Introduction

Prevalence studies conducted in most jurisdictions reveal that approximately 1–2% of the population suffer from pathological gambling (Shaffer & Hall, 2001). Although pathological gambling is wide-spread, very few gamblers seek professional help for this problem. According to the National Gambling Impact Study Commission (1999), 97% of problem gamblers in the United States do not seek treatment. This low rate could be explained by the fact that many pathological gamblers have a tendency to deny their problem. For some, lack of available effective treatment or treatments adapted to the gambler’s needs could explain their resistance to seeking help (Petry, 2005).

In addition to professional treatments offered, other resources are available. Amongst these resources are self-help groups (GA) that give participants the opportunity to discuss their difficulties and discover a way to remain abstinent. Yet some pathological gamblers solve their gambling problems without undertaking any specific steps. It is estimated that about 39% of people who have had gambling problems are successful in solving them by themselves. These gamblers are considered “naturally recovered” (Hodgins, Wynne, & Makarchuk, 1999). Natural recovery seems to be a common process amongst gamblers, as it is with individuals suffering from substance addiction (Hodgins et al., 1999; Nathan, 2003). However, it seems that recovered gamblers who succeed without specific professional help have less severe gambling problems than those who have undergone formal therapy (Hodgins et al., 1999).

For quite a long time, the gambling industry has offered self-exclusion programs. Such programs are designed to limit access to gaming opportunities and provide problem gamblers help to cease or limit their gambling behaviour (Blaszczynski, Ladouceur, & Nower, 2004).

Self-exclusion services are relatively well known and are available in many casinos. Nonetheless, the characteristics of people who have used these services and the effectiveness of these programs are still not well known. Few self-exclusion programs have been evaluated and their long-term impact deserves further study.

Self-exclusion is an industry-based program allowing individuals to sign an agreement to ban themselves from entering, or allow themselves to be removed from, specified gaming venues. The ban may be for a limited time, for example from 6 months to 5 years, or a lifetime. When individuals decide to be self-excluded, they sign an agreement not to enter the gambling venue for a fixed period of time. With this agreement, gamblers agree to being refused entry into the casino, and being asked to leave if identified (Blaszczynski et al., 2004).

Effectiveness of Self-Exclusion Programs

To date, two studies have evaluated the effectiveness of self-exclusion programs. In Ladouceur, Jacques, Giroux, Ferland, and Leblond (2000) conducted an empirical study of 220 individuals who excluded themselves from the Province of Quebec’s casinos. 62% of this group were males, with an average age of 41. Most participants (95%) were identified as pathological gamblers. The self-exclusion period varied, with 66% barring themselves for 12 months or less and 25% for 60 months. Almost all participants (97%) reported confidence that they would succeed in staying away from casinos during the self-exclusion period. However, 36% revealed that they had breached the self-exclusion agreement by entering the casino. They went back an average of 6 times during their self-exclusion period. In addition, 50% reported having gambled on other games, such as video-lottery games, during their self-exclusion period. Finally, the most significant finding of the Ladouceur et al. study (2000) is that 30% of the participants complied with their initial agreement and remained abstinent during their self-exclusion period.

In Australia another study aimed to evaluate the impact of self-exclusion programs (O’Neil et al., 2003). Between 1997 and 2002, 4,083 interviews were conducted with gamblers. About half (56%) chose to self-exclude from an average number of 16.4 specified venues for an average period of 1.7 years. Of 933 individuals in the self-exclusion program between 1996 and 2002, 137 (15% of the population of self-excluded gamblers) were detected breaching their self-exclusion orders. These individuals reported an average of 3.2 breaches per person. Between 0.4 and 1.5% of problem gamblers utilise self-exclusion programs in Australia (O’Neil et al., 2003).

The study conducted by O’Neil et al. (2003) suffers from many methodological flaws. First, the authors provide insufficient information on the data collection procedures, the sample recruited, and the overall response rate. It is therefore impossible to determine whether their data are representative and generalizable. In addition, no behavioural measures were used to assess its effectiveness; the impact of the program relies exclusively on the participants’ self reports and detected breaches. These issues reduce the validity and reliability of the data.

