have gambled on private games,
sports, horseracing, non-casino machines or “other” activities.
Endorsement of
the usually residual “Other” category was higher in this survey than in some other
gambling surveys. Respondents who said
that they had done some other type of gambling in the past year were somewhat more
likely than those who did not endorse this item to be female and to be employed
fulltime. These respondents were
significantly more likely than those who did not gamble on “other” activities
to be between the ages of 45 and 64, White, to have attended college and to
have annual household incomes over $50,000.
This suggests that endorsement of participation in these activities is probably
more closely related to charitable gambling than to illegal or
culturally-specific activities such as cockfighting.
To understand patterns of gambling participation,
it is helpful to examine the demographics of respondents who wager at
increasing levels of frequency. To analyze
levels of gambling participation, respondents were divided into five groups:
·
non-gamblers who have never participated in any type of
gambling (15% of the total sample);
·
infrequent
gamblers who have
participated in one or more types of gambling but not in the past year (17% of
the total sample);
·
past year
gamblers who have
participated in one or more types of gambling in the past year but not on a
weekly basis (48% of the total sample); and
·
monthly
gamblers who participate in
one or more types of gambling on a monthly basis (11% of the total sample).
·
weekly
gamblers who participate in
one or more types of gambling on a weekly basis (9% of the total sample).
Table 5 presents information about the demographic
characteristics of these different groups in New Mexico. For easier comprehension, non-gamblers and
infrequent gamblers have been collapsed into a single group, as have monthly
and weekly gamblers.
There are some important differences between
non- and infrequent gamblers in New Mexico.
Non-gamblers are significantly more likely than infrequent gamblers to
be under 35, Hispanic, keeping house and to have an annual household income
under $25,000. Non-gamblers in New
Mexico are significantly less likely than infrequent gamblers to have attended
college and to have military experience.
The only significant difference between monthly and weekly gamblers in New
Mexico is that weekly gamblers are more likely than monthly gamblers to have
graduated from college.
Table
5: Demographics
of Gamblers in New Mexico
|
|
|
Non- &
Infrequent Gamblers
(923)
%
|
Past Year Gamblers
(1363)
%
|
Monthly & Weekly Gamblers
(564)
%
|
Sig.
|
|
|
|
|
|
|
|
|
Gender
|
Male
|
43.3
|
48.4
|
59.2
|
|
|
|
Female
|
56.7
|
51.6
|
40.8
|
.000
|
|
|
|
|
|
|
|
|
Age
|
18 – 34
|
28.9
|
33.8
|
30.0
|
|
|
|
35 – 54
|
26.9
|
32.8
|
37.2
|
.000
|
|
|
55+
|
44.2
|
33.5
|
32.8
|
|
|
|
|
|
|
|
|
|
Ethnicity
|
White
|
43.4
|
45.9
|
41.2
|
|
|
|
Hispanic
|
41.5
|
39.5
|
48.7
|
.002
|
|
|
Other*
|
15.1
|
14.6
|
10.1
|
|
|
|
|
|
|
|
|
|
Marital Status
|
Married
|
62.9
|
62.8
|
65.2
|
|
|
|
Widowed
|
11.0
|
5.3
|
5.4
|
.000
|
|
|
Divorced/Separated
|
11.9
|
12.7
|
12.1
|
|
|
|
Never Married
|
14.3
|
19.1
|
17.3
|
|
|
|
|
|
|
|
|
Education
|
Elementary / Some HS
|
18.9
|
7.0
|
10.2
|
|
|
|
HS Grad
|
24.2
|
26.9
|
33.6
|
.000
|
|
|
Some College
|
26.4
|
32.9
|
30.6
|
|
|
|
BA Degree
|
15.5
|
19.5
|
17.2
|
|
|
|
Graduate Study
|
15.0
|
13.6
|
8.4
|
|
* Includes Native American, African American and Other.