The present study has two main goals: (1) to assess changes in gambling behaviour and problems of self-excluded patrons, and (2) to follow self-excluded gamblers for 2 years (during and after the self-exclusion period). More specifically, this study focuses on the following questions: (1) Will gamblers change their gambling pattern during the self-exclusion period? (2) How will gamblers cope with their decision? (3) What happens at the end of the self-exclusion period?

Method

Participants

A total of 161 individuals who excluded themselves from a Quebec casino participated in the study. They were recruited at the time they signed the self-exclusion agreement. Participants received no monetary compensation for their participation.

Procedure

Among different gambling options in the province of Quebec (such as lotteries, VLT, Bingo, Horse betting), there are three casinos (Montreal, Gatineau, and Charlevoix) in which a self-exclusion program is offered. Self-exclusion periods range from 6 months to 2 years. The program described in this study is run by each casino’s security department. It is publicized through a pamphlet available in different areas of the casino. Individuals who decided to exclude themselves approached a security agent who took them into a private office. Then, the security agent invited the person to participate in this study to evaluate the self-exclusion program. When a patron agreed, the security agent faxed us the name and phone number of the self-excluded patron who was called within the next few days. The participant was then contacted every 6 months for 2 years, for a total of 5 interviews. The participants were divided into three groups, according to the length of the self-exclusion period they selected themselves: 6, 12 and 24 months. The graph below illustrates the design used in this study.

The interviewers were either experienced clinical psychologists or graduate students in psychology. They received prior training and were supervised by the first author during the course of the study. The interview lasted 30–45 min.

Instruments

The questionnaire included 4 sections: (1) the first section examined the motives for self-exclusion, the triggers that led to this decision and certain elements related to the person’s gambling history; (2) the second section assessed the urge to gamble with a 10 point likert scale, consequences of gambling, confidence in the success of the self-exclusion program and compliance in the program; (3) the third section concentrated on an assessment of pathological gambling using The South Oaks Gambling Screen for a 6-month time frame (SOGS) (Lesieur & Blume, 1987) and the DSM-IV criteria for pathological gambling (APA, 1994); and (4) the fourth section was aimed at collecting socio-demographic data. During the 6, 12, 18 and 24-month follow-ups, the same questionnaires were used, except for the first section.

Results

Descriptive Results

A total of 239 files of individuals who met with a casino security agent were transmitted to the researchers. During the first telephone contact, 161 people agreed to participate in the study. The others were no longer interested in participating or could not be reached, despite 15 attempts made to contact them. 60% of the sample was men. The average age of participants was 43.5 years (SD = 12.3). Almost half (45%) had completed high school while 20 and 26.3%, respectively, held college or university degrees. The majority (72.6%) were employed. 15.8% had a household income of $25,000 or less, 34% earned between $25,000 and $50,000, and 43.4% had an income of over $50,000. More than half (56.9%) of the participants were married or living with a partner.

This was the first self-exclusion contract for all the participants. A third of the participants (33.3%) excluded themselves for a period of 6 months, 45.9% for 12 months and 20.8% for 24 months or more. They heard about the program through friends or relatives (45.3%), the media (18.2%) or information available at the casinos (20.1%). For the majority (90.7%), a specific event motivated their decision to enrol in the self-exclusion program. For 74.5% of the participants, financial problems constituted the main reason for excluding themselves. It was a personal decision for 83.8%. In addition, when asked why they had chosen to exclude themselves rather than reduce their gambling or stop gambling on their own, 88.2% of participants reported that they had no control over their gambling and felt that they could not do it on their own.

Ninety percent of the participants preferred casino games. Amongst them, 41% preferred them because of the atmosphere and 20.4% believed that the probability of winning was higher at casinos. Most popular were electronic gaming machines (60.9%), blackjack (16.8%) and roulette (9.9%). To the question “Why do you gamble?”, 69.5% mentioned the possibility of monetary gain, 39.2% the fun of the game and 18.1% reported they gamble to get away or to escape boredom.