Table
5: Demographics of Gamblers in New Mexico (cont’d)
|
|
Non- &
Infrequent Gamblers
(923)
%
|
Past Year Gamblers
(1363)
%
|
Monthly & Weekly Gamblers
(564)
%
|
Sig.
|
|
|
|
|
|
|
|
|
Employment
|
Working Full Time
|
37.7
|
55.6
|
55.9
|
.000
|
|
|
Working Part Time
|
12.0
|
11.2
|
10.8
|
|
|
|
Keeping House
|
14.3
|
6.7
|
5.7
|
|
|
|
Retired
|
22.5
|
16.1
|
17.6
|
|
|
|
Disabled / Unemployed
|
6.2
|
4.8
|
6.2
|
|
|
|
|
|
|
|
|
Income
|
Up to $25,000
|
21.6
|
18.3
|
15.7
|
.000
|
|
|
$25,001 - $35,000
|
11.7
|
10.5
|
13.0
|
|
|
|
$35,001 - $50,000
|
9.5
|
15.2
|
11.2
|
|
|
|
$50,001 - $75,000
|
11.7
|
16.1
|
17.7
|
|
|
|
$75,001 - $125,000
|
8.3
|
13.0
|
19.6
|
|
|
|
Over $125,000
|
3.2
|
4.8
|
6.3
|
|
|
|
Refused
|
34.0
|
22.1
|
16.5
|
|
|
|
|
|
|
|
|
|
Religion
|
Fundamentalist/Christian
|
24.6
|
23.3
|
22.8
|
.000
|
|
|
Protestant
|
28.3
|
22.8
|
20.1
|
|
|
|
Catholic
|
31.0
|
37.7
|
44.1
|
|
|
|
Other
|
11.8
|
11.1
|
9.4
|
|
|
|
None
|
4.3
|
4.9
|
3.6
|
|
|
|
|
|
|
|
|
Armed Forces Service
|
15.9
|
17.3
|
24.5
|
.000
|
|
|
|
|
|
|
.000
|
|
Interviewed in Spanish
|
11.5
|
2.6
|
2.0
|
|
Overall, Table 5 shows that significant
differences in gambling participation are associated with gender, age, ethnicity,
marital status, education and employment status. Important differences in gambling participation are also
associated with income, religion and military experience. Non- and infrequent gamblers are
significantly more likely than past-year, monthly and weekly gamblers in New
Mexico to be female, aged 55 and over, widowed, to have less than a high school
education and to be retired or keeping house.
Non- and infrequent gamblers are also significantly more likely than
more frequent gamblers to have refused to provide information about their
annual household income and to have been interviewed in Spanish.
Monthly and weekly gamblers are significantly
more likely than past-year gamblers to be male, to be Hispanic, to have an annual
household income over $50,000 and to have military experience. Monthly and weekly gamblers are
significantly less likely than past-year gamblers to have attended college.
For several types of gambling, respondents who acknowledged
participating in the past year were asked about their preferences for
particular games. These types of gambling included lottery, casino, pari-mutuel
and non-casino gaming machines.
Lottery. Respondents who had played the
lottery in the past year (N=1473) were asked what kinds of tickets they usually
purchased. Respondents were permitted
multiple answers to this question. The
most popular lottery games in New Mexico are Powerball and Scratchers. Eight in ten of these respondents (80%)
reported that they usually bought Powerball tickets and another 27% said that
they usually bought instant tickets, or Scratchers. Two-thirds (67%) of these respondents reported that they usually
only bought Powerball tickets, 14% reported only buying Scratchers and 13%
reported that they usually bought one or the other of these lottery
products. Only 6% of these respondents
reported that they usually bought other kinds of lottery tickets.
Casino. Respondents who had gambled at a
casino in the past year (N=1044) were asked what casino game they usually
played. The majority (74%) said that
they usually played slot machines or video poker at the casino. Another 19% said that they usually played
card games such as blackjack or poker. Only
7% of these respondents indicated that they usually gambled on anything besides
card games or machines at the casino.
These respondents were also asked what city or location they
visited on the last occasion when they went to a casino. Four out of five of these respondents (80%)
indicated that their last visit was to a casino in New Mexico while 20%
indicated that their last visit was to a casino outside of New Mexico. Among respondents whose last visit was to a
casino in New Mexico, 88% indicated that this was a tribal casino and 4% were
not sure if the casino were tribally owned or not. Among respondents who last visit to a casino was outside New
Mexico, 78% indicated that the casino was located in Nevada, 16% said they
visited a casino in Arizona or Colorado and 5% visited a casino even further
afield.
Pari-mutuel. Respondents who had wagered on horse or dog
races in the past year (N=187) were asked whether they usually did so at a
racetrack in New Mexico, an off-track betting facility in New Mexico, a tribal
casino or somewhere else. Eight in ten of
these respondents (83%) indicated that they usually wagered at a racetrack in
New Mexico. Another 4% of these
respondents said that they usually wagered at an off-track betting facility in New
Mexico. The small group of remaining
pari-mutuel gamblers usually wagered at an off-track facility outside New
Mexico.
Gaming Machines. Respondents who had wagered on gaming
machines outside a casino in the past year (N=153) were asked where they
usually played these machines. Nearly
one-quarter of these respondents (23%) indicated that they usually played
gaming machines at a racetrack. Only
about one in eight of these respondents (12%) played gaming machines at social
or fraternal organizations. Other
places where respondents said that they usually played gaming machines included
bars, taverns or restaurants (15%) and grocery and convenience stores
(20%).
Favorite Gambling Activities
Table 6 on the following page presents information about
favorite gambling activities among infrequent, past-year, monthly and weekly
gamblers. Questions about preferred
gambling activities were only asked of respondents who indicated that they had
gambled five or more times in their lifetime.
If an individual acknowledged gambling once a month or more often on any
of the activities included in the questionnaire, this variable was
automatically coded “Yes.” If a person
had ever gambled or had gambled in the past year but said “No” to this
question, this variable was coded “No” and these items were not asked.
Table 6 shows that electronic gaming machines were the
preferred gambling activity across all of these groups. Infrequent gamblers were significantly more
likely to say that they had no favorite gambling activity than other gamblers
and significantly less likely than more frequent gamblers to identify the
lottery as their favorite gambling activity.
Table
6: Favorite Gambling
Activities Among New Mexico Gamblers
|
|
Infrequent Gamblers
(93)
%
|
Past Year Gamblers
(736)
%
|
Monthly Gamblers
(319)
%
|
Weekly
Gamblers
(244)
%
|
Sig.
|
|
|
|
|
|
|
.000
|
|
Slot machines (casino & non)
|
19.1
|
32.6
|
28.5
|
26.6
|
|
|
Casino table games
|
18.1
|
15.8
|
12.9
|
17.6
|
|
|
Lottery
|
3.2
|
14.9
|
19.7
|
18.0
|
|
|
Private or sports
|
14.9
|
12.8
|
20.4
|
15.6
|
|
|
Pari-mutuel
|
8.5
|
4.2
|
3.4
|
6.1
|
|
|
Bingo
|
4.3
|
2.0
|
3.4
|
2.5
|
|
|
Other/None
|
31.9
|
17.8
|
11.6
|
13.5
|
|
Another important question in gambling studies is why people
choose whether or not to gamble. Respondents
who had gambled five or more times in their lifetime were asked why they
generally gambled, and to indicate whether any of several different reasons was
“very important,” “somewhat important,” or “not at all important.” Table 7 presents information on the
proportion of respondents who indicated that each of these reasons was “very
important” or “somewhat important.”
Table
7: Reasons for Gambling
Among New Mexico Gamblers
|
Somewhat or very important
|
Infrequent Gamblers
(93)
%
|
Past Year Gamblers
(736)
%
|
Monthly Gamblers
(319)
%
|
Weekly
Gamblers
(244)
%
|
Sig.