Problem Gambling and Negative Consequences

According to the SOGS, 88.8% met the criteria for pathological gambling, 6.8% were considered at-risk gamblers and 4.3% had no gambling problems. Pathological gambling was also assessed with DSM-IV criteria. According to the DSM-IV, 73.1% of the participants were pathological gamblers. Some participants may have been at risk of becoming a pathological gambler, and decided to self exclude themselves in a preventive way. A small percentage (2.6%) reported not having lost any money in a casino, while 50.3% had lost more than $25,000. In addition, 60.5% of the participants had borrowed money to gamble during the past 6 months.

The urge to gamble over the past 6 months was reported to be at a very high level for 81.4% of the participants. Moreover, 65.4% considered themselves to have very little or no control over their gambling habits during the same period. Gambling activities had various negative consequences for the participants. 65.8% believed that gambling interfered greatly with their daily activities. Specifically, some felt that gambling had a great impact on their social life (32.7%), their work (10.9%) and their mood (27.9%). A little over a third of the participants (37.5%) had suicidal ideas during the last 6 months. For 58.1%, this train of thought was related to their gambling problem. Finally, 6.5% had attempted suicide during the same period.

Level of Confidence in the Self-Exclusion Program

During the first interview, 62% of participants believed that self-exclusion would be an effective program and 79.8% thought that taking this step will be a great way to help themselves. This step brought a feeling of freedom and peace of mind for 44.1% of participants, an improvement in their financial situation for 19.8%, a feeling of pride and accomplishment for 18%, and achievement of abstinence for 16.7%. Finally, 11.8% of participants were able to take a step back and realize that they wanted to stop gambling. Some participants (19.5%) stated that self-exclusion would change their gambling habits outside the casinos. However, 45.3% intended to return to a casino once their self-exclusion period was over. Among the latter, 29.1% hoped to do this in the context of vacationing and recreation.

Impact of the Program

During the follow-ups, some patrons were no longer interested in continuing their participation in the study and others could not be reached. The analyses for the 6-month follow-up interview included 117 participants. The 12, 18 and 24-month follow-ups had 83, 60 and 53 participants respectively. The urge to gamble was significantly reduced over time F(4, 45) = 18.41, P < .001. Analysis revealed that this reduction happened between the first interview and the 6-month follow-up, F(1, 48) = 54.262, P < .001 (see Fig. 1). The perception of control increased significantly for all participants, F(4, 44) = 23.879, P < .001. Again, this change happened between the first interview and the 6-month follow-up, F(1, 47) = 66.974, P < .001 (see Fig. 2).

Fig. 1
figure 1

Urge to Gamble over a 2 year period according to the 6, 12 and 24 months of self-exclusion groups

Full size image
Fig. 2
figure 2

Perception of Control Over Gambling Behaviour over a 2 year period according to the 6, 12 and 24 months of self-exclusion groups

Full size image

The intensity of negative consequences for gambling was significantly reduced over time for daily activities, F(4, 45) = 13.776, P < .001, social life, F(4, 43) = 3.262, P < .05, and work, F(4, 44) = 3.295, P < .05 as well as mood, F(4, 44) = 3.516, P < .05. These changes were significant between the first interview and the 6-month follow-up for the majority of the categories mentioned (daily activities, F(1, 48) = 44.088, P < .001; work, F(1, 47) = 9.086, P < .01; and mood, F(1, 47) = 5.616, P < .05), with the exception of social life (P = 0.52).

The SOGS score was significantly reduced over time, F(4, 36) = 17.948, P < .001. This change was observed between the first interview and the 6-month follow-up, F(1, 39) = 59.984, P < .001. A similar trend was observed for the DSM-IV score, F(4, 45) = 29.635, P < .001. This reduction also took place between the baseline and the 6-month follow-up, F(1, 48) = 56.977, P < .001 (see Fig. 3).

Fig. 3
figure 3

Proportion of pathological gamblers (DSM-IV) according to the 6, 12 and 24 months of self-exclusion groups

Full size image

Several participants returned to gamble at a casino during the self-exclusion period. At the 6-month follow-up, 40.5%, 42.3%, and 22.2% of the self-excluded patrons had returned to a casino (6, 12, and 24 months groups respectively). At the 12 month follow-up, self exclusion was still active for the 12 and 24 months groups. Results revealed that 55.3% and 10.5% of the 12 and 24 month groups breached. Finally, at the 18 month follow-up, 26.6% of the 24 month group had returned to a casino. Furthermore, some participants were not flagged or identified when they breached their contract. Amongst those who did not return to a casino, 45.3% said that they had decided to respect their commitment. For 38.5% of participants, the idea of being caught during the self-exclusion period did not invoke any particular feelings, although 34.1% stated that they would feel shame, guilt and humiliation if they returned and were caught.

During the first interview, 93% of the participants revealed that they would opt for self-exclusion again, given the opportunity. These percentages remained relatively high at the 6-month follow-up (77.5%), the 12-month follow-up (79.7%) and the 18-month follow-up (79.1%), but then the percentage falls to 50.3% by the 24-month follow-up interview.

Next, those for whom the self-exclusion period was still in effect were compared to other participants. Some differences emerged on several variables. Thus, in contrast to those who were no longer bound by the self-exclusion contract, people still enrolled in the program had a greater perception of control, F(1, 56) = 4.098, P < .05; they considered that gambling was interfering less with their daily activities, F(1, 57) = 4.495, P < .05 and they believed more strongly in the self-exclusion program’s effectiveness in helping gamblers, F(1, 54) = 4.006, P = .05. Furthermore, those whose self-exclusion period was still in effect were more convinced that the program had helped them, F(3, 53) = 3.172, P < .05.

Discussion

This study showed that the majority of participants in the self-exclusion program were pathological gamblers. Most had a problem with electronic gaming machines and had either very little or no control over their gambling. This program mainly targeted problem gamblers having difficulties controlling their gambling habits.

The self-exclusion program had many positive results. On several variables in the study, changes were produced during the first six months, that is, between the first interview and the 6-month follow-up. For example, major improvements were recorded on the urge to gamble, perceived control over gambling, and the intensity of the negative consequences of gambling on daily activities, social life, work and mood. SOGS scores as well as DSM-IV scores were reduced. In light of these changes, it can be concluded that the self-exclusion program had a beneficial impact.

However, some issues need to be addressed. The study demonstrated that, over time, participants seemed to perceive the self-exclusion program as less effective in helping gamblers. In addition, by the 6-month follow-up interview, more than half of the participants for whom the self-exclusion program was still in effect had returned to a casino or breached their contracts. Moreover, several participants were not identified when entering a casino. It appears that the detection system needs to be greatly improved. Furthermore, people subscribing to self-exclusion services sometimes have unclear expectations. They ask themselves: “do I have to stop or will someone stop me if I come back?” Several authors agree that responsibility for self-exclusion lies with both the gambling industry and the individual who signs the agreement. According to Nowatzki and Williams (2002), stringent application of the self-exclusion program seems to be a universal problem in the gambling industry. It is very difficult for casino security agents to memorize thousands of individuals’ pictures. In addition, it is not rare for some gamblers to change their appearance; identification is then much more difficult, if not simply impossible. These authors recommend a computerized identity verification system. However, this would imply a fundamental change in government’s involvement in gambling. Collins and Kelly (2002) suggest keeping a register of self-excluded persons, which would be made available to all sectors of the industry because they are responsible for ensuring compliance. Verifying everyone’s identity would resolve the problem but is contrary to the prevailing values of North America, Australia and New Zealand.