|
|
|
|
|
|
|
|
|
Entertainment or fun
|
69.9
|
79.9
|
81.8
|
82.0
|
.025
|
|
To win money
|
46.2
|
57.9
|
65.8
|
63.9
|
.003
|
|
Excitement or challenge
|
46.2
|
55.5
|
56.4
|
60.2
|
.150
|
|
To be with people
|
41.9
|
46.7
|
44.8
|
49.2
|
.606
|
|
Convenience
|
31.2
|
40.9
|
50.0
|
50.0
|
.001
|
|
Inexpensive entertainment
|
28.0
|
50.0
|
49.5
|
50.8
|
.001
|
|
As a distraction
|
10.8
|
19.7
|
24.8
|
19.6
|
.046
|
Table 7 shows that the majority of New Mexicans gamble for
entertainment although infrequent gamblers are significantly less likely to
endorse this reason than more frequent gamblers. As gambling participation increases, winning money becomes an
increasingly important reason for gambling as does excitement or challenge,
inexpensive entertainment and convenience.
The importance of gambling in order to be with people is not significantly
different for these different groups of gamblers. However, infrequent gamblers are significantly less likely than
more frequent gamblers to say that distraction is a somewhat or very important
reason for gambling.
Given differences in gambling participation by gender, age
and ethnicity, differences in reasons for gambling associated with these
important demographic variables were examined.
The only difference between men and women was that men were
significantly more likely to say that they gamble because it is exciting and
challenging and because it is easy and convenient to do. Respondents under the age of 35 were
significantly more likely than older respondents to say that winning money,
excitement and being around or with other people were important reasons for
gambling. White respondents were
significantly more likely than respondents from other ethnic groups to say that
they gambled for entertainment or fun and significantly less likely to say that
they gambled to distract themselves from everyday problems. Hispanic respondents were significantly less
likely than other respondents to say that they gambled because it was
inexpensive entertainment.
In the New Mexico survey, respondents who had never gambled
or gambled infrequently
were asked whether any of several different reasons to not gamble was “very
important,” “somewhat important” or “not at all important.” Losing money was the most important reason
for not gambling among these respondents, followed by moral or ethical
concerns. Women in this group were
significantly more likely than men to say that moral or ethical concerns, the
possibility of losing money and inconvenience were all important reasons that
they did not gamble. Hispanic
respondents were significantly more likely to say that losing money and inconvenience
were important reasons not to gamble while White respondents were significantly
more likely to say that moral and ethical concerns were important reasons not
to gamble. There were no significant
differences in reasons for not gambling by age.
Two problem gambling screens were used in the New Mexico
survey. The NORC DSM-IV Screen for
Gambling Problems (NODS) was used to provide a measure of problem gambling
based on the most recent psychiatric criteria for pathological gambling as well
as comparability with recent national and statewide surveys. The Problem Gambling Severity Index (PGSI)
from the recently developed Canadian Problem Gambling Index (Ferris &
Wynne, 2001) was used in New Mexico as a secondary measure of gambling-related
impacts and to provide a first opportunity to compare the performance of these
two problem gambling screens in a single survey.
In 1998 the
National Gambling Impact Study Commission contracted the National Opinion
Research Center (NORC) and partner organisations to undertake a national survey
of problem gambling in the United States.
The Commission specified the use of DSM-IV criteria in this survey which
meant that neither the SOGS nor any of its variants could not be used. After reviewing the available DSM-IV
screens, the research team elected to develop a new measure designed
specifically for administration in large population surveys. This instrument has 17 lifetime and 17 past-year
items. Several items are only
administered if a preliminary screening question is endorsed and past-year
items are only administered if the corresponding lifetime item is
endorsed. Each criterion item is scored
zero or one, to produce maximum scores of ten for each of the “lifetime” and “current”
frames. Scores of zero were interpreted
as indicating low risk, one or two at risk, three to four problem gambling, and
five or more pathological gambling.
One important step in developing the NODS was a validation
study with a national clinical sample of 40 individuals enrolled in outpatient
problem gambling treatment programs and an additional random telephone sample
of 45 respondents in the Chicago metropolitan area. Ninety-five percent of the clinical sample scored five or more
points on the lifetime NODS; the remaining two cases scored four points. The test-retest reliability of the NODS was
examined in a half-sample of 44 cases drawn equally from the clinical and telephone
pilot samples. The lifetime and
past-year scores on the NODS were found to be highly reliable (r=0.99 and 0.98, respectively) (Gerstein
et al, 1999). Based on the field test,
the research team concluded that the NODS had strong internal consistency,
retest reliability and good validity.