Blaszczynski et al. (2004) recognize that monitoring systems are imperfect and breaches will occur but contend that venues should be accountable through public reporting of data. They have adequate staff training and the financial means, as well as the surveillance necessary, to provide a workable infrastructure for gamblers who desire to self-exclude. According to Collins and Kelly (2002), self-exclusion should be considered a resource available to gamblers. Thus, once gamblers have excluded themselves, they must understand that they have to assume responsibility for their problem, rather than blame it on the gambling industry. The model proposed by Blaszczynski et al. (2004) places the primary responsibility on the gambler but recognizes the ongoing need for gaming venues to participate actively in self-exclusion programs and provide vigilant, continuous support to participants.

Another question must be raised: “How intrusive or repressive should we become?” In Quebec, there is no penalty for individuals who return to a casino during the duration of the self-exclusion program. In some cases, the lack of penalties is related to the frequency with which some people return to a casino despite their self-exclusion contract (Nowatzki & Williams, 2002). These authors, as well as Blaszczynski et al. (2004) believe that gambling venues should be subjected to some form of penalty for non-compliance. However, it is not clear what that penalty should be. In their recommendations, Collins and Kelly (2002) suggest sanctions such as confiscating wins, fines and even arrests for trespassing. Nevertheless, Nowatzki and William (2002) believe that this latter sanction would make their behaviour criminal and that it would disregard the compulsive nature of their gambling problem. This question of punishment could be avoided if a computerized registration system were developed.

What are the criteria that would determine whether a program is effective? In this study, several changes were observed in the participants. However, some returned to gambling. Does this mean that the program was not effective? According to Blaszczynski et al. (2004), even if some people return to a casino, despite having been self-excluded, this course of action could curtail gambling activity and thus lead to a potential improvement in behaviour control and a reduction in harm amongst more gamblers.

How long should the program last? The results of this study show that gamblers still enrolled in the program believe more strongly that this course of action has helped them. How can these results be explained? Perhaps they perceive the self-exclusion contract to be a symbolic commitment and are therefore more motivated to refrain from gambling again. At present, a person can exclude himself for a period of 6 months to 5 years. Collins and Kelly (2002) suggest that self-exclusion be offered for a period of one year and that it be renewable either by prepaid postcard, e-mail or letter, so that the client does not have to return to the gambling establishment once the contract has ended. These authors also suggest that a longer period, over one year in duration, could have negative effects, in that the person might end up changing his mind and wanting to revoke his contract. Nowatzki and Williams (2002) believe that such programs will have little value if it is possible to change the conditions at any time. We believe that the self-exclusion contract should be irreversible and irrevocable. The longer exclusion could help reduce the risk of a relapse. However, there is little empirical data about the period of abstinence required to prevent a relapse. Blaszczynski et al. (2004) suggest leaving the choice of duration to the discretion of the gambler, while establishing a minimum period of one year and a maximum of 5 years.

Who should be in charge of the program: casino employees, independent experts or both? Operators should play a more active role in self-exclusion in order to optimize its effectiveness. Collins and Kelly (2002) believe that the casinos’ responsibility towards the gambler is to do everything possible to help him succeed in his course of action. An important element is to stop all forms of direct advertising to the client. Nonetheless, they feel that the casinos should not be responsible for the application of the self-exclusion program. Rather, this responsibility should be given to an independent organization in charge of establishing a national register distributed to all sectors of the industry. According to Blaszczynski et al. (2004), the gambling industry must participate actively in the self-exclusion program and ensure that they provide constant support to clients. Casinos should (1) train their employees and put in place the necessary infrastructure to support the entire self-exclusion process; (2) publicize the self-exclusion service in order to make it accessible and demonstrate the value of such a course of action with empirical data; and (3) develop and establish an identification protocol for individuals who do not comply with their contract. In conclusion, the self-exclusion program needs systematic monitoring and a continuous evaluation of the outcome. In order to do this, casinos would do well to co-operate with independent evaluators who would periodically and randomly verify adherence to the program. This final measure would increase understanding of the process and would motivate the gambling industry to maintain its commitment.