In addition
to the U.S. national survey the NODS has been used in several state level
prevalence surveys and an older persons study in the U.S. (Shapira et al, 2002; Volberg, 2001a, 2001b, 2001c,
2002, 2003a; Volberg & McNeilly, 2003).
It has also been used in a Norwegian national survey (Lund & Nordlund,
2003) and in a Spanish provincial study (Becońa, 2004). The NODS is increasingly being used in North
American clinical settings as an assessment and outcome measure (Hodgins, 2002,
2004) as well as in research studies of problem gamblers in the community (Sartor
et al, in press; Scherrer et al, 2005) and its use is mandatory for all clients
entering drug and alcohol treatment programs in Michigan (Herriff, personal
communication). In this section
of the report and the two that follow, the lifetime NODS serves as the primary
measure of at-risk, problem and pathological gambling in New Mexico.
In epidemiological
research, prevalence is a measure of the number of individuals in the population with a disorder
at one point in time. In epidemiology,
prevalence contrasts with incidence which is a measure of the number of new
cases that arise over a specific period of time. In problem gambling prevalence surveys, individuals are classified as at-risk, problem
or pathological
gamblers on the basis of their responses to a previously established number of
items from a valid and reliable problem gambling screen.
Prevalence rates are based on samples rather than the entire
population. One important source of
uncertainty in generalizing from a sample to the population—sampling error—is
generally presented as a measure of the uncertainty around the identified
value. Calculations of the size of this
variation—sometimes called the confidence interval and sometimes referred to as
the margin of error—are based on the percentage of the sample with a particular
characteristic and the size of the sample.
To illustrate, the margin of error for the main sample of
respondents in New Mexico (N=2,850) is ±1.8%. The margin of error for an entire sample is
generally calculated for a situation in which half of the respondents answer a
question “Yes” and the other half answer “No.”
The confidence interval allows us to assume with reasonable certainty—95
times out of 100—that the “true” value is somewhere between 48.2% and 51.8%.
The confidence interval narrows as the value approaches
either 0% or 100%. For example, a value
of 5% in the New Mexico survey has a margin of error of ±0.8%. This means that we
can be reasonably certain that the “true” value falls between 4.2% and
5.8%. As values near these extremes,
the confidence interval can approach or exceed the value itself. The closer the confidence interval comes to
the value, the less reliable the value itself is considered to be. In several of the tables that follow,
confidence intervals that equal or exceed 50% of the value of the prevalence
estimate are flagged with an asterisk and readers are advised to treat these
estimates with caution.
Table 8 on the
following page presents information about the proportion of the main sample
(N=2,850) who scored on an increasing number of items on the lifetime and
past-year NODS. Table 8 also summarizes
the prevalence of problem and pathological gambling based on established
criteria for discriminating between respondents without gambling-related
difficulties and those with moderate and severe problems (Gerstein et al, 1999;
Toce-Gerstein, Gerstein & Volberg, 2003).
Table
8: Scores
on Lifetime and Past Year NODS
|
Number of Items
|
Lifetime
|
Past
Year
|
|
|
(2850)
|
(2850)
|
|
|
|
|
|
Non-Gamblers
|
15.0
|
32.4
|
|
0
|
76.5
|
62.9
|
|
Non Problem Gamblers
|
76.5
|
62.9
|
|
1
|
4.6
|
2.4
|
|
2
|
1.8
|
1.2
|
|
At-Risk Gamblers
|
6.4
|
3.6
|
|
3
|
0.7
|
0.4
|
|
4
|
0.4
|
0.3
|
|
Problem
|
1.1
|
0.7
|
|
5
|
0.4
|
0.2
|
|
6
|
0.4
|
0.2
|
|
7
|
0.1
|
0.1
|
|
8
|
0.1
|
0.0
|
|
9
|
0.0
|
0.0
|
|
10
|
0.1
|
0.1
|
|
Pathological
|
1.1
|
0.6
|
|
|
|
|
|
Combined Problem/Path
|
2.2
|
1.3
|
According to the most
recent census of the population (U.S. Bureau of the Census, 2001), the
population of New Mexico aged 18 and over in 2000 was 1,310,472. Based on these figures, we estimate that
between 9,400 (0.7%) and 19,400 (1.5%) New Mexico adults can be classified as pathological
gamblers.
Another 9,400 (0.7%) to 19,400
(1.5%) New Mexico adults can be classified as problem gamblers. Finally, an additional 72,100 (5.5%) to 95,600
(7.3%) New Mexico adults can be classified as at-risk gamblers.
Prevalence Across Demographic Groups
Problem gambling prevalence rates can be significantly different among
subgroups in the population. Because
the confidence intervals around prevalence estimates can be large, most
comparisons between these groups must be interpreted with caution. However, the size of the main sample in New
Mexico means that, in this instance, confidence intervals exceed 50% of the
variance for relatively few of the prevalence estimates for subgroups in the
population. In presenting these data,
all instances where the confidence interval equals or exceeds the prevalence
estimate have been suppressed. Table 9 on
the following page presents information about the size of each group as well as
the confidence interval for the combined problem and pathological gambling
prevalence rate.
Table 9: Differences
in Prevalence by Demographic Group
|
|
|
Group
Size
|
Prevalence Rate
(3+)
|
Conf.
Interval
|
|
|
|
|
|
|
|
Gender
|
Male
|
1393
|
2.6
|
±0.8
|
|
.
|
Female
|
1457
|
1.6
|
±0.6
|
|
|
|
|
|
|
|
Age
|
18 – 34
|
859
|
2.8
|
±1.1
|
|
.
|
35 – 54
|
868
|
2.3
|
±1.0
|
|
|
55+
|
1002
|
1.3*
|
±0.7
|
|
|
|
|
|
|
|
Ethnicity
|
White
|
1257
|
1.4
|
±0.6
|
|
.
|
Hispanic
|
1196
|
2.8
|
±0.9
|
|
|
Other**
|
396
|
2.3*
|
±1.5
|
|
|
|
|
|
|
|
Marital Status
|
Married
|
1786
|
1.4
|
±0.5
|
|
|
Never Married
|
486
|
4.5
|
±1.8
|
|
|
|
|
|
|
|
Education
|
HS Graduate
|
768
|
3.4
|
±1.3
|
|
|
Some college
|
852
|
2.9
|
±1.1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Employment
|
Fulltime
|
1421
|
1.9
|
±0.7
|
|
|
Disabled / Unemployed
|
157
|
8.3
|
±4.3
|
|
|
|
|
|
|
|
Religion
|
Fundamentalist
|
621
|
2.9
|
±1.3
|
|
|
Catholic
|
966
|
2.6
|
±1.0
|
* Confidence
interval equals or exceeds 50% of the prevalence estimate.
**
Includes Native American, African American and Other.
Table 9 shows that there are substantial differences in the
prevalence of problem gambling across different subgroups in the population in New
Mexico. Differences in prevalence rates
by gender, ethnicity, marital status, education, employment status and religion
are all statistically significant. The
prevalence of problem and pathological gambling in New Mexico is significantly
higher among men, among non-Caucasians, among respondents who have never
married, among respondents who are disabled or unemployed and among respondents
who are fundamentalist Christians or Catholic.
Differences in prevalence rates by age, household income and military
service are not statistically significant.
Another approach
to understanding the relationship between gambling involvement and
gambling-related problems is to examine the prevalence of problem gambling
among individuals who participate in specific types of gambling. Table 10 on
the following page shows the prevalence of problem and pathological
gambling among respondents who have ever gambled, among those who have gambled
in the past year and among those who gamble monthly and weekly. Table 10 also shows the prevalence of
problem and pathological gambling among respondents who have participated in
specific types of gambling in the past year.
All results where the confidence interval exceeds 50% of the prevalence
estimate have been flagged with an asterisk. As in Table 9, all instances where the confidence interval equals
or exceeds the prevalence estimate have been suppressed. This includes weekly gamblers as well as
past year gambling on the Internet and gambling on non-lottery numbers
games.
Table 10:
Prevalence by Type of Gambling
|
|
Group
Size
|
Prevalence
(3+)
%
|
Conf.
Interval
|
|
All Gamblers
|
2422
|
2.5
|
±0.6
|
|
Past-Year Gamblers
|
1926
|
3.0
|
±0.8
|
|
Monthly
Gamblers
|
563
|
6.2
|
±2.0
|
|
Among Past Year Players
|
|
|
|
|
Non-Casino Bingo
|
157
|
8.9
|
±4.5
|
|
Private
|
378
|
6.1
|
±2.4
|
|
Sports
|
418
|
5.0
|
±2.1
|
|
Casino
|
1044
|
4.5
|
±1.3
|
|
Other
|
268
|
3.7*
|
±2.3
|
|
Pari-mutuel
|
200
|
3.5*
|
±2.6
|
|
Lottery
|
1473
|
3.3
|
±0.9
|
*Confidence
interval equals or exceeds 50% of the prevalence estimate.
Table 10 shows that problem gambling prevalence rates increase
along with gambling participation. Although
the group is quite small, the prevalence of problem gambling is highest among
past-year players of non-casino bingo.
Problem gambling prevalence rates are also high among past-year private
bettors, past-year sports bettors and past-year casino gamblers. Prevalence rates among these groups of
past-year players are more than twice as high as the problem gambling
prevalence rate in the population as a whole.
As with gambling participation, it is helpful to compare the
prevalence of problem and pathological gambling in New Mexico with comparable
prevalence estimates elsewhere in the United States. Although the jurisdictions
where problem gambling surveys have been done in the United States differ
substantially in the types of gambling available, in levels of gambling
participation and in the demographic characteristics of the general population,
it is helpful to understand how New Mexico compares with other
jurisdictions.
Figure 3 on the following page presents lifetime NODS
prevalence rates for states where similar surveys have been conducted in the
United States. Overall, Figure 3 shows
that the prevalence of at-risk, problem and pathological gambling in New Mexico
is at the lower end of a range of problem gambling prevalence rates based on
the same problem gambling screen. The
prevalence of at-risk, problem and pathological gambling in New Mexico is
somewhat higher than the prevalence rate obtained in North Dakota in 2000 but
lower than prevalence rates identified in Oregon in 2000 and the United States
as a whole in 1998. The prevalence of
at-risk, problem and pathological gambling in New Mexico is well below the
prevalence rates identified in recent surveys carried out in Arizona and Nevada
(Volberg, 2002, 2003a).
Figure
3: Comparing NODS
Rates Across States (Lifetime)
The comparison between New Mexico and Arizona is
particularly interesting, not only because the two states are contiguous but
also because their demographics are quite similar with large Hispanic and
Native American populations. Although
the population of Arizona is nearly three times greater than the population of
New Mexico, the two states have nearly identical levels of educational
attainment and workforce participation.
The median household income in Arizona is only slightly higher than in
New Mexico (U.S. Census, 2001). There
are similar numbers of tribal casinos and racetracks in both states as well as
mature state lotteries. One important
difference between the two states is that neither racetracks nor veterans and
fraternal clubs in Arizona are permitted to operate gaming machines.
The finding that New Mexico has approximately half the rate
of at-risk gambling as Arizona and twice the rate of pathological gambling at a
point in time when the duration of exposure to expanded gambling opportunities
is about the same in the two states suggests several intriguing possibilities. Perhaps the rate of at-risk gambling in New
Mexico is increasing and will eventually rise to the level identified in
Arizona. Another possibility is that
the rate of at-risk gambling in Arizona is decreasing and will eventually reach
the level identified in New Mexico.
With regard to pathological gambling, it is possible that pathological
gambling in New Mexico is decreasing and will eventually drop to the level in
Arizona. Alternatively, it is possible
that pathological gambling in Arizona is increasing and will eventually reach
the level identified in New Mexico.
Another possibility is that there are moderating factors that affect the
prevalence of at-risk, problem and pathological gambling in ways that are not
yet recognized or understood. Future
surveys of gambling and problem gambling in these two states would help test these
and other hypotheses, as would prospective, longitudinal research on the
development of gambling problems within individuals over time.
In considering how best to develop and refine
policies and programs for problem gamblers, it is important to direct these
efforts in an effective and efficient way.
The most effective efforts at prevention, outreach and treatment are
targeted at individuals who are at greatest risk of experiencing gambling-related
difficulties. Since the purpose of this
section is to examine vulnerable individuals, our focus will be on differences
between individuals who gamble, with and without problems, rather than on the
entire New Mexico sample.
As noted above, the lifetime NODS serves as
the primary measure of at-risk, problem and pathological gambling in New
Mexico. In this section of the report,
we examine differences between groups of respondents who score at increasing
levels of severity on the lifetime NODS in terms of demographics, gambling
participation and other important correlates of problem and pathological
gambling.
Table 11 shows that, as in many other
jurisdictions, problem and at-risk gamblers in New Mexico are demographically distinct
from non-problem gamblers. At-risk and
problem gamblers in New Mexico are significantly more likely than non-problem
gamblers to be male, Hispanic, unmarried and disabled or unemployed. At-risk and problem gamblers in New Mexico
are significantly less likely than non-problem gamblers to have graduated from
college. Problem gamblers are
significantly less likely than non-problem and at-risk gamblers to be
retired.
Table
11: Demographics of Non-Problem, At-Risk and
Problem Gamblers
|
|
|
Non-Problem
Gamblers
(2180)
%
|
At-Risk
Gamblers
(182)
%
|
Problem &
Pathological Gamblers
(60)
%
|
Sig.
|
|
|
|
|
|
|
|
|
Gender
|
Male
|
49.6
|
62.6
|
60.0
|
.001
|
|
|
Female
|
50.4
|
37.4
|
40.0
|
|
|
|
|
|
|
|
|
|
Age
|
18 – 34
|
30.2
|
33.1
|
42.1
|
.173
|
|
|
35 – 54
|
32.7
|
32.0
|
35.1
|
|
|
|
55 +
|
37.2
|
34.8
|
22.8
|
|
|
|
|
|
|
|
|
|
Ethnicity
|
White
|
46.4
|
47.5
|
28.3
|
.012
|
|
|
Hispanic
|
39.8
|
44.2
|
56.7
|
|
|
|
Other*
|
13.8
|
8.3
|
15.0
|
|
|
|
|
|
|
|
|
|
Marital
Status
|
Married
|
63.5
|
59.7
|
41.7
|
.005
|
|
|
Widowed
|
6.3
|
8.7
|
6.7
|
|
|
|
Divorced/Separated
|
12.9
|
14.4
|
15.0
|
|
|
|
Never Married
|
17.2
|
17.7
|
36.7
|
|
|
|
|
|
|
|
|
* Includes Native
American, African American and Other.
Table 11: Demographics
of Non-Problem, At-Risk and Problem Gamblers (cont’d)
|
|
|
Non-Problem
Gamblers
(2180)
%
|
At-Risk
Gamblers
(182)
%
|
Problem &
Pathological Gamblers
(60)
%
|
Sig.
|
|
|
|
|
|
|
|
|
Education
|
Less than HS
|
9.0
|
11.6
|
8.5
|
.001
|
|
|
HS Graduate
|
26.7
|
29.8
|
44.1
|
|
|
|
Some College
|
31.4
|
29.3
|
42.4
|
|
|
|
BA Degree
|
19.3
|
19.3
|
3.4
|
|
|
|
Graduate
Study
|
13.6
|
9.9
|
1.7
|
|
|
|
|
|
|
|
|
|
Employment
|
Working Full
Time
|
52.8
|
50.0
|
45.8
|
.000
|
|
|
Working Part
Time
|
11.3
|
12.6
|
16.9
|
|
|
|
Keeping House
|
6.9
|
4.9
|
3.4
|
|